openSUSE Build Service

Overview Repositories Revisions Requests Users Attributes Meta

File wicked.changes of Package wicked (Revision 9838603e0c85d662a34cae1bce75a742)

Currently displaying revision 9838603e0c85d662a34cae1bce75a742 , Show latest

-------------------------------------------------------------------
Fri May 17 17:23:22 CEST 2013 - olaf@aepfle.de

- Update prototype of ni_objectmodel_signal_to_event
  main.c: In function ‘ni_nanny_netif_state_change_signal_receive’:
  main.c:204:2: warning: comparison of unsigned expression < 0 is always false
  main.c: In function ‘ni_nanny_modem_state_change_signal_receive’:
  main.c:292:2: warning: comparison of unsigned expression < 0 is always false
- commit a3396b7

-------------------------------------------------------------------
Fri May 17 17:00:12 CEST 2013 - olaf@aepfle.de

- Correct ni_sysconfig_get_boolean usage in compat-redhat.c
  compat-redhat.c: In function ‘try_vlan’:
  compat-redhat.c:354:2: warning: comparison is always false due to limited range of data type
- commit 987951e

-------------------------------------------------------------------
Fri May 17 16:57:01 CEST 2013 - olaf@aepfle.de

- Fix logic error in __ni_redhat_define_interface
  compat-redhat.c: In function ‘__ni_redhat_read_interface’:
  compat-redhat.c:150:2: warning: comparison is always false due to limited range of data type
- commit 6b364a1

-------------------------------------------------------------------
Fri May 17 16:50:35 CEST 2013 - olaf@aepfle.de

- Fix warning in ni_ifcondition_type
  fsm-policy.c: In function ‘ni_ifcondition_type’:
  fsm-policy.c:1086:2: warning: comparison of unsigned expression < 0 is always false
- commit 0b3e929

-------------------------------------------------------------------
Fri May 17 16:47:18 CEST 2013 - olaf@aepfle.de

- Correct return type of ni_ifworker_type_from_string
  fsm.c: In function ‘ni_fsm_workers_from_xml’:
  fsm.c:975:3: warning: comparison of unsigned expression < 0 is always false
- commit 9f04805

-------------------------------------------------------------------
Fri May 17 14:32:41 CEST 2013 - mt@suse.de

- Fixed to build noinst-testing with RPM_OPT_FLAGS
- commit 115f5bf

-------------------------------------------------------------------
Fri May 17 14:19:59 CEST 2013 - mt@suse.de

- Defined to own also the datadir/wicked dir in spec
- commit da21266

-------------------------------------------------------------------
Fri May 17 13:42:09 CEST 2013 - mt@suse.de

- Added infiniband child pkey check to validate
- commit 235125d

-------------------------------------------------------------------
Fri May 17 13:38:45 CEST 2013 - mt@suse.de

- Fixed infiniband sign warnings
  Further, added defines with default/unset settings.
  infiniband.c:108:15: warning: comparison between signed and unsigned integer expressions
  infiniband.c:110:17: warning: comparison between signed and unsigned integer expressions
- commit a516731

-------------------------------------------------------------------
Fri May 17 12:47:02 CEST 2013 - mt@suse.de

- Fixed sign warning in __ni_lease_owns_address
  interface.c:513:30: warning: comparison between signed and unsigned integer expressions
- commit 103932a

-------------------------------------------------------------------
Fri May 17 12:43:05 CEST 2013 - mt@suse.de

- Fixed warning in ni_dhcp_set_client_id
  device.c:645:19: warning: comparison between signed and unsigned integer expressions
- commit ddbfce1

-------------------------------------------------------------------
Fri May 17 12:39:20 CEST 2013 - mt@suse.de

- Fixed warnings in ni_stringbuf_trim_empty_lines
  util.c:1154:23: warning: comparison between signed and unsigned integer expressions
  util.c:1602:16: warning: comparison between signed and unsigned integer expressions
- commit 9686a48

-------------------------------------------------------------------
Fri May 17 12:37:50 CEST 2013 - mt@suse.de

- Fixed sign warning in __ni_rfkill_recv
  rfkill.c:92:9: warning: comparison between signed and unsigned integer expressions
- commit 8e40624

-------------------------------------------------------------------
Fri May 17 12:33:35 CEST 2013 - mt@suse.de

- Changed int types to unsigned in ni_capture_devinfo
  capture.c:543:22: warning: comparison between signed and unsigned integer expressions
  capture.c:547:23: warning: comparison between signed and unsigned integer expressions
- commit 342f6f6

-------------------------------------------------------------------
Fri May 17 12:26:27 CEST 2013 - mt@suse.de

- Avoid warnings by consistent unsigned ifindex use
  ifevent.c:187:42: warning: comparison between signed and unsigned integer expressions
  iflist.c:190:37: warning: comparison between signed and unsigned integer expressions
- commit 4c2f5eb

-------------------------------------------------------------------
Fri May 17 12:24:57 CEST 2013 - mt@suse.de

- Fixed warning in ni_duid_parse_hex
  duid.c:197:10: warning: comparison between signed and unsigned integer expressions
- commit 1a099bf

-------------------------------------------------------------------
Fri May 17 11:28:51 CEST 2013 - olaf@aepfle.de

- Fix warming in ni_bridge_port_validate
  bridge.c:341:5: warning: comparison of unsigned expression < 0 is always false
- commit 5bb045d

-------------------------------------------------------------------
Fri May 17 11:19:09 CEST 2013 - olaf@aepfle.de

- Make NI_BRIDGE_VALUE_NOT_SET unsigned
  bridge.c: In function ‘ni_bridge_port_validate’:
  bridge.c:340:21: warning: comparison between signed and unsigned integer expressions
  bridge.c:345:22: warning: comparison between signed and unsigned integer expressions
- commit a15da67

-------------------------------------------------------------------
Thu May 16 22:48:24 CEST 2013 - mt@suse.de

- Fixed also ni_sockaddr_format to not write crap
- commit 8d43257

-------------------------------------------------------------------
Thu May 16 17:58:31 CEST 2013 - olaf@aepfle.de

- Fix warning in ni_bridge_del_port_ifindex
  bridge.c:186:21: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
- commit 3d0caaa

-------------------------------------------------------------------
Thu May 16 17:55:19 CEST 2013 - olaf@aepfle.de

- Fix warning in ni_bonding_set_option
  bonding.c:1157:43: warning: 'tmp' may be used uninitialized in this function [-Wuninitialized]
- commit 5242fd0

-------------------------------------------------------------------
Thu May 16 17:52:59 CEST 2013 - olaf@aepfle.de

- Fix warning in ni_sockaddr_print
  ordered comparison of pointer with integer zero
- commit 026e0d7

-------------------------------------------------------------------
Thu May 16 15:38:16 CEST 2013 - olaf@aepfle.de

- Move schema files to /usr/share/wicked
- commit 874c975

-------------------------------------------------------------------
Thu May 16 15:32:52 CEST 2013 - mt@suse.de

- Initial systemd service units for >= 12.3 only
  The supplicants are started as part of wicked daemon,
  disabled installation of dbus autostart services.
- commit b0599fb

-------------------------------------------------------------------
Thu May 16 12:34:08 CEST 2013 - mt@suse.de

- Removed --decorate=short from git log in make dist
  Now the ChangeLog.git inside of the tar archive is
  using the default git log format.
- commit ca95894

-------------------------------------------------------------------
Thu May 16 11:38:18 CEST 2013 - olaf@aepfle.de

- Rename *-supplicant to wicked-* in dbus sevice files
- commit 6a89b04

-------------------------------------------------------------------
Thu May 16 11:00:15 CEST 2013 - olaf@aepfle.de

- Rename *-supplicant to wicked-*
- commit 4c9200c

-------------------------------------------------------------------
Thu May 16 10:00:54 CEST 2013 - olaf@aepfle.de

- Use port interface when applying sysfs values
- commit 979f7db

-------------------------------------------------------------------
Wed May 15 17:45:43 CEST 2013 - olaf@aepfle.de

- package libexec subdirs
- commit 3e78644

-------------------------------------------------------------------
Wed May 15 17:07:09 CEST 2013 - olaf@aepfle.de

- move supplicants from sbindir to libexecdir/wicked/bin
- commit e41a800

-------------------------------------------------------------------
Wed May 15 16:52:15 CEST 2013 - olaf@aepfle.de

- remove unused wicked_devellibdir
- commit 1b3fbb6

-------------------------------------------------------------------
Wed May 15 16:16:50 CEST 2013 - olaf@aepfle.de

- remove usage of absolute paths to sysvinit-tools
- commit 50e9d8b

-------------------------------------------------------------------
Wed May 15 15:58:22 CEST 2013 - olaf@aepfle.de

- specfile: replace _devellibdir wit _libdir
- commit 1cf5369

-------------------------------------------------------------------
Wed May 15 15:46:42 CEST 2013 - olaf@aepfle.de

- install binaries into /usr
- commit c9af35d

-------------------------------------------------------------------
Wed May 15 15:38:18 CEST 2013 - olaf@aepfle.de

- autogen.sh: use /usr as default when building from scratch
- commit c537b55

-------------------------------------------------------------------
Tue May 14 16:17:03 CEST 2013 - mt@suse.de

- Validate infiniband settings, fixed mode in schema
- commit e5efa67

-------------------------------------------------------------------
Tue May 14 15:37:23 CEST 2013 - mt@suse.de

- Use infiniband-child xml config node for children
- commit a8f1e22

-------------------------------------------------------------------
Mon May 13 20:39:19 CEST 2013 - mt@suse.de

- Updated infiniband interface xml schema
- commit 54923c6

-------------------------------------------------------------------
Mon May 13 20:32:01 CEST 2013 - mt@suse.de

- Infiniband dbus objectmodel implementation
- commit 5c07400

-------------------------------------------------------------------
Mon May 13 20:12:18 CEST 2013 - mt@suse.de

- Infiniband interface sysfs setup and discovery
- commit aabfa91

-------------------------------------------------------------------
Mon May 13 20:10:00 CEST 2013 - mt@suse.de

- Added ni_sysfs_netif_exists utility
- commit 7110136

-------------------------------------------------------------------
Mon May 13 20:09:08 CEST 2013 - mt@suse.de

- Added ni_sysfs_netif_printf utility
- commit dd118b5

-------------------------------------------------------------------
Mon May 13 20:07:20 CEST 2013 - mt@suse.de

- Added infiniband ifcfg parsing and xml generation
- commit f2dfec6

-------------------------------------------------------------------
Mon May 13 20:04:41 CEST 2013 - mt@suse.de

- Added initial infiniband structure to netdev
- commit feaa5e5

-------------------------------------------------------------------
Mon May 13 19:37:08 CEST 2013 - mt@suse.de

- Adopted vlan to use new netdev_ref bind/unbind
- commit 1117cad

-------------------------------------------------------------------
Mon May 13 19:36:15 CEST 2013 - mt@suse.de

- Added netdev_ref_t type and bind/unbind helpers
- commit 09737af

-------------------------------------------------------------------
Mon May 13 12:55:46 CEST 2013 - mt@suse.de

- Fixed ifcfg parser to not fail on empty routes file
- commit cd177f0

-------------------------------------------------------------------
Mon May 13 12:54:34 CEST 2013 - mt@suse.de

- Fixed man page file names in RPM file lists
- commit 67eb3f2

-------------------------------------------------------------------
Mon May 13 12:04:44 CEST 2013 - olaf@aepfle.de

- Check for iwlib.h instead of wireless.h
  wireless.h depends on other includes, use iwlib.h to fix configure
  "present but cannot be compiled" errors.
- commit 77417b0

-------------------------------------------------------------------
Mon May 13 11:50:12 CEST 2013 - olaf@aepfle.de

- Add configure check for wireless.h
- commit 9f5e8ae

-------------------------------------------------------------------
Mon May  6 22:39:39 CEST 2013 - mt@suse.de

- Fixed utils/schema2html build
- commit 33d6343

-------------------------------------------------------------------
Mon May  6 18:48:22 CEST 2013 - olaf@aepfle.de

- fix typo in schema/bonding.xml description
- commit 1393a4d

-------------------------------------------------------------------
Fri Apr 26 09:47:10 CEST 2013 - mt@suse.de

- Corrected to return error on ifcfg parsing failures
- commit 8faf843

-------------------------------------------------------------------
Fri Apr 26 09:37:39 CEST 2013 - mt@suse.de

- Fixed bridge port validation
- commit 3ab5a3c

-------------------------------------------------------------------
Fri Apr 26 09:33:09 CEST 2013 - mt@suse.de

- Fixed bridge port priority validation
- commit cc065b5

-------------------------------------------------------------------
Fri Apr 26 08:52:00 CEST 2013 - mt@suse.de

- Consistently use x-alloc variants in string utils
- commit f53da15

-------------------------------------------------------------------
Thu Apr 25 15:52:52 CEST 2013 - mt@suse.de

- Renamed format/parse_int to uint, added parse long
  Added parse_long/ulong utilities used in parse_int/uint functions,
  that accept only bare numbers (and minus) now and make more input
  and limit checks to catch e.g. parsing of -5 into a uint.
  Renamed the format_int and parse_int_mapped to _uint_ as they are.
- commit c763a5a

-------------------------------------------------------------------
Thu Apr 25 08:47:17 CEST 2013 - mt@suse.de

- Added bridge validate, enabled for dbus and ifcfg
- commit 7d9a8a6

-------------------------------------------------------------------
Thu Apr 25 08:46:15 CEST 2013 - mt@suse.de

- Parse/write sysfs bridge ageing time as ulong
- commit eea2c6a

-------------------------------------------------------------------
Tue Apr 23 18:25:38 CEST 2013 - mt@suse.de

- Adopted and improved the bridge xml schema
  Now, there is a configuration-only dict used for changeDevice
  and a properties dict that is returned as interface status.
- commit 1a8c4f1

-------------------------------------------------------------------
Tue Apr 23 18:22:34 CEST 2013 - mt@suse.de

- Completed bridge options parsing and ifcfg converter
  Splitted bridge stp into enablement and state, changed
  time values to use double, add ports in bridge_setup.
- commit ad80b0d

-------------------------------------------------------------------
Wed Apr 17 18:04:10 CEST 2013 - mt@suse.de

- Added xml deserialization element-name tag to arrays
  When the array definition does not specify an element-name attibute, the
  name of the element type is used for xml deserialization / displaying
  purposes with fallback to anonymous "e" element nodes.
- commit 6d0ebb0

-------------------------------------------------------------------
Wed Apr 17 13:59:52 CEST 2013 - mt@suse.de

- Added int16,int64,uint64,double dbus properties
- commit ce59226

-------------------------------------------------------------------
Wed Apr 17 13:55:26 CEST 2013 - mt@suse.de

- Check for ERANGE in parse_int/parse_double
- commit 5cbd1ff

-------------------------------------------------------------------
Wed Apr 17 13:50:16 CEST 2013 - mt@suse.de

- Corrected to validate against bonding defaults
- commit 60dea68

-------------------------------------------------------------------
Tue Apr 16 09:10:27 CEST 2013 - mt@suse.de

- Added ni_string_printf utility function
- commit 067ea91

-------------------------------------------------------------------
Fri Apr 12 16:14:55 CEST 2013 - mt@suse.de

- Bonding mode deps, unavailable slaves, all attrs
- commit 6dbe396

-------------------------------------------------------------------
Thu Apr 11 14:12:02 CEST 2013 - mt@suse.de

- Destroy arrays in ni_sysfs_bonding_set_list_attr
  Further, changed to add before deleting attrs.
- commit 05d4618

-------------------------------------------------------------------
Thu Apr 11 13:48:39 CEST 2013 - mt@suse.de

- Removed ni_sysfs_bonding_get_slaves declaration dup
- commit 76f90ec

-------------------------------------------------------------------
Thu Apr 11 13:40:09 CEST 2013 - mt@suse.de

- Added a ni_bonding_has_slave utility function
- commit 1536ef9

-------------------------------------------------------------------
Thu Apr 11 13:39:06 CEST 2013 - mt@suse.de

- Fixed ifup/ifdown no interface status messages
- commit a0b558a

-------------------------------------------------------------------
Thu Apr 11 11:12:03 CEST 2013 - mt@suse.de

- Improved bonding schema comments a bit
- commit d15d77b

-------------------------------------------------------------------
Thu Apr 11 10:34:19 CEST 2013 - mt@suse.de

- Fixed a typo in add bonding slave causing SEGV
- commit 8970d05

-------------------------------------------------------------------
Thu Apr 11 09:08:01 CEST 2013 - mt@suse.de

- Use shared netif-reference in bonding slave schema
- commit 65b8cdf

-------------------------------------------------------------------
Thu Apr 11 09:07:12 CEST 2013 - mt@suse.de

- Removed bond monitoring check again
- commit 43192d8

-------------------------------------------------------------------
Wed Apr 10 18:30:16 CEST 2013 - mt@suse.de

- Added a bonding target hack and schema constraints
- commit 5576807

-------------------------------------------------------------------
Wed Apr 10 18:29:11 CEST 2013 - mt@suse.de

- Added bonding properties to dbus declarations
- commit b626836

-------------------------------------------------------------------
Wed Apr 10 18:28:23 CEST 2013 - mt@suse.de

- Added more basic checks to bonding_validate
- Added more bonding setting checks
- commit a1a3bba

-------------------------------------------------------------------
Wed Apr 10 18:26:50 CEST 2013 - mt@suse.de

- Updated bonding schema, synced tags with converter
- Fixed typo in bonding arpmon target xml generation
- commit 3edf75c

-------------------------------------------------------------------
Wed Apr 10 15:06:40 CEST 2013 - mt@suse.de

- Updated bonding schema, synced tags with converter
- commit df373b9

-------------------------------------------------------------------
Wed Apr 10 11:30:57 CEST 2013 - mt@suse.de

- Completed bonding options parsing and ifcfg converter
- commit fbc93fd

-------------------------------------------------------------------
Wed Apr 10 09:39:26 CEST 2013 - mt@suse.de

- Added a limit parameter to ni_string_split
- commit 2bd96b3

-------------------------------------------------------------------
Tue Apr  9 14:39:07 CEST 2013 - mt@suse.de

- Moved modprobe bonding to ni_bonding_load function
- commit 31efd1d

-------------------------------------------------------------------
Fri Apr  5 13:01:10 CEST 2013 - mt@suse.de

- Be strict in ni_parse_int_mapped, new maybe variant
  Added ni_parse_int_maybe_mapped function, corresponding to
  ni_format_int_maybe_mapped, which is falling back to parse
  the input string as uint. On success, the uint is backward
  mapped, and 1 is returned, when it does not map back.
- commit 7827ac3

-------------------------------------------------------------------
Fri Apr  5 10:44:33 CEST 2013 - mt@suse.de

- Added a base arg to ni_parse_int, adopted sources
- commit fd831e8

-------------------------------------------------------------------
Thu Apr  4 15:15:27 CEST 2013 - mt@suse.de

- Disabled broken try to report if bond device exists
- commit 43b75c9

-------------------------------------------------------------------
Thu Apr  4 15:10:33 CEST 2013 - mt@suse.de

- Enable modprobe of bonding max_bonds=0 module
  Default is max_bonds=1 that automatically creates a bond0
  interface -- this may be not the interface name we need.
  We request the bond interface creation in separate step.
- commit 0f6a32e

-------------------------------------------------------------------
Thu Apr  4 15:09:20 CEST 2013 - mt@suse.de

- Complain about missed ifcfg file prefix separately
- commit 9a31cd0

-------------------------------------------------------------------
Thu Apr  4 15:08:25 CEST 2013 - mt@suse.de

- Fixed ifcfg BRIDGE and BONDING_MASTER checks
- commit 925f92e

-------------------------------------------------------------------
Thu Apr  4 14:55:56 CEST 2013 - mt@suse.de

- Added a helper function to call modprobe
- commit 0ca78d6

-------------------------------------------------------------------
Thu Apr  4 14:50:56 CEST 2013 - mt@suse.de

- Use shellcmd_parse with parsed cmd, _new with argv
- commit f5441ca

-------------------------------------------------------------------
Thu Apr  4 14:37:36 CEST 2013 - mt@suse.de

- Simple string split into array and join helpers
- commit 3bdb1fb

-------------------------------------------------------------------
Fri Mar 22 17:17:31 CET 2013 - mt@suse.de

- Install missing /etc/wicked/ifconfig directory
- commit 7228ed7

-------------------------------------------------------------------
Thu Mar 21 14:11:40 CET 2013 - mt@suse.de

- One call in boot mode, allow cores in init.d/network
- commit a1cc8c1

-------------------------------------------------------------------
Thu Mar 21 14:09:51 CET 2013 - mt@suse.de

- Allow core files in init.d/wickedd for now
- commit 81cdb5d

-------------------------------------------------------------------
Thu Mar 21 13:04:27 CET 2013 - mt@suse.de

- Added ni_config_sources returning sources of a type
- commit 235c496

-------------------------------------------------------------------
Thu Mar 21 11:23:18 CET 2013 - mt@suse.de

- Compat-suse ifcfg file parsing improvements
- commit c9d9bcf

-------------------------------------------------------------------
Thu Mar 21 11:11:12 CET 2013 - mt@suse.de

- ifup: fetch configured/default ifconfig sources
- commit 6ca51a1

-------------------------------------------------------------------
Thu Mar 21 11:08:07 CET 2013 - mt@suse.de

- Added default ifconfig sources to client.xml
- commit 1085f44

-------------------------------------------------------------------
Thu Mar 21 11:07:56 CET 2013 - mt@suse.de

- Parse <sources><ifconfig=... /> from the config
- commit 1e588f3

-------------------------------------------------------------------
Wed Mar 20 20:14:58 CET 2013 - olaf@aepfle.de

- move wicked(8) into client subpackage
- commit c2b8b5d

-------------------------------------------------------------------
Wed Mar 20 15:19:38 CET 2013 - mt@suse.de

- Return count of added directory in ni_scandir
- commit 304707e

-------------------------------------------------------------------
Fri Mar 15 11:44:42 CET 2013 - mt@suse.de

- Fixed allocation to format complete duid as hex
- commit b36268f

-------------------------------------------------------------------
Fri Mar 15 10:25:12 CET 2013 - mt@suse.de

- Write DHCPv6 default duid file to persistent store
- commit c94a910

-------------------------------------------------------------------
Thu Mar 14 21:29:44 CET 2013 - mt@suse.de

- Updated .gitignore file
- commit a41af70

-------------------------------------------------------------------
Thu Mar 14 21:28:15 CET 2013 - mt@suse.de

- Configure variables in all man pages
- commit 3964798

-------------------------------------------------------------------
Thu Mar 14 21:26:32 CET 2013 - mt@suse.de

- Configure variables in etc xml and dbus service files
- commit bf10fe4

-------------------------------------------------------------------
Thu Mar 14 21:17:51 CET 2013 - mt@suse.de

- Renamed all etc xml and dbus service files to *.in
- commit 80419c7

-------------------------------------------------------------------
Thu Mar 14 21:17:45 CET 2013 - mt@suse.de

- Renamed all man pages to *.in to set paths
- commit f4418b8

-------------------------------------------------------------------
Thu Mar 14 21:16:25 CET 2013 - mt@suse.de

- Added piddir and storedir parsing in common config
- commit 8ef2998

-------------------------------------------------------------------
Thu Mar 14 19:25:57 CET 2013 - mt@suse.de

- Rebased --with-schemadir,extensionsdir to configdir
- commit 1bfc39c

-------------------------------------------------------------------
Thu Mar 14 19:24:18 CET 2013 - mt@suse.de

- Added --with-piddir,statedir,storedir configure options
- commit 6720c59

-------------------------------------------------------------------
Thu Mar 14 11:32:45 CET 2013 - mt@suse.de

- Fixed compiler warnings, cleaned up a bit
- commit d908332

-------------------------------------------------------------------
Tue Mar 12 13:58:18 CET 2013 - mt@suse.de

- Added TAGS to .gitignore file
- commit d6f0f7f

-------------------------------------------------------------------
Tue Mar 12 13:42:21 CET 2013 - mt@suse.de

- Fixed to copy ifname to capture, addr-label leak
- commit 747ccbb

-------------------------------------------------------------------
Tue Mar 12 09:31:22 CET 2013 - mt@suse.de

- dhcp6: fixed reparse to use buffer not its data
- commit 47b04cf

-------------------------------------------------------------------
Sat Mar  9 15:19:18 CET 2013 - mt@suse.de

- Fixed automake link variables use, tweaked a bit
- commit 43caa74

-------------------------------------------------------------------
Fri Mar  8 13:29:47 CET 2013 - mt@suse.de

- rcnetwork: fixed ifconfig arg, reset onboot init
- commit 3559faf

-------------------------------------------------------------------
Thu Mar  7 16:28:36 CET 2013 - mt@suse.de

- For now, just use ifcfg files in /etc/init.d/network
  Updated LSB tags, removed use of sysconfig functions,
  disabled obsolete localfs and remotefs stages/options.
- commit 7ff2d15

-------------------------------------------------------------------
Thu Mar  7 11:41:52 CET 2013 - mt@suse.de

- Reset suse-dbus-unauthorized-service rpmlink error
  Added wicked-rpmlintrc to reset the badness causing
  build system to fail.
  The Security-Audit is in progress, see:
  https://bugzilla.novell.com/show_bug.cgi?id=783932
- commit ffa6e24

-------------------------------------------------------------------
Thu Mar  7 11:37:45 CET 2013 - mt@suse.de

- Removed shebang from extensions/functions
- commit cfd5320

-------------------------------------------------------------------
Wed Mar  6 17:54:27 CET 2013 - mt@suse.de

- Added rpm spec dependencies for sysconfig >= 0.81
  The sysconfig-0.81 version provides network-scripts as
  sub-package allowing us to provide /etc/init.d/network.
- commit de36bd2

-------------------------------------------------------------------
Tue Mar  5 14:15:17 CET 2013 - mt@suse.de

- Adopted wicked.spec for automake
- commit 709eee5

-------------------------------------------------------------------
Tue Mar  5 14:03:44 CET 2013 - mt@suse.de

- Adopted .gitignore for use with automake
- commit 6c6929a

-------------------------------------------------------------------
Tue Mar  5 14:02:32 CET 2013 - mt@suse.de

- Initial switch to automake instead of autoconf
- commit 58439c9

-------------------------------------------------------------------
Tue Mar  5 13:52:11 CET 2013 - mt@suse.de

- Remove also obsolete testing/ibft-test.c
- commit 77624b9

-------------------------------------------------------------------
Wed Feb 27 13:28:21 CET 2013 - mt@suse.de

- Use require-link in link-detection same to fsm
- commit 915fb78

-------------------------------------------------------------------
Wed Feb 27 12:05:27 CET 2013 - okir@suse.de

- Support compat loading of old ifcfg files
- commit 0c4d314

-------------------------------------------------------------------
Wed Feb 27 11:40:54 CET 2013 - okir@suse.de

- Remove obsolete iBFT handling code in C
- commit 67411e3

-------------------------------------------------------------------
Fri Dec 14 10:25:06 CET 2012 - mt@suse.de

- Adopted to pass program name to ni_init
- commit 1da188e

-------------------------------------------------------------------
Fri Dec 14 10:24:28 CET 2012 - mt@suse.de

- Added missed wicked/logging.h inclusion
- commit 61b4e43

-------------------------------------------------------------------
Fri Dec 14 10:19:46 CET 2012 - mt@suse.de

- Adopted use ni_sockaddr_format instead address
- commit 75fa0f0

-------------------------------------------------------------------
Wed Dec 12 16:03:51 CET 2012 - mt@suse.de

- Applied ppp username/password sanity checks
- commit e5d84fb

-------------------------------------------------------------------
Tue Nov  6 11:00:22 CET 2012 - mt@suse.de

- Install dbus system service start files
- commit 3f9717e

-------------------------------------------------------------------
Tue Nov  6 10:18:37 CET 2012 - mt@suse.de

- Applied dhcp lease object model sanity checks
- commit c90d4ff

-------------------------------------------------------------------
Fri Nov  2 14:10:26 CET 2012 - mt@suse.de

- Fixed netdev ethernet allocation from dbus dict
- commit 55ea509

-------------------------------------------------------------------
Fri Nov  2 10:25:27 CET 2012 - mt@suse.de

- Fixed netbios node type use in dhcp4 and schema
- commit 3bed070

-------------------------------------------------------------------
Fri Nov  2 10:25:01 CET 2012 - mt@suse.de

- Applied dhcp domain and path name sanity checks
- commit e9803ea

-------------------------------------------------------------------
Tue Oct 16 11:37:16 CEST 2012 - mt@suse.de

- Attempt to fix makefile rebuild / config.h deps
- commit a31b69f

-------------------------------------------------------------------
Mon Oct 15 15:12:01 CEST 2012 - mt@suse.de

- spec file: Fixed version check for libnl build-require
- commit 0782528

-------------------------------------------------------------------
Wed Oct 10 09:03:37 CEST 2012 - okir@suse.de

- ni_ipv6_ra_rdnss: use new ni_sockaddr_array rather than open coding the array functionality
- commit f1c8d84

-------------------------------------------------------------------
Wed Oct 10 08:57:45 CEST 2012 - okir@suse.de

- Introduce ni_sockaddr_arrays and functions
- commit fa5acbb

-------------------------------------------------------------------
Wed Oct 10 04:58:13 CEST 2012 - mt@suse.de

- Fixed ni_daemonize to write pid file in final process
- commit a9e749c

-------------------------------------------------------------------
Wed Oct 10 04:23:22 CEST 2012 - mt@suse.de

- Adopted to install /etc/init.d/wickedd
- commit f6f99e2

-------------------------------------------------------------------
Wed Oct 10 03:49:26 CEST 2012 - mt@suse.de

- Renamed /etc/init.d/wicked to wickedd (LSB)
- commit d9c6fd0

-------------------------------------------------------------------
Tue Oct  9 19:23:50 CEST 2012 - mt@suse.de

- Added optional rtnelink RDNSS update events
- commit 8425108

-------------------------------------------------------------------
Tue Oct  9 18:46:14 CEST 2012 - mt@suse.de

- rtnetlink: fixed message data length check
- commit 71e90e7

-------------------------------------------------------------------
Tue Oct  9 11:33:38 CEST 2012 - okir@suse.de

- Fixed SEGV caused by ni_dbus_message_get_args
- commit 7d28a8d

-------------------------------------------------------------------
Mon Oct  8 16:00:42 CEST 2012 - mt@suse.de

- Fixed ni_netdev_clear_event_filters
- commit 63df5f5

-------------------------------------------------------------------
Mon Oct  8 14:55:54 CEST 2012 - mt@suse.de

- man: cleaned up wicked man page references a bit
- commit 7d91531

-------------------------------------------------------------------
Mon Oct  8 14:55:00 CEST 2012 - mt@suse.de

- wicked.spec: create ghost run dir in rpm post
- commit f1a96f4

-------------------------------------------------------------------
Mon Oct  8 14:40:50 CEST 2012 - mt@suse.de

- makefile,spec: added configure --with-devellibdir
- commit eac1b11

-------------------------------------------------------------------
Mon Oct  8 13:59:46 CEST 2012 - mt@suse.de

- init-scripts: refer to fsf via url as in gpl v3.
- commit 4e176a3

-------------------------------------------------------------------
Mon Oct  8 13:52:22 CEST 2012 - mt@suse.de

- makefile, spec: install man5 manual pages
- commit 5447f4d

-------------------------------------------------------------------
Mon Oct  8 13:22:17 CEST 2012 - mt@suse.de

- wicked.spec: marked nanny config as config
- commit da215c2

-------------------------------------------------------------------
Mon Oct  8 12:57:31 CEST 2012 - mt@suse.de

- wicked.spec: post actions in sysvinit package now
- commit 7999d3e

-------------------------------------------------------------------
Mon Oct  8 12:07:01 CEST 2012 - mt@suse.de

- wicked.spec: Added nanny and sysvinit subpackages
- commit b69f1f3

-------------------------------------------------------------------
Mon Oct  8 12:06:06 CEST 2012 - mt@suse.de

- install: install-bin already triggers install-lib
- commit 072c048

-------------------------------------------------------------------
Mon Oct  8 11:39:20 CEST 2012 - mt@suse.de

- Handle asprintf failures
- commit 94f8204

-------------------------------------------------------------------
Mon Oct  8 11:38:48 CEST 2012 - mt@suse.de

- Removed -Werror from CFLAGS
- commit 6124f4b

-------------------------------------------------------------------
Fri Oct  5 20:39:52 CEST 2012 - mt@suse.de

- dhcp6: Enabled a dummy/tracing ra prefix handler
- commit 1649bcf

-------------------------------------------------------------------
Fri Oct  5 17:50:56 CEST 2012 - mt@suse.de

- Removed unimplemented ni_rtevent_refresh_all declaration
- commit e39015d

-------------------------------------------------------------------
Fri Oct  5 17:49:40 CEST 2012 - mt@suse.de

- Process interface ipv6 ra prefix info events
- commit dadd612

-------------------------------------------------------------------
Fri Oct  5 17:42:03 CEST 2012 - mt@suse.de

- Process ipv6 link ra info (managed/config) flags
- commit 71fc87d

-------------------------------------------------------------------
Fri Oct  5 17:40:03 CEST 2012 - mt@suse.de

- Subscribed to newlink ipv6 info rtnetlink group
- commit 25268ff

-------------------------------------------------------------------
Fri Oct  5 17:38:23 CEST 2012 - mt@suse.de

- Moved ipv6 devinfo variables to devconf sub-struct
- commit 0f3e9e5

-------------------------------------------------------------------
Thu Sep 13 21:58:55 CEST 2012 - okir@suse.de

- FSM: when starting to work on a device, clear the server side event filters
- commit feb9b92

-------------------------------------------------------------------
Thu Sep 13 21:47:16 CEST 2012 - okir@suse.de

- Get rid of dev->link_info.event_uuid
  Use the new event filter mechanism instead.
- commit 551abd3

-------------------------------------------------------------------
Thu Sep 13 19:48:01 CEST 2012 - okir@suse.de

- Introduce event_filter functionality in ni_netdev_t
  This will be used to control more closely how events are reported
  and with what uuids.
- commit 47ad48c

-------------------------------------------------------------------
Thu Sep 13 19:24:22 CEST 2012 - okir@suse.de

- Added sample file with an open WLAN config
- commit 3a939a9

-------------------------------------------------------------------
Thu Sep 13 15:59:15 CEST 2012 - okir@suse.de

- wpa-supplicant: Fix some issues when using WPA-EAP
- commit 3210071

-------------------------------------------------------------------
Thu Sep 13 15:30:45 CEST 2012 - okir@suse.de

- WLAN: do not print error messages during netdev refresh if rfkill is in effect
- commit caa5061

-------------------------------------------------------------------
Wed Sep 12 19:07:29 CEST 2012 - mt@suse.de

- dhcp4: process packets with partial checksum (xen)
- commit 93ba00b

-------------------------------------------------------------------
Wed Sep 12 16:23:21 CEST 2012 - mt@suse.de

- init.d/wicked: check for server.xml not config.xml
- commit 2154ff1

-------------------------------------------------------------------
Tue Sep 11 14:28:20 CEST 2012 - okir@suse.com

- Fix redhat config file parser to handle more recent Fedora features
- commit dbc3bf4

-------------------------------------------------------------------
Tue Sep 11 13:21:08 CEST 2012 - okir@suse.com

- make install no longer overwrites /etc/init.d scripts
  If you want to install the init scripts, you now have to invoke
  "make install-init" explicitly. This is so that people trying out
  wicked do not accidentally ruin their systems.
  Fix up the definition of devellibdir while we're at it.
- commit 2b02777

-------------------------------------------------------------------
Tue Sep 11 13:20:25 CEST 2012 - okir@suse.com

- gcc 4.7 warns about variables that are set but not used. Fix these warnings.
- commit b041327

-------------------------------------------------------------------
Tue Sep 11 13:18:13 CEST 2012 - okir@suse.com

- Work around an issue in more recent libdbus versions that check for boolean values
  Recent versions of libdbus validate whether a boolean we encode actually has a
  0/1 value, and bail out if not.
- commit 9c0b613

-------------------------------------------------------------------
Mon Sep 10 17:50:26 CEST 2012 - mt@suse.de

- dhcp4,autoip4: device created debug msg on success
- commit 3106a65

-------------------------------------------------------------------
Mon Sep 10 17:15:40 CEST 2012 - mt@suse.de

- autoip4: initialize link ifindex in device_new
- commit 8191fb3

-------------------------------------------------------------------
Mon Sep 10 17:13:41 CEST 2012 - mt@suse.de

- dhcp: Use same create messages for dhcp4 and dhcp6
- commit 4e55d0d

-------------------------------------------------------------------
Mon Sep 10 17:09:10 CEST 2012 - mt@suse.de

- autoip4: delete devices and deactivate events on exit
- dhcp4: delete devices and deactivate events on exit
- commit 0de0b4b

-------------------------------------------------------------------
Mon Sep 10 16:52:31 CEST 2012 - mt@suse.de

- autoip4: drop device ref after register in create
  register autoip4 device obtains an own reference on success
- dhcp4: drop device ref after register in create
  register dhcp4 device obtains an own reference on success
- commit b49cc45

-------------------------------------------------------------------
Mon Sep 10 16:51:54 CEST 2012 - mt@suse.de

- dhcp4: write debug message about device deletion
- commit b5e20cc

-------------------------------------------------------------------
Mon Sep 10 16:39:16 CEST 2012 - mt@suse.de

- dhcp4: initialize link ifindex in device_new
- commit 0150ba9

-------------------------------------------------------------------
Mon Sep 10 16:26:46 CEST 2012 - mt@suse.de

- dhcp6: fixed return value of dhcp6 device create
- commit 210a749

-------------------------------------------------------------------
Mon Sep 10 14:34:51 CEST 2012 - mt@suse.de

- dhcp6: delete devices and deactivate events on exit
- commit 2d19f35

-------------------------------------------------------------------
Mon Sep 10 14:30:44 CEST 2012 - mt@suse.de

- dhcp6: drop device ref after register in create
  register dhcp6 device obtains an own reference on success
- commit 55fdbb1

-------------------------------------------------------------------
Mon Sep 10 14:25:29 CEST 2012 - mt@suse.de

- autoip4: write debug message about device deletion
- dhcp6: write debug message about device deletion
- commit 1664d55

-------------------------------------------------------------------
Mon Sep 10 14:24:43 CEST 2012 - mt@suse.de

- dhcp6: close socket on send failures
- commit f51f507

-------------------------------------------------------------------
Mon Sep 10 14:17:45 CEST 2012 - mt@suse.de

- dhcp6: don't hold device reference in the socket
  The device owns the socket, not another way around.
  Don't reuse inactive or broken sockets while open.
- commit fe4383f

-------------------------------------------------------------------
Mon Sep 10 09:40:08 CEST 2012 - okir@suse.de

- Experimental support for WPA-EAP
- .gitignore: ignore html/ subdir
- Make sure "make htmldoc" sets LD_LIBRARY_PATH=.
- Update wicked(8) manpage to describe getnames
- Client-side support for getNames()
- Add support for InterfaceList.deviceByName()
- dbus config: lacked permission to send to org.opensuse.Network.InterfaceList
- Add server side support for Interface.getNames()
- schema: document linkMonitor method
- Support device identification by PCI attributes
- No longer include ipv4.h and ipv6.h in netinfo.h
- Move some debug related functions from netinfo.h to logging.h
- schema2html: fix up a warning
- commit 1621dab

-------------------------------------------------------------------
Thu Sep  6 17:17:17 CEST 2012 - mt@suse.de

- dhcp6: first enabled renew,rebind work in progress
- commit 50e4c30

-------------------------------------------------------------------
Thu Sep  6 17:07:19 CEST 2012 - mt@suse.de

- Added ni_address_list_count utility
- commit 493de78

-------------------------------------------------------------------
Tue Sep  4 13:58:50 CEST 2012 - okir@suse.de

- Add support for autogenerating DBus API doc from schema
- commit 2611827

-------------------------------------------------------------------
Tue Sep  4 09:22:21 CEST 2012 - mt@suse.de

- Added ni_sockaddr_is_loopback,linklocal variants
- commit b54d558

-------------------------------------------------------------------
Mon Sep  3 18:23:07 CEST 2012 - mt@suse.de

- ifevent: do not clear ifindex before device delete
  This caused, that all the dhcp,autoip,... device objects
  and their dbus services were not found and cleaned up.
- commit 438830a

-------------------------------------------------------------------
Mon Sep  3 10:55:17 CEST 2012 - okir@suse.de

- xml-schema: rearrange code in ni_xs_process_define() a bit
- commit b15e10c

-------------------------------------------------------------------
Fri Aug 31 18:05:17 CEST 2012 - mt@suse.de

- dhcp6: use fixed ia-pd comment from errata 248
- commit fb51f88

-------------------------------------------------------------------
Fri Aug 31 17:57:52 CEST 2012 - mt@suse.de

- dhcp6: removed obsolete / unused variable in main
- commit b62d7d9

-------------------------------------------------------------------
Fri Aug 31 17:55:35 CEST 2012 - mt@suse.de

- dhcp6: fixed to use correct ia-pd paragraph check
- commit e793f3e

-------------------------------------------------------------------
Fri Aug 31 17:09:52 CEST 2012 - mt@suse.de

- dhcp6: renew,rebind,decline work in progress (disabled)
- commit 0067b26

-------------------------------------------------------------------
Fri Aug 31 15:27:29 CEST 2012 - mt@suse.de

- dhcp6: use options directly of the ia type enum
- commit f03c97d

-------------------------------------------------------------------
Fri Aug 31 15:16:15 CEST 2012 - okir@suse.de

- Improving FSM debug output a little
- commit 63c0437

-------------------------------------------------------------------
Fri Aug 31 15:09:50 CEST 2012 - mt@suse.de

- dhcp6: Use one ia list in ni_addrconf_lease.dhcp6
- commit 38d5fd3

-------------------------------------------------------------------
Fri Aug 31 14:56:49 CEST 2012 - okir@suse.de

- When receiving a linkDown event, schedule the interface for take-down, not for recheck
- commit 57bad4a

-------------------------------------------------------------------
Fri Aug 31 14:56:44 CEST 2012 - mt@suse.de

- Makefile: Changed depend to use -MM instead -M
  This is to avoid adding compiler version and system header
  dependencies to Makefile.
  Also added all sources as prerequisites to depend target.
- commit 6482f30

-------------------------------------------------------------------
Fri Aug 31 14:55:15 CEST 2012 - okir@suse.de

- Fix a minor nuisance in fsm
  When refreshing the interface state, we would always change a
  device that didn't have link to a maximum state of DEVICE_EXISTS.
  This caused useless state transitions (which were fixed automatically,
  but nevertheless). Instead, cap the state at (LINK_UP - 1)
- commit f72e5a2

-------------------------------------------------------------------
Fri Aug 31 14:54:32 CEST 2012 - okir@suse.de

- schema: always call linkUp, not just when there's a <link> element
- commit 5ddbe0a

-------------------------------------------------------------------
Fri Aug 31 14:54:08 CEST 2012 - okir@suse.de

- network-nanny: allow users to call ManagedInterface.disable()
- commit 6bba67f

-------------------------------------------------------------------
Fri Aug 31 14:15:00 CEST 2012 - mt@suse.de

- Added ni_address_is_temporary,permanent,deprecated
- commit 54c38d4

-------------------------------------------------------------------
Fri Aug 31 13:35:50 CEST 2012 - okir@suse.de

- Added DBus service files for the addrconf supplicants
- commit 7043dcf

-------------------------------------------------------------------
Fri Aug 31 13:35:32 CEST 2012 - okir@suse.de

- Add missing config file for network-nanny
- commit 48a844f

-------------------------------------------------------------------
Fri Aug 31 11:21:24 CEST 2012 - okir@suse.de

- ni_address_t: change label member from a char[IFNAMSZ] to a char *
- commit 012d044

-------------------------------------------------------------------
Fri Aug 31 11:15:37 CEST 2012 - okir@suse.de

- Rename a number of ni_address* functions to ni_sockaddr_*
  All of these functions were operating on a ni_sockaddr_t, so use
  the ni_sockaddr_ prefix to distinguish them from ni_address_t
  related functions.
- commit 88fcbec

-------------------------------------------------------------------
Fri Aug 31 11:01:05 CEST 2012 - okir@suse.de

- wicked convert: support redhat ifcfg files
- commit 3291661

-------------------------------------------------------------------
Fri Aug 31 11:00:24 CEST 2012 - okir@suse.de

- Changed ni_var_array_set*() functions from an int return to void
- commit e7b94ef

-------------------------------------------------------------------
Fri Aug 31 10:56:20 CEST 2012 - okir@suse.de

- Add missing includes to <wicked/client.h>
- commit 134e864

-------------------------------------------------------------------
Fri Aug 31 10:53:17 CEST 2012 - okir@suse.de

- objectmodel: in the on-the-wire representation of addresses, we forgot broadcast and label
- commit b726827

-------------------------------------------------------------------
Fri Aug 31 10:52:54 CEST 2012 - okir@suse.de

- When processing RTM_NEWADDR messages, we missed IFA_LABEL
- commit b829ef9

-------------------------------------------------------------------
Thu Aug 30 14:32:51 CEST 2012 - okir@suse.de

- Redesign of the compat code to make redhat support easier
- commit e423a5d

-------------------------------------------------------------------
Thu Aug 30 14:25:06 CEST 2012 - okir@suse.de

- ifworker->control struct now has a type of its own (ni_ifworker_control_t)
- commit 59c3ac0

-------------------------------------------------------------------
Thu Aug 30 14:24:20 CEST 2012 - okir@suse.de

- __ni_parse_ipv4shorthand: fix a bug when parsing addresses like 127/8
  We forgot to convert the result to network byte order :-/
- commit 034c6a7

-------------------------------------------------------------------
Thu Aug 30 14:23:50 CEST 2012 - okir@suse.de

- Re-introduce ni_route_clone, make ni_route_list_append public
- commit e5cd639

-------------------------------------------------------------------
Thu Aug 30 12:02:11 CEST 2012 - okir@suse.de

- Change ni_netdev_new() call signature - remove netconfig_t argument
- commit 9515b31

-------------------------------------------------------------------
Thu Aug 30 11:56:49 CEST 2012 - okir@suse.de

- Clean up some dead code in compat-suse.c
- commit 3b120a2

-------------------------------------------------------------------
Thu Aug 30 11:54:35 CEST 2012 - okir@suse.de

- Update wicked manpage to descrive legacy conversion
- commit 9a55e4c

-------------------------------------------------------------------
Thu Aug 30 11:53:36 CEST 2012 - okir@suse.de

- wicked convert: support splitting up the document into per-interface files
- commit b1df46b

-------------------------------------------------------------------
Thu Aug 30 11:42:32 CEST 2012 - okir@suse.de

- Remove obsolete convert/suse.c
- commit ca3134b

-------------------------------------------------------------------
Thu Aug 30 11:41:29 CEST 2012 - okir@suse.de

- Add support for "wicked convert" to convert ifcfg files to XML
  So far, this supports SUSE ifcfg files only.
- commit bac30f5

-------------------------------------------------------------------
Thu Aug 30 11:41:10 CEST 2012 - okir@suse.de

- Make __ni_address_list_find public and rename to ni_address_list_find
- commit 0d6f8ad

-------------------------------------------------------------------
Thu Aug 30 11:40:37 CEST 2012 - okir@suse.de

- ni_sysconfig_get_value: if the string is empty, return NULL
- commit 967b72f

-------------------------------------------------------------------
Thu Aug 30 11:39:54 CEST 2012 - okir@suse.de

- Bugfix: when calling xml_document_print with a NULL file, we actually mean "stdout"
- commit 629222e

-------------------------------------------------------------------
Thu Aug 30 11:39:43 CEST 2012 - okir@suse.de

- Add ni_isdir, ni_isreg
- commit ef1f120

-------------------------------------------------------------------
Thu Aug 30 10:05:26 CEST 2012 - okir@suse.de

- Fix a bug in ni_address_new; publish ni_address_list_dedup
- commit bf403cc

-------------------------------------------------------------------
Thu Aug 30 09:38:18 CEST 2012 - okir@suse.de

- Introduce ni_sockaddr_prefix_parse
- commit 46b4a0a

-------------------------------------------------------------------
Thu Aug 30 09:21:05 CEST 2012 - okir@suse.de

- Rename ni_address_new()
  ni_address_new becomes ni_netdev_add_address,
  and __ni_address_new becomes ni_address_new
- commit e387aba

-------------------------------------------------------------------
Thu Aug 30 09:06:38 CEST 2012 - okir@suse.de

- Change the calling convention of all ni_sysconfig_get/set functions
   * do not strdup strings
   * use ni_bool_t when dealing with booleans
   * return a bool rather than an error (ie zero/negative)
- commit f159193

-------------------------------------------------------------------
Thu Aug 30 09:06:07 CEST 2012 - okir@suse.de

- Introduce several path related functions
  extern const char *    ni_dirname(const char *path);
  extern const char *    ni_sibling_path(const char *path, const char *file);
  extern const char *    ni_sibling_path_printf(const char *path, const char *fmt, ...);
- commit 222beb2

-------------------------------------------------------------------
Wed Aug 29 18:08:00 CEST 2012 - okir@suse.de

- Got rid of __ni_route_new()
- commit e1ab3ba

-------------------------------------------------------------------
Wed Aug 29 18:00:25 CEST 2012 - okir@suse.de

- Remove some ni_route_* functions from netinfo.h that duplicated from route.h
- commit 0dbb39f

-------------------------------------------------------------------
Wed Aug 29 17:55:23 CEST 2012 - okir@suse.de

- Moved several ni_sockaddr related functions from netinfo.h to address.h
  Also added ni_sockaddr_prefix_print()
- commit 0d9fa9d

-------------------------------------------------------------------
Wed Aug 29 17:20:38 CEST 2012 - okir@suse.de

- Moved src/sysconfig.h to include/wicked
- commit b676331

-------------------------------------------------------------------
Wed Aug 29 14:18:31 CEST 2012 - okir@suse.de

- move sysconfig.h from src/ to include/wicked
- commit fde278e

-------------------------------------------------------------------
Wed Aug 29 14:18:01 CEST 2012 - okir@suse.de

- Another documentation update
- commit 09ca96b

-------------------------------------------------------------------
Wed Aug 29 11:55:43 CEST 2012 - okir@suse.de

- Update WhyWicked.txt
- commit aefdee0

-------------------------------------------------------------------
Tue Aug 28 17:57:24 CEST 2012 - okir@suse.de

- Remove a lot of obsolete server-side ibft stuff
- commit c4efc81

-------------------------------------------------------------------
Tue Aug 28 17:45:24 CEST 2012 - okir@suse.de

- Replace much of the advanced device-identification code
  Previously, we used somewhat klunky constructs like
  	<identify>
  	  <ethernet>
  	    <permanent-address>00:11:22:33:44:55</permanent-address>
  	  </ethernet>
  	</identify>
  with some equally klunky rules for abbreviating the XML somewhat.
  Change this to use a namespace="..." attribute in order to specify
  the naming service that should be consulted to resolve the name.
   *	Config files that want to refer to a device using a specific naming
  	service should now use <name namespace="...">:
  	<name namespace="ethernet">
  	  <permanent-address>00:11:22:33:44:55</permanent-address>
  	</name>
  	This is just a more generic form of the <name> element.
  	IOW, the default namespace is to look up by kernel device name.
   *	When referencing a subordinate device, you should use <device>
   	like this:
  	   <device namespace="alias">
  	     myfunkydevice0
  	   </device>
  	or
  	   <device namespace="modem">
  	     <equipment-id>123456789</equipment-id>
  	   </device>
  Currently, the old naming convention is still supported, but slated to
  disappear any time now.
- commit a7c0e41

-------------------------------------------------------------------
Tue Aug 28 17:42:21 CEST 2012 - okir@suse.de

- Remove obsolete sample file
- commit 70bef28

-------------------------------------------------------------------
Tue Aug 28 15:27:08 CEST 2012 - okir@suse.de

- Added a document describing the design and its rationale
- commit 61ca7af

-------------------------------------------------------------------
Tue Aug 28 15:26:27 CEST 2012 - okir@suse.de

- schema: linkUp() didn't specify which XML element to use as input
- commit 76dc67b

-------------------------------------------------------------------
Tue Aug 28 14:08:38 CEST 2012 - okir@suse.de

- Add several new marshal/unmarshal functions to dbus-objects/misc.c
  In addition to passing address and address/prefix data along in a dict,
  also support functions for extracting these from a variant object directly.
- commit fdb4c49

-------------------------------------------------------------------
Tue Aug 28 14:01:47 CEST 2012 - okir@suse.de

- Renamed a few local functions in dbus-objects/misc.c
- commit a3cf0e2

-------------------------------------------------------------------
Mon Aug 27 17:35:06 CEST 2012 - okir@suse.de

- network-nanny: Do a better job handling device enablement
  Up to now, you had to be root in order to enable a device, and
  it always had to happen at runtime.
  Now you can control behavior through the config file:
   <nanny>
     <allow link-type="wireless" />
     <allow class="modem" auto="true" />
     <allow device="eth0" />
   </nanny>
  A device matched by such an "allow" element is now subject to
  user management through network-nanny. By default, the device
  still has to be enabled explicitly for nanny to monitor it;
  but now any user can call "wicked nanny enable $DEVICE" on such
  a device.
  If the auto="true" attribute is given, then a matching device is
  set up for monitoring automatically when detected.
  The root user is allowed to enable or disable any device.
- commit 5158ea4

-------------------------------------------------------------------
Mon Aug 27 17:28:50 CEST 2012 - okir@suse.de

- Introduce ni_objectmodel_link_class() helper
- commit 39f0d37

-------------------------------------------------------------------
Mon Aug 27 17:09:11 CEST 2012 - okir@suse.de

- network-nanny: rename managed_device.user_controlled -> monitor
- commit 1511c80

-------------------------------------------------------------------
Mon Aug 27 16:42:47 CEST 2012 - okir@suse.de

- ni_config: Change the return type of various ni_config_parse_* helpers from int to bool
  We always returned -1, and never inspected the return code anyway.
- commit c8c265f

-------------------------------------------------------------------
Mon Aug 27 16:35:43 CEST 2012 - okir@suse.de

- Introduce a variant of ni_init that takes a callback for processing application-specific config options
- commit 4c2abe7

-------------------------------------------------------------------
Mon Aug 27 14:45:09 CEST 2012 - okir@suse.de

- Update wicked-config(5)
- commit 290a9fa

-------------------------------------------------------------------
Mon Aug 27 14:39:14 CEST 2012 - okir@suse.de

- ni_dbus_xml_register_methods: validate signature of built-in methods
  When associating the method information from the schema with a built-in
  method, verify the signature to make sure that schema and implementation
  actually agree.
- commit 6225d5e

-------------------------------------------------------------------
Mon Aug 27 14:39:04 CEST 2012 - okir@suse.de

- dhcp4: get rid of a silly debug message
- commit 222e63a

-------------------------------------------------------------------
Mon Aug 27 14:32:02 CEST 2012 - okir@suse.de

- Split up configuration file
- commit 5829c55

-------------------------------------------------------------------
Mon Aug 27 14:30:01 CEST 2012 - okir@suse.de

- ni_config_parse: change the syntax of <include>
- commit a624f78

-------------------------------------------------------------------
Mon Aug 27 14:29:21 CEST 2012 - okir@suse.de

- ni_init: when there's no $appname.xml config file, fall back to config.xml
- commit 15b90ea

-------------------------------------------------------------------
Mon Aug 27 14:19:28 CEST 2012 - okir@suse.de

- ni_config_parse: move handling of <statedir> into the element loop
- commit a6e5f77

-------------------------------------------------------------------
Mon Aug 27 14:17:13 CEST 2012 - okir@suse.de

- Got rid of obsolete ni_config->{ipv4,ipv6} members
- commit 960fe5f

-------------------------------------------------------------------
Mon Aug 27 14:13:58 CEST 2012 - okir@suse.de

- ni_config_parse: move handling of several elements into the element loop
- commit 7544d62

-------------------------------------------------------------------
Mon Aug 27 14:10:21 CEST 2012 - okir@suse.de

- Support <include> statement in config file
- commit d332b2c

-------------------------------------------------------------------
Mon Aug 27 13:54:22 CEST 2012 - okir@suse.de

- ni_init() takes an additional application name to select config file
- commit ff741eb

-------------------------------------------------------------------
Mon Aug 27 12:48:18 CEST 2012 - mt@suse.de

- /etc/init.d/network: fixed to use ifup --mode boot
- commit 90feed9

-------------------------------------------------------------------
Mon Aug 27 12:46:36 CEST 2012 - okir@suse.de

- Introduce <route-priority> field in <ipv4:dhcp> request
- commit f90f151

-------------------------------------------------------------------
Mon Aug 27 12:45:26 CEST 2012 - okir@suse.de

- When comparing two routes, also verify their priority
  Routes with different priorities are distinct, from the kernel's
  point of view.
- commit 3c95ab7

-------------------------------------------------------------------
Mon Aug 27 12:43:55 CEST 2012 - okir@suse.de

- ifconfig: we did not write all route attributes correctly when installing a route
- commit b062d81

-------------------------------------------------------------------
Mon Aug 27 12:42:11 CEST 2012 - okir@suse.de

- addrconf: fix a small bug when handling dropLease requests
  When a lease does not exist (anymore), we want to catch the
  AddrconfNoLease error, and just pretend it succeeded. We did
  not do this correctly (missing minus sign in front of the error
  code).
- commit cdbb68c

-------------------------------------------------------------------
Mon Aug 27 12:37:38 CEST 2012 - mt@suse.de

- etc/init.d/network: defined WICKED_CLIENT_BIN variable
- commit c830a2d

-------------------------------------------------------------------
Mon Aug 27 12:16:58 CEST 2012 - mt@suse.de

- dhcp6: change how we register device objects.
  Each object representing a network device is now bound to its interfaces
  using ni_objectmodel_bind_compatible_interfaces. This way, we also enable
  the generic org.opensuse.Network.Addrconf service on each device; so that
  the signal sending code stops complaining about missing signal definitions.
- commit dfb05cb

-------------------------------------------------------------------
Mon Aug 27 12:09:03 CEST 2012 - okir@suse.de

- network-nanny: isolate secrets set by different users
  Introduce a "nanny-user", and associate a separate secret-db with
  each user. In Nanny.addSecret, new secrets are added to the DB
  for the caller of the method. Similarly, when updating a profile, store
  the caller's uid in the profile.
  Then, when we bring up an interface based on a given profile, look
  up any secrets in the DB of the user who last updated the profile.
- commit cd7e671

-------------------------------------------------------------------
Mon Aug 27 12:07:48 CEST 2012 - okir@suse.de

- dbus-server: optionally, pass caller's uid to method handler
  Introduce an additional type of method handler (method->handler_ex)
  which is given the caller's uid in addition to the arguments
- commit 8326743

-------------------------------------------------------------------
Mon Aug 27 11:53:08 CEST 2012 - mt@suse.de

- dhcp6: get rid of wicked prefixes, dhcp6 not 4 fix
- commit 722ff76

-------------------------------------------------------------------
Mon Aug 27 11:30:02 CEST 2012 - okir@suse.de

- network-nanny: allow normal users to set policies and secrets
- commit 40699a0

-------------------------------------------------------------------
Mon Aug 27 11:19:51 CEST 2012 - okir@suse.de

- Implement ni_dbus_object_get_caller_uid
- commit e7daa56

-------------------------------------------------------------------
Mon Aug 27 10:41:42 CEST 2012 - okir@suse.de

- renamed etc/network-mgr.conf -> etc/network-nanny.conf
- commit f9c2f5e

-------------------------------------------------------------------
Mon Aug 27 10:34:57 CEST 2012 - okir@suse.de

- Add some comments to sample nanny policy files
- commit b283fea

-------------------------------------------------------------------
Mon Aug 27 10:28:12 CEST 2012 - okir@suse.de

- Renamed samples/manager -> samples/nanny
- commit 8554765

-------------------------------------------------------------------
Mon Aug 27 10:27:28 CEST 2012 - okir@suse.de

- network-nanny: rename {manager,nanny}.[hc]
- commit 980bb5b

-------------------------------------------------------------------
Mon Aug 27 10:19:39 CEST 2012 - okir@suse.de

- network-nanny: more manager->nanny rename
- commit 4e658a3

-------------------------------------------------------------------
Mon Aug 27 10:06:42 CEST 2012 - okir@suse.de

- Rename org.opensuse.Network.Manager -> Nanny
- commit 204f788

-------------------------------------------------------------------
Mon Aug 27 10:03:53 CEST 2012 - okir@suse.de

- client: rename manager.c to nanny.c
- commit 7fe8c8b

-------------------------------------------------------------------
Mon Aug 27 10:03:27 CEST 2012 - okir@suse.de

- wicked client: rename manager subcommand to nanny
- commit 3721ad7

-------------------------------------------------------------------
Mon Aug 27 09:47:11 CEST 2012 - okir@suse.de

- Work around a segfault in dhcp4-supplicant shutdown
- commit 7c45e20

-------------------------------------------------------------------
Mon Aug 27 09:44:14 CEST 2012 - okir@suse.de

- autoip4: change how we register device objects.
  Each object representing a network device is now bound to its interfaces
  using ni_objectmodel_bind_compatible_interfaces. This way, we also enable
  the generic org.opensuse.Network.Addrconf service on each device; so that
  the signal sending code stops complaining about missing signal definitions.
- commit 65ba22e

-------------------------------------------------------------------
Mon Aug 27 09:34:00 CEST 2012 - okir@suse.de

- dhcp4: change how we register device objects.
  Each object representing a network device is now bound to its interfaces
  using ni_objectmodel_bind_compatible_interfaces. This way, we also enable
  the generic org.opensuse.Network.Addrconf service on each device; so that
  the signal sending code stops complaining about missing signal definitions.
- commit 1829741

-------------------------------------------------------------------
Mon Aug 27 09:33:21 CEST 2012 - okir@suse.de

- Introduce an addrconf-device class
- commit c1b7fde

-------------------------------------------------------------------
Mon Aug 27 09:17:00 CEST 2012 - okir@suse.de

- ni_objectmodel_register_service() should take a const service pointer as argument
- commit 35db8c4

-------------------------------------------------------------------
Fri Aug 24 18:20:22 CEST 2012 - mt@suse.de

- ifevent: report error on add membership failure
- commit 8071f88

-------------------------------------------------------------------
Fri Aug 24 16:47:16 CEST 2012 - mt@suse.de

- Makefile: Fixed config.h dependencies
- commit 4153aac

-------------------------------------------------------------------
Fri Aug 24 16:19:54 CEST 2012 - okir@suse.de

- addrconf supplicants use generic org.opensuse.Network.Addrconf interface to send signals
  Currently, addrconf supplicants for dhcp4, dhcp6 and autoip4 use their flavor-specific
  interfaces to send Lease{Acquired,Released,Lost} events. Make them use the generic
  org.opensuse.Network.Addrconf interface, and document that interface in the schema.
- commit 53cf275

-------------------------------------------------------------------
Fri Aug 24 12:58:44 CEST 2012 - mt@suse.de

- Makefile: fixed some build / cleanup dependencies
- commit 64582aa

-------------------------------------------------------------------
Fri Aug 24 12:11:07 CEST 2012 - mt@suse.de

- Makefile: moved init script install to install-init
- commit 00d2976

-------------------------------------------------------------------
Fri Aug 24 11:53:46 CEST 2012 - okir@suse.de

- Make use of schema information explicit in ni_dbus_{service,method}_t
  Up to now, we stored the pointer to the schema information in a
  "void *user_data" pointer inside method and service objects. Which
  is a bit pointless, given that there's only one user of that pointer.
  Make the purpose of this field explicit by using the proper
  ni_xs_{method,service}_t types, and calling them "schema" rather
  than "user_data".
- commit 44c91b2

-------------------------------------------------------------------
Fri Aug 24 11:51:52 CEST 2012 - mt@suse.de

- Don't complain for devices not supporting ethtool
- commit a6b63f8

-------------------------------------------------------------------
Fri Aug 24 11:38:51 CEST 2012 - okir@suse.de

- First stab at introspection
- commit 71617a7

-------------------------------------------------------------------
Thu Aug 23 16:54:59 CEST 2012 - okir@suse.de

- network-nanny: change managed device to track "missing secrets" through a flag, not a state
- commit dba757a

-------------------------------------------------------------------
Thu Aug 23 16:51:24 CEST 2012 - okir@suse.de

- network-nanny: change event handlers to map signal names to ni_event_t
- commit f988de8

-------------------------------------------------------------------
Thu Aug 23 16:43:36 CEST 2012 - okir@suse.de

- Make ni_objectmodel_event_to_signal public, and add ni_objectmodel_signal_to_event
- commit 76f513d

-------------------------------------------------------------------
Thu Aug 23 11:46:30 CEST 2012 - mt@suse.de

- config.xml: removed -- from comment, confuses vim
- commit db3ffeb

-------------------------------------------------------------------
Thu Aug 23 11:09:50 CEST 2012 - mt@suse.de

- client: ininitialized rv in do_manager_addsecret
- commit 73a1eb4

-------------------------------------------------------------------
Thu Aug 23 10:46:05 CEST 2012 - mt@suse.de

- ifconfig: Fixed return value of __ni_rtnl_link_up
- commit 0ea8794

-------------------------------------------------------------------
Thu Aug 23 10:44:26 CEST 2012 - mt@suse.de

- dhcp6: use new ni_address_is util functions
- commit b9cd826

-------------------------------------------------------------------
Thu Aug 23 09:57:37 CEST 2012 - mt@suse.de

- Check also ipv6 in ni_address_is_loopback
- commit 7fcc2da

-------------------------------------------------------------------
Thu Aug 23 09:55:47 CEST 2012 - mt@suse.de

- Added ni_address_is_linklocal,duplicate,tentative
- commit 4f83932

-------------------------------------------------------------------
Thu Aug 23 09:40:24 CEST 2012 - okir@suse.de

- network-nanny: hande deviceCreate events for netdevs correctly
- commit d55ab01

-------------------------------------------------------------------
Thu Aug 23 09:19:20 CEST 2012 - okir@suse.de

- network-nanny: in ni_managed_device_apply_policy, change how we act in various states
- commit d0dc100

-------------------------------------------------------------------
Thu Aug 23 08:57:13 CEST 2012 - mt@suse.de

- Fixed ni_address_list_remove utility function
- commit 40665ec

-------------------------------------------------------------------
Thu Aug 23 08:42:40 CEST 2012 - okir@suse.de

- dbus-connection: do not re-enter dbus_connection_dispatch()
  In some cases, we could deadlock:
  	incoming signal on dbus socket
  	-> dbus_connection_dispatch()
  	-> invoke signal handler
  	-> signal handler places a dbus call
  	-> ni_dbus_connection_call call dbus_connection_dispatch()
  		in order to flush out any signals that may
  		arrive while we're waiting for the response.
  	-> this deadlocks forever
  This patch prevents re-entering dbus_connection_dispatch().
  Let's hope that this is good enough.
- commit 4f2e64b

-------------------------------------------------------------------
Wed Aug 22 17:54:16 CEST 2012 - mt@suse.de

- dhcp6: check address family in addr update events
- commit 343af99

-------------------------------------------------------------------
Wed Aug 22 15:13:05 CEST 2012 - mt@suse.de

- Use int format precision to not fail with -Werror
- commit 0c682bf

-------------------------------------------------------------------
Wed Aug 22 15:05:13 CEST 2012 - mt@suse.de

- dhcp: wait until ipv6 dad finished before starting
- commit b873821

-------------------------------------------------------------------
Wed Aug 22 15:04:28 CEST 2012 - mt@suse.de

- dhcp6: try to load duid using same filename argument
- commit 25038da

-------------------------------------------------------------------
Wed Aug 22 14:18:35 CEST 2012 - okir@suse.de

- dhcp4: get rid of more wicked_ function prefixes
- commit 71a8d75

-------------------------------------------------------------------
Wed Aug 22 14:05:58 CEST 2012 - okir@suse.de

- client: get rid of a few wicked_ function prefixes
- commit b4daa3a

-------------------------------------------------------------------
Wed Aug 22 13:59:24 CEST 2012 - okir@suse.de

- server: get rid of a few wicked_ function prefixes
- commit 8e539d5

-------------------------------------------------------------------
Wed Aug 22 13:54:03 CEST 2012 - okir@suse.de

- The big namespace change: replace com.suse.Wicked with org.opensuse.Network
- commit c311c38

-------------------------------------------------------------------
Wed Aug 22 13:53:37 CEST 2012 - okir@suse.de

- Fix dbus method signature of Manager.addSecret()
- commit ae61e29

-------------------------------------------------------------------
Wed Aug 22 13:37:26 CEST 2012 - okir@suse.de

- Clean up some places where we use com.suse.Wicked verbatim
- commit 6412753

-------------------------------------------------------------------
Wed Aug 22 13:22:41 CEST 2012 - okir@suse.de

- Recognize and map org.freedesktop.DBus.Error.ServiceUnknown errors
- commit 62e81ab

-------------------------------------------------------------------
Wed Aug 22 13:10:54 CEST 2012 - okir@suse.de

- network-nanny: change security_id type
  Previously, this was a string like "essid:okirnet". We need something
  more flexible, so change it to a class + list-of-attributes thing.
  The string syntax looks like this:
  	wireless:essid=okirnet,ap=00:11:22:33:44:55
  	modem:equipment-id=123456
  When setting a secret in network-nanny, you do not need to send all
  attributes; leaving an attribute unset means "I don't care".
- commit 35fe62e

-------------------------------------------------------------------
Wed Aug 22 13:00:17 CEST 2012 - okir@suse.de

- Introduce ni_quote, ni_unquote
- commit 3d77b4b

-------------------------------------------------------------------
Wed Aug 22 12:59:05 CEST 2012 - okir@suse.de

- ni_stringbuf_vprintf: make it configurable whether to reset the buffer or not
  Also, use vasprintf rather than doing our own vsnprintf() loop.
- commit f77d6f0

-------------------------------------------------------------------
Tue Aug 21 15:12:32 CEST 2012 - mt@suse.de

- dhcp6: Removed obsolete wireless scanning disabling
- commit 9ef451c

-------------------------------------------------------------------
Tue Aug 21 09:43:50 CEST 2012 - okir@suse.de

- network-nanny: support device disable
- commit 69011b6

-------------------------------------------------------------------
Tue Aug 21 09:43:28 CEST 2012 - okir@suse.de

- In ni_dbus_object_get_managed_objects, automatically purge stale client objects
- commit 4deefec

-------------------------------------------------------------------
Tue Aug 21 09:25:44 CEST 2012 - okir@suse.de

- Fix up a bug introduced by one of the previous commits
  When bringing up a virtual device, we would bail out prematurely
  because the worker didn't have a dbus object associated with it.
  Change this: only error out if the worker doesn't have a device
  factory method bound to it.
  Also, we weren't always signalling success correctly, due to some
  half-baked code duplication between ni_ifworker_update_state
  and ni_ifworker_set_state.
- commit 65f622e

-------------------------------------------------------------------
Mon Aug 20 20:11:29 CEST 2012 - okir@suse.de

- Add sample wlan policy for network-nanny
- commit 0073544

-------------------------------------------------------------------
Mon Aug 20 18:32:04 CEST 2012 - mt@suse.de

- Pass ifnode as config to ni_ifworker_set_config
- commit 63aa036

-------------------------------------------------------------------
Mon Aug 20 12:53:06 CEST 2012 - mt@suse.de

- rcnetwork: quoted mode variable (unary operator)
- commit 94e5d8c

-------------------------------------------------------------------
Fri Aug 17 17:09:41 CEST 2012 - okir@suse.de

- network-nanny: improve the way we treat missing secrets
- commit 761526a

-------------------------------------------------------------------
Fri Aug 17 16:35:39 CEST 2012 - okir@suse.de

- Add general ni_secret_db functionality and change network-nanny to use it
- commit 974db40

-------------------------------------------------------------------
Fri Aug 17 14:09:17 CEST 2012 - mt@suse.de

- Test utility showing rtnetlink netdev events
- commit 63d57b3

-------------------------------------------------------------------
Fri Aug 17 14:08:35 CEST 2012 - mt@suse.de

- Use wildcard to ignore testing binaries
- commit dccadb9

-------------------------------------------------------------------
Fri Aug 17 14:07:14 CEST 2012 - mt@suse.de

- Added interface address change events and cleanup
- commit 3fb31d3

-------------------------------------------------------------------
Fri Aug 17 14:04:23 CEST 2012 - mt@suse.de

- Expose rtnl newaddr parsing into an address
- commit 453cd1a

-------------------------------------------------------------------
Fri Aug 17 11:26:41 CEST 2012 - okir@suse.de

- Cleaned up lots unneeded #includes
- commit 603457f

-------------------------------------------------------------------
Fri Aug 17 10:42:23 CEST 2012 - mt@suse.de

- Added ni_address_list_remove utility function
- commit 8738559

-------------------------------------------------------------------
Thu Aug 16 22:40:24 CEST 2012 - martin@vidner.net

- Allow everyone to introspect us.
  OTOH ObjectManager access for everyone means no property is a secret.
  That's probably a problem, FIXME.
- commit 1d85652

-------------------------------------------------------------------
Thu Aug 16 22:31:32 CEST 2012 - martin@vidner.net

- Added a dummy handler for o.fd.DBus.Introspectable.Introspect
- commit 5f2fd4d

-------------------------------------------------------------------
Thu Aug 16 21:07:39 CEST 2012 - martin@vidner.net

- Makefile: don't let install-data overwrite /etc/init.d/network
  Moved installing the init scripts from install-data to install-bin.
- commit 5d914b7

-------------------------------------------------------------------
Thu Aug 16 17:51:43 CEST 2012 - okir@suse.de

- expire wireless scan info more aggressively
- commit 637c8c9

-------------------------------------------------------------------
Thu Aug 16 17:22:42 CEST 2012 - okir@suse.de

- network-nanny: trigger a recheck when interface sends a linkScanUpdated event
- commit 952f277

-------------------------------------------------------------------
Thu Aug 16 16:12:46 CEST 2012 - okir@suse.de

- Working on sending out notifications when our list of visible wlan cells changes
- commit deef8ec

-------------------------------------------------------------------
Thu Aug 16 16:12:26 CEST 2012 - okir@suse.de

- dhcp4-supplicant: fix a FIXME :-)
- commit 4c8e3fe

-------------------------------------------------------------------
Thu Aug 16 13:55:38 CEST 2012 - okir@suse.de

- Work on getting wireless supported by network-nanny
- commit fb78927

-------------------------------------------------------------------
Thu Aug 16 13:34:06 CEST 2012 - okir@suse.de

- fsm-policy: implement <wireless:essid> check
- commit 014b7b4

-------------------------------------------------------------------
Thu Aug 16 13:33:18 CEST 2012 - okir@suse.de

- wireless objectmodel: when receiving a wireless object, unmarshal the scan info (unimplemented until now)
- commit fe6f628

-------------------------------------------------------------------
Thu Aug 16 13:23:27 CEST 2012 - okir@suse.de

- Add new function ni_wireless_parse_ssid; make ni_wireless_scan_{new,free} public
- commit 8a13b2b

-------------------------------------------------------------------
Thu Aug 16 13:12:50 CEST 2012 - okir@suse.de

- Move __ni_objectmodel_{get,set}_hwaddr to misc.c and make them visible to other services
- commit 7148c81

-------------------------------------------------------------------
Thu Aug 16 12:22:36 CEST 2012 - okir@suse.de

- FSM: factor our API binding code from fsm_build_hierachy
  Move this code into a new function called ni_ifworker_bind_early;
  and call this function from network-nanny when preparing to bring
  up a device.
- commit 961a5f3

-------------------------------------------------------------------
Thu Aug 16 11:35:48 CEST 2012 - okir@suse.de

- network-nanny: fix up some dbus-class/service related variable names
- commit 3d05996

-------------------------------------------------------------------
Thu Aug 16 11:33:53 CEST 2012 - okir@suse.de

- network-nanny: combine mgr->{modem,netdev}_list into one device_list
- commit f889e16

-------------------------------------------------------------------
Thu Aug 16 11:30:06 CEST 2012 - okir@suse.de

- Get rid of obsolete ni_layer_t enum
- commit cfe1f36

-------------------------------------------------------------------
Thu Aug 16 11:28:46 CEST 2012 - okir@suse.de

- Renamed ni_objectmodel_modem_get_proxy_class* function to ni_objectmodel_modem_get_class
  Moved these functions from src/modem-manager.c to src/dbus-objects/modem.c
- commit b62adea

-------------------------------------------------------------------
Thu Aug 16 11:24:12 CEST 2012 - okir@suse.de

- Renamed dbus class modem-proxy to modem
- commit 9e71516

-------------------------------------------------------------------
Thu Aug 16 11:18:01 CEST 2012 - okir@suse.de

- Rename the dbus class for modem manager client objects from "modem" to "mm-modem"
  Same goes for all derived classes (modem-gsm, modem-cdma).
- commit 9f0bba1

-------------------------------------------------------------------
Thu Aug 16 10:01:04 CEST 2012 - okir@suse.de

- Sort out the mess with dbus modem class naming
  Client objects for objects owned by modem-manager should have a
  class name starting with "mm-". The first step is to untangle the
  names of the macro referring to these classes, and the functions
  returning the class names for specific modem sub-types (gsm, cdma).
- commit 25fc9af

-------------------------------------------------------------------
Thu Aug 16 09:43:37 CEST 2012 - okir@suse.de

- Only wickedd should request wireless scans from wpa-supplicant directly
  All other services should go through the daemon explicitly.
- commit 20cae1a

-------------------------------------------------------------------
Thu Aug 16 09:29:17 CEST 2012 - okir@suse.de

- In __ni_manager_rfkill_event, update a device's rfkill info even if it's not user-controlled
- commit 5b14f7d

-------------------------------------------------------------------
Thu Aug 16 09:28:58 CEST 2012 - okir@suse.de

- Remove an incorrect warning in ni_managed_device_new
- commit cdc38d6

-------------------------------------------------------------------
Thu Aug 16 09:26:20 CEST 2012 - okir@suse.de

- Remove obsolete enum ni_evaction
- commit 5a462cb

-------------------------------------------------------------------
Wed Aug 15 19:02:55 CEST 2012 - okir@suse.de

- Add rfkill support
- commit 9850dff

-------------------------------------------------------------------
Wed Aug 15 18:28:50 CEST 2012 - okir@suse.de

- Remove a lot of old cruft from the socket code
- commit fd5c44e

-------------------------------------------------------------------
Wed Aug 15 17:43:46 CEST 2012 - okir@suse.de

- In __ni_dbus_object_unregister, stop destroying after we've run the first destroy() function
- commit fa76e68

-------------------------------------------------------------------
Wed Aug 15 17:34:14 CEST 2012 - okir@suse.de

- ifconfig: propagate RFKILL error from __ni_rtnl_link_up
- commit 7522ba7

-------------------------------------------------------------------
Wed Aug 15 17:16:01 CEST 2012 - okir@suse.de

- Add new error RADIO_DISABLED
- commit 9fc9dcd

-------------------------------------------------------------------
Wed Aug 15 16:23:59 CEST 2012 - okir@suse.de

- Fix one debug message in fsm
- commit 86b7731

-------------------------------------------------------------------
Wed Aug 15 16:19:57 CEST 2012 - okir@suse.de

- Fix return types of a bunch of functions
  These functions were returning an integer, but were really providing
  a boolean. Change them to ni_bool_t return value
- commit f783836

-------------------------------------------------------------------
Wed Aug 15 16:03:51 CEST 2012 - okir@suse.de

- network-nanny: removed one useless function
- commit 35627a0

-------------------------------------------------------------------
Wed Aug 15 16:02:15 CEST 2012 - okir@suse.de

- network-nanny: handle ethernet link down correctly
- commit cb83086

-------------------------------------------------------------------
Wed Aug 15 15:35:40 CEST 2012 - okir@suse.de

- network-nanny: fix Ethernet link detection use case
- commit 74f2f1c

-------------------------------------------------------------------
Wed Aug 15 15:31:48 CEST 2012 - okir@suse.de

- FSM: fix two bugs with state updates
   a)	when the server signalled that a link went down, etc,
   	we failed to update the the worker's current state due to
  	a stupid bug
   b)	we only ever updated the state of devices we were actively
   	configuring.
- commit 02fff0d

-------------------------------------------------------------------
Wed Aug 15 15:11:12 CEST 2012 - okir@suse.de

- Fix a bug in __ni_lease_owns_address
  This routine would return TRUE if the address did *not* match :-(
- commit 7d8e273

-------------------------------------------------------------------
Wed Aug 15 15:05:21 CEST 2012 - mt@suse.de

- Changed ifevent to use nl_recvmsgs with callback
- commit 73cf348

-------------------------------------------------------------------
Wed Aug 15 14:55:34 CEST 2012 - mt@suse.de

- Fixed tags target in the makefile
- commit 97830e3

-------------------------------------------------------------------
Wed Aug 15 14:43:59 CEST 2012 - okir@suse.de

- Try to fix up object destruction in dbus-server
- commit 3f0992d

-------------------------------------------------------------------
Wed Aug 15 14:41:36 CEST 2012 - okir@suse.de

- Remove a superfluous ni_trace() call
- commit 2b9f891

-------------------------------------------------------------------
Wed Aug 15 13:01:03 CEST 2012 - okir@suse.de

- Fix a bug in ni_fsm_refresh_state
  When we had brought up a device, w->done would be true, and on the next
  refresh, we would clear w->object, and subsequently w->device.
- commit d9bb6d2

-------------------------------------------------------------------
Wed Aug 15 12:11:20 CEST 2012 - okir@suse.de

- Extend the sample umts policy to include a template for instantiating a PPP device
- commit 50c4f3e

-------------------------------------------------------------------
Wed Aug 15 12:10:38 CEST 2012 - okir@suse.de

- client: fix nanny addpolicy to process more than one policy/template per file
- commit ab2364a

-------------------------------------------------------------------
Wed Aug 15 12:00:39 CEST 2012 - okir@suse.de

- Introduce "template" policies
  A template can be used to create virtual devices in response to
  events (usually a new device becoming available). For instance, you
  can define a template to create a PPP device whenever a new modem
  becomes available. Or you can create a template that creates a PPPoE
  device whenever it finds a specific Ethernet device has link.
- commit dfa43c1

-------------------------------------------------------------------
Wed Aug 15 12:00:32 CEST 2012 - okir@suse.de

- Make ni_objectmodel_netif_class visible globally
- commit 9519f51

-------------------------------------------------------------------
Wed Aug 15 10:24:16 CEST 2012 - okir@suse.de

- fsm-policy: rename NI_IFPOLICY_TYPE_* to NI_IFPOLICY_ACTION_*
- commit 944b252

-------------------------------------------------------------------
Wed Aug 15 10:20:25 CEST 2012 - okir@suse.de

- fsm-policy: introduce <sharable> condition when checking devices
- commit 59b3463

-------------------------------------------------------------------
Wed Aug 15 09:59:42 CEST 2012 - okir@suse.de

- FSM: when resetting an ifworker, fix up the shared/exclusive ownership info of all children
- commit ee1ae8f

-------------------------------------------------------------------
Wed Aug 15 08:27:12 CEST 2012 - okir@suse.de

- network-nanny: replace mdev->running with mdev->state and support several distinct states
- commit 4a48e65

-------------------------------------------------------------------
Wed Aug 15 08:00:56 CEST 2012 - okir@suse.de

- fix ni_debug_xml to use new xml_node_print_debug() function
- commit cc7a313

-------------------------------------------------------------------
Wed Aug 15 08:00:14 CEST 2012 - okir@suse.de

- introduce ni_debug_nanny and use it
- commit 6643d83

-------------------------------------------------------------------
Wed Aug 15 07:52:54 CEST 2012 - okir@suse.de

- network-nanny: do not apply policies to devices that were removed
- commit f8b5a5a

-------------------------------------------------------------------
Wed Aug 15 07:51:11 CEST 2012 - okir@suse.de

- network-nanny: handle deviceDown for netdevs, too
- commit 2ac9899

-------------------------------------------------------------------
Tue Aug 14 23:49:26 CEST 2012 - okir@suse.de

- network-nanny: some code reshuffle
- commit 1affb3a

-------------------------------------------------------------------
Tue Aug 14 18:45:13 CEST 2012 - okir@suse.de

- network-nanny: inspect the return status of an ifworker and deal with errors
- commit 984b5d2

-------------------------------------------------------------------
Tue Aug 14 18:44:27 CEST 2012 - okir@suse.de

- Fix the netdev event sending code
  When sending a deleteDevice event for a netdev, there is a race condition
  because we delete the server-side dbus object after sending the event(s).
  It may happen that the client receives the deleteDevice event, deletes its
  local object handle, then calls GetManagedObject, which still finds the
  active server-side object and returns it - appearing as a *new* object
  on the client side.
  This patch changes this by
  	- unregistering the server-side object before sending events
  	- modifying the netdev event sending code to take a dbus object
  	  as argument, rather than the netdev device handle
- commit 4187bfd

-------------------------------------------------------------------
Tue Aug 14 18:32:41 CEST 2012 - okir@suse.de

- Fix the modem event sending code
  When sending a deleteDevice event for a modem, there is a race condition
  because we delete the server-side dbus object after sending the event(s).
  It may happen that the client receives the deleteDevice event, deletes its
  local object handle, then calls GetManagedObject, which still finds the
  active server-side object and returns it - appearing as a *new* object
  on the client side.
  This patch changes this by
  	- unregistering the server-side object before sending events
  	- modifying the modem event sending code to take a dbus object
  	  as argument, rather than the modem device handle
- commit 7cd0732

-------------------------------------------------------------------
Tue Aug 14 18:31:47 CEST 2012 - okir@suse.de

- FSM: propagate errors in schema validation to the caller, and don't start the FSM
- commit 39a1295

-------------------------------------------------------------------
Tue Aug 14 15:10:32 CEST 2012 - okir@suse.de

- FSM: do better tracking of active vs completed ifworkers
- commit 1f4e709

-------------------------------------------------------------------
Tue Aug 14 15:07:08 CEST 2012 - okir@suse.de

- network-nanny: update sequence number when updating a policy
- commit 33f3d41

-------------------------------------------------------------------
Tue Aug 14 14:41:58 CEST 2012 - okir@suse.de

- network-nanny: unify ni_manager_{netdev,modem}_apply_policy
- commit e42929f

-------------------------------------------------------------------
Tue Aug 14 14:41:35 CEST 2012 - okir@suse.de

- Introduce ni_ifworker_type_to_string
- commit 85f5faf

-------------------------------------------------------------------
Tue Aug 14 14:39:16 CEST 2012 - okir@suse.de

- Introduce xml_node_print_debug()
- commit 28750fb

-------------------------------------------------------------------
Tue Aug 14 14:14:40 CEST 2012 - okir@suse.de

- network-nanny: unify ni_manager_{netdev,modem}_new
- commit f128a4d

-------------------------------------------------------------------
Tue Aug 14 13:11:44 CEST 2012 - okir@suse.de

- network-nanny: more cleanup of netdev/modem unification
- commit 6dbaab3

-------------------------------------------------------------------
Tue Aug 14 13:04:05 CEST 2012 - okir@suse.de

- network-nanny: unify ni_manager_remove_{netdev,modem}
- commit 167e074

-------------------------------------------------------------------
Tue Aug 14 12:52:53 CEST 2012 - okir@suse.de

- network-nanny: unify ni_manager_get_{netdev,modem}
- commit 7893c3b

-------------------------------------------------------------------
Tue Aug 14 12:45:18 CEST 2012 - okir@suse.de

- network-nanny: unify ni_objectmodel_unregister_managed_{modem,netdev}
- commit d01ad4a

-------------------------------------------------------------------
Tue Aug 14 12:40:21 CEST 2012 - okir@suse.de

- network-nanny: introduce generic managed_device_{new,free}
- commit e79853e

-------------------------------------------------------------------
Tue Aug 14 12:35:51 CEST 2012 - okir@suse.de

- When destroying a dbus object, apply destructors of super classes as well
- commit e6997b4

-------------------------------------------------------------------
Tue Aug 14 12:30:43 CEST 2012 - okir@suse.de

- Unify managed_{netdev,modem} into one struct
- commit 0a0ef5f

-------------------------------------------------------------------
Tue Aug 14 12:28:09 CEST 2012 - okir@suse.de

- network-nanny: get rid of modem field in managed_modem
- commit fd13082

-------------------------------------------------------------------
Tue Aug 14 12:27:12 CEST 2012 - okir@suse.de

- network-nanny: get rid of netdev field in managed_netdev
- commit fd2af0e

-------------------------------------------------------------------
Tue Aug 14 12:22:46 CEST 2012 - okir@suse.de

- network-nanny: fix various issues with modem hotplug
- commit c1ece96

-------------------------------------------------------------------
Tue Aug 14 12:21:48 CEST 2012 - okir@suse.de

- Introduce completion callback for ifworkers
- commit 023080a

-------------------------------------------------------------------
Tue Aug 14 12:16:39 CEST 2012 - okir@suse.de

- Introduce ni_fsm_destroy_worker
- commit 02dd55d

-------------------------------------------------------------------
Tue Aug 14 10:08:34 CEST 2012 - okir@suse.de

- Ongoing work on modem handling
- commit f48ae78

-------------------------------------------------------------------
Tue Aug 14 10:07:34 CEST 2012 - okir@suse.de

- Mark ifworkers for disappeared interfaces as "dead"
- commit 31be2e6

-------------------------------------------------------------------
Tue Aug 14 10:06:22 CEST 2012 - okir@suse.de

- Add ni_dbus_object_send_property and ni_dbus_object_send_property_string
- commit 8f492e8

-------------------------------------------------------------------
Mon Aug 13 15:37:53 CEST 2012 - okir@suse.de

- Work on modem bring-up in network-nanny
- commit 0a4426b

-------------------------------------------------------------------
Mon Aug 13 15:37:21 CEST 2012 - okir@suse.de

- Clean up handling of modem-manager signals a little bit
- commit 99f571a

-------------------------------------------------------------------
Mon Aug 13 14:59:27 CEST 2012 - okir@suse.de

- Add debug facility for modem handling
- commit 9a530b3

-------------------------------------------------------------------
Mon Aug 13 14:44:43 CEST 2012 - okir@suse.de

- Introduce a setClientInfo method for modem objects
  As part of the process, rename ni_netdev_clientinfo_t to
  ni_device_clientinfo_t and a client_info pointer to ni_modem_t.
- commit eecf99a

-------------------------------------------------------------------
Mon Aug 13 14:36:58 CEST 2012 - okir@suse.de

- Add ni_ifworker_check_config, ni_ifworker_is_running
- commit e731080

-------------------------------------------------------------------
Mon Aug 13 13:43:53 CEST 2012 - okir@suse.de

- network-nanny: manage a cache store of secrets
- commit f53140f

-------------------------------------------------------------------
Mon Aug 13 13:39:27 CEST 2012 - okir@suse.de

- Fix several issues in the FSM
  ni_ifworker_by_modem would always fail, even if the modem we were looking for existed
  When receiving current state from the server, we would always set existing devices to
  state DEVICE_UP, even if they had not associated yet. The correct state is DEVICE_EXISTS.
- commit 6215a0a

-------------------------------------------------------------------
Mon Aug 13 12:11:09 CEST 2012 - okir@suse.de

- Add a security_id field to the ifworker
- commit 618ba05

-------------------------------------------------------------------
Mon Aug 13 11:42:49 CEST 2012 - okir@suse.de

- XML schema validation: when prompting for a username/password, add the new node to the XML tree
  If prompting failed, remove it again.
  We do it this way round because it helps us identify the device owning this config item
- commit 6d5e62d

-------------------------------------------------------------------
Mon Aug 13 11:09:59 CEST 2012 - okir@suse.de

- FSM: introduce a callback mechanism when prompting for user names, passwords and the like
- commit 05f4746

-------------------------------------------------------------------
Mon Aug 13 10:34:34 CEST 2012 - okir@suse.de

- Simplify fsm policy condition checks a little
- commit a1cbccd

-------------------------------------------------------------------
Mon Aug 13 10:25:19 CEST 2012 - okir@suse.de

- Add a sample network-nanny policy for a umts stick
- commit d117523

-------------------------------------------------------------------
Mon Aug 13 10:23:41 CEST 2012 - okir@suse.de

- fsm-policy: add checks for modem properties
- commit 93a8330

-------------------------------------------------------------------
Mon Aug 13 10:21:31 CEST 2012 - okir@suse.de

- objectmodel for modems: send identifcation info as properties
- commit 63d0c77

-------------------------------------------------------------------
Mon Aug 13 10:20:52 CEST 2012 - okir@suse.de

- network-nanny: always enable user control for modems (for now)
- commit 3e45df8

-------------------------------------------------------------------
Mon Aug 13 10:20:26 CEST 2012 - okir@suse.de

- network-nanny: when policy is changed, check modems, too
- commit c334b32

-------------------------------------------------------------------
Mon Aug 13 10:19:42 CEST 2012 - okir@suse.de

- Add inline helper functions ni_ifworker_get_{netdev,modem}
- commit 12c246f

-------------------------------------------------------------------
Mon Aug 13 08:35:39 CEST 2012 - okir@suse.de

- Start to work on modem support in network-nanny
- commit 21d668d

-------------------------------------------------------------------
Mon Aug 13 08:08:25 CEST 2012 - okir@suse.de

- Refactor __ni_ifworker_refresh_modems
  Factor out some code and place it in ni_fsm_recv_new_modem()
- commit a7593c8

-------------------------------------------------------------------
Mon Aug 13 08:01:34 CEST 2012 - okir@suse.de

- Make ni_dbus_object_lookup recognize absolute vs relative path names
- commit 77512cf

-------------------------------------------------------------------
Mon Aug 13 08:00:38 CEST 2012 - okir@suse.de

- emit deviceCreate/deviceDelete events when new modems/netifs are created/deleted
- commit 32e96da

-------------------------------------------------------------------
Fri Aug 10 16:00:09 CEST 2012 - okir@suse.de

- modem-manager: handle older-style modem-manager signals
  These signals provided a string rather than an object-path.
- commit 871597a

-------------------------------------------------------------------
Fri Aug 10 15:59:41 CEST 2012 - okir@suse.de

- wickedd: always emit an event when a device (modem or netif) is created or destroyed
- commit e498c6c

-------------------------------------------------------------------
Tue Aug  7 09:20:36 CEST 2012 - okir@suse.de

- Unify the SIGTERM/SIGINT handling code in various mainloops
- commit 793b2de

-------------------------------------------------------------------
Tue Aug  7 09:04:33 CEST 2012 - okir@suse.de

- Renamed network-mgr to network-nanny
- commit a3dfa0e

-------------------------------------------------------------------
Fri Jul 27 16:12:45 CEST 2012 - okir@suse.de

- Make ManagedInterface.enable work for wireless, too
- commit 35db5be

-------------------------------------------------------------------
Fri Jul 27 15:45:43 CEST 2012 - okir@suse.de

- Another step towards a NetworkManager like service
  The new manager facilities can now detect link events on ethernet
  interfaces and bring up the interface in response.
  My test case right now:
  	ifconfig eth0 down
  	./bin/wicked --debug all manager enable eth0
  	./bin/wicked --debug all manager addpolicy samples/manager/eth-all.xml
- commit 640d138

-------------------------------------------------------------------
Fri Jul 27 15:41:34 CEST 2012 - okir@suse.de

- Some more FSM adjustments for the manager
- commit 616d4a7

-------------------------------------------------------------------
Fri Jul 27 15:41:13 CEST 2012 - okir@suse.de

- Add ni_dbus_message_append_object_path
- commit 612a1cf

-------------------------------------------------------------------
Fri Jul 27 15:38:48 CEST 2012 - okir@suse.de

- Implement new linkMonitor() method
  This allows NetworkManager-like clients to bring up an interface for
  monitoring (link nego events; wireless events; ...)
  This patch adds the requisite functions at the low-level, server side
  method and a client-side function.
- commit 3e13418

-------------------------------------------------------------------
Fri Jul 27 11:57:07 CEST 2012 - okir@suse.de

- Minor code cleanup: get rid of some old crud
- commit 18b0c47

-------------------------------------------------------------------
Fri Jul 27 11:54:35 CEST 2012 - okir@suse.de

- Fold ni_fsm_kickstart into ni_fsm_schedule()
- commit cd04490

-------------------------------------------------------------------
Fri Jul 27 11:47:21 CEST 2012 - okir@suse.de

- Add NI_OBJECTMODEL_MANAGED_POLICY_LIST_PATH
- commit d7b3b28

-------------------------------------------------------------------
Fri Jul 27 11:02:14 CEST 2012 - okir@suse.de

- FSM: expose a few parts of the FSM to the new manager code
- commit ce00c49

-------------------------------------------------------------------
Fri Jul 27 11:01:39 CEST 2012 - okir@suse.de

- Introduce ni_fsm_policies_changed_since
- commit 32000cc

-------------------------------------------------------------------
Thu Jul 26 13:29:29 CEST 2012 - mt@suse.de

- Fixed incorrect string copy to netdev name
- commit b9d69af

-------------------------------------------------------------------
Thu Jul 26 12:44:51 CEST 2012 - okir@suse.de

- Fix a bug in xml_document_from_string
- commit b25614a

-------------------------------------------------------------------
Thu Jul 26 12:34:05 CEST 2012 - okir@suse.de

- Fix up ni_dbus_translate_error to use our full error conversion maps
- commit 429c9e6

-------------------------------------------------------------------
Thu Jul 26 12:29:06 CEST 2012 - okir@suse.de

- Added NI_POLICY_EXISTS error code and strings.
- commit 252c3eb

-------------------------------------------------------------------
Thu Jul 26 12:28:30 CEST 2012 - okir@suse.de

- Added functions xml_document_sprint and xml_node_sprint
- commit 863c28e

-------------------------------------------------------------------
Thu Jul 26 11:25:41 CEST 2012 - okir@suse.de

- More work on the manager part
- commit 56e91de

-------------------------------------------------------------------
Thu Jul 26 11:23:05 CEST 2012 - okir@suse.de

- Minor improvement to debug output in fsm.c:interface_state_change_signal()
- commit 1610b76

-------------------------------------------------------------------
Thu Jul 26 10:33:00 CEST 2012 - okir@suse.de

- Made ni_ifworker_array_{append,index,destroy} public
- commit 847eb2d

-------------------------------------------------------------------
Thu Jul 26 10:00:30 CEST 2012 - okir@suse.de

- Renamed ni_ifworker_by_object_path to ni_fsm_ifworker_by_object_path and made it public
- commit 10785af

-------------------------------------------------------------------
Thu Jul 26 09:52:23 CEST 2012 - okir@suse.de

- Change ni_fsm_create_client() to return a dbus_client rather than a bool
- commit 3f68b3b

-------------------------------------------------------------------
Thu Jul 26 09:35:43 CEST 2012 - okir@suse.de

- Add magic cookie checks to dbus_variant handling
- commit 37ffb94

-------------------------------------------------------------------
Wed Jul 25 16:38:27 CEST 2012 - okir@suse.de

- Fix errors in ni_dbus_variant handling
  A couple of patches back, I introduced the ability to store the message
  pointer in the dbus_variant_t. This caused some pretty prominent oopses
  when we run ni_dbus_variant_destroy() on a uninitialized variant.
- commit 7ab6b47

-------------------------------------------------------------------
Wed Jul 25 16:02:22 CEST 2012 - okir@suse.de

- Started to work on manager
- commit 038f1c5

-------------------------------------------------------------------
Wed Jul 25 12:40:18 CEST 2012 - mt@suse.de

- Correctly copy na addrs to granted lease, cleanup
- commit 57393ee

-------------------------------------------------------------------
Wed Jul 25 09:59:09 CEST 2012 - mt@suse.de

- Added dhcp6 address configuration service scheme
- commit 2534b33

-------------------------------------------------------------------
Wed Jul 25 09:58:07 CEST 2012 - mt@suse.de

- Added dbus addrconf forwarder glue for dhcp6
- commit a8d2b91

-------------------------------------------------------------------
Tue Jul 24 17:40:44 CEST 2012 - mt@suse.de

- Added optional tags make target [ctags]
- commit e964097

-------------------------------------------------------------------
Tue Jul 24 17:39:19 CEST 2012 - mt@suse.de

- Add ia_na addrs with status OK to lease addrs
- commit d319036

-------------------------------------------------------------------
Tue Jul 24 17:21:41 CEST 2012 - okir@suse.de

- Make ni_objectmodel_register_service public
- commit c3bf2a7

-------------------------------------------------------------------
Tue Jul 24 12:57:21 CEST 2012 - okir@suse.de

- When backgrounding server processes, use the proper program name in syslog
- commit 4867610

-------------------------------------------------------------------
Tue Jul 24 11:27:12 CEST 2012 - okir@suse.de

- Renamed a bunch of function ni_ifworker_foo -> ni_fsm_foobar
- commit 40bf644

-------------------------------------------------------------------
Tue Jul 24 11:18:55 CEST 2012 - okir@suse.de

- fsm: allow run-time registration of new dependency checks,
  and use it to register the "reachable" check (which is not part of the
  library)
- commit c979024

-------------------------------------------------------------------
Tue Jul 24 11:07:43 CEST 2012 - okir@suse.de

- Move most of the fsm/ifworker/ifpolicy stuff into the library
- commit 2cf41ab

-------------------------------------------------------------------
Tue Jul 24 10:52:37 CEST 2012 - okir@suse.de

- Get rid of client/policy.h
- commit ef4fba4

-------------------------------------------------------------------
Tue Jul 24 10:47:50 CEST 2012 - okir@suse.de

- Rename ni_ifpolicy to ni_fsm_policy
- commit d7b1131

-------------------------------------------------------------------
Tue Jul 24 10:31:25 CEST 2012 - okir@suse.de

- Move client/calls.c into the library
- commit 9b8615f

-------------------------------------------------------------------
Tue Jul 24 10:25:31 CEST 2012 - okir@suse.de

- Rename wicked_get_{interface,modem}_object to ni_call_get_{netif,modem}_list_object
- commit 91663fa

-------------------------------------------------------------------
Tue Jul 24 10:12:20 CEST 2012 - okir@suse.de

- Move client/fsm.h to include/wicked/fsm.h
- commit b55d76f

-------------------------------------------------------------------
Tue Jul 24 10:10:20 CEST 2012 - okir@suse.de

- sed -i 's:ni_iftransition_:ni_fsm_transition_:g'
- commit 763f849

-------------------------------------------------------------------
Tue Jul 24 10:04:28 CEST 2012 - okir@suse.de

- ifup: get rid of global __root_object variable
  Move this into ni_fsm instead
- commit b72a8f6

-------------------------------------------------------------------
Tue Jul 24 09:57:58 CEST 2012 - okir@suse.de

- sed -i 's:ni_objectmodel_fsm_:ni_fsm_:g'
- commit dc49a58

-------------------------------------------------------------------
Tue Jul 24 09:55:29 CEST 2012 - okir@suse.de

- sed -i 's:\<STATE_:NI_FSM_&:g'
- commit 52b6894

-------------------------------------------------------------------
Tue Jul 24 09:53:06 CEST 2012 - okir@suse.de

- sed -i 's:ni_ifworker_req_:ni_fsm_require_:g'
- commit 00ea7f7

-------------------------------------------------------------------
Tue Jul 24 09:46:05 CEST 2012 - okir@suse.de

- Some more work on the zookeeper stuff - WIP
- commit 697a6c1

-------------------------------------------------------------------
Tue Jul 24 09:45:19 CEST 2012 - okir@suse.de

- Add new function xml_document_from_string()
- commit 2b77fa5

-------------------------------------------------------------------
Tue Jul 24 09:34:09 CEST 2012 - okir@suse.de

- Rename ni_objectmodel_register_device_service and make it public
  Rename ni_objectmodel_register_device_service to
  ni_objectmodel_register_netif_service and add it to objectmodel.h
- commit 2be9728

-------------------------------------------------------------------
Tue Jul 24 09:27:09 CEST 2012 - okir@suse.de

- Make ni_objectmodel_register_{netif,modem}_classes public
  Also, split off the service registrations into a separate
  ni_objectmodel_register_{netif,modem}_services()
- commit 7e95a2a

-------------------------------------------------------------------
Tue Jul 24 09:19:28 CEST 2012 - okir@suse.de

- Add com.suse.Wicked.Manager bus name
- commit 9da559f

-------------------------------------------------------------------
Mon Jul 23 15:40:24 CEST 2012 - mt@suse.de

- Removed package changelog entries from spec file
- commit 3ebfb2c

-------------------------------------------------------------------
Mon Jul 23 15:39:53 CEST 2012 - mt@suse.de

- Adopted spec file to use new init scripts
- commit 8c18126

-------------------------------------------------------------------
Mon Jul 23 15:36:57 CEST 2012 - mt@suse.de

- Removed obsolete wickedd init script
- commit 4d32f06

-------------------------------------------------------------------
Mon Jul 23 15:36:23 CEST 2012 - mt@suse.de

- Install new init scripts
- commit 12da630

-------------------------------------------------------------------
Mon Jul 23 15:34:53 CEST 2012 - mt@suse.de

- Added missed action dummies to etc/init.d/network
- commit 53e5985

-------------------------------------------------------------------
Mon Jul 23 14:03:55 CEST 2012 - okir@suse.de

- Start to work on zookeeper daemon
- commit a80c9da

-------------------------------------------------------------------
Mon Jul 23 13:56:27 CEST 2012 - okir@suse.de

- Make a few ni_ifworker_* functions public
- commit 6fad019

-------------------------------------------------------------------
Mon Jul 23 13:50:22 CEST 2012 - okir@suse.de

- Get rid of target_state range in ni_objectmodel_fsm_t
- commit 41f0631

-------------------------------------------------------------------
Mon Jul 23 13:37:24 CEST 2012 - okir@suse.de

- wicked client: introduce program_name variable (same as in servers)
- commit d775c5c

-------------------------------------------------------------------
Mon Jul 23 12:57:41 CEST 2012 - okir@suse.de

- Get rid of obsolete <pidfile> element in config.xml
- commit 1ce88c6

-------------------------------------------------------------------
Mon Jul 23 12:55:54 CEST 2012 - okir@suse.de

- All daemons were using the same pidfile path. Not good.
  Changed ni_server_background() to take an additional "appname" argumnt;
  build the pidfile path from that.
- commit c21f290

-------------------------------------------------------------------
Mon Jul 23 12:32:12 CEST 2012 - mt@suse.com

- Use NL_SKIP to ignore unwanted netlink messages
- commit d35894f

-------------------------------------------------------------------
Mon Jul 23 10:06:55 CEST 2012 - mt@suse.com

- Start dhcp6 supplicant in background by default
- commit ef5afc1

-------------------------------------------------------------------
Mon Jul 23 09:43:41 CEST 2012 - mt@suse.com

- Next incarnation step in dhcp6 supplicant
- commit be083b2

-------------------------------------------------------------------
Mon Jul 23 09:42:18 CEST 2012 - mt@suse.com

- Added ni_buffer_reset, return status in _putc
- commit 0559ceb

-------------------------------------------------------------------
Mon Jul 23 09:39:29 CEST 2012 - mt@suse.com

- Increased buffer in ni_print_hex for opaque data
- commit 6954456

-------------------------------------------------------------------
Mon Jul 23 09:37:44 CEST 2012 - mt@suse.com

- Use dhcp6.server_addr instead of server_unicast
- commit 254d7dc

-------------------------------------------------------------------
Mon Jul 23 09:37:18 CEST 2012 - mt@suse.com

- Added ni_opaque_eq utility function
- commit 167e66a

-------------------------------------------------------------------
Mon Jul 23 09:36:07 CEST 2012 - mt@suse.com

- Updated to use fqdn rfc reference instead of draft
- commit 143c0f2

-------------------------------------------------------------------
Thu Jul 19 14:13:37 CEST 2012 - okir@suse.de

- Cleanup TODO file a bit
- commit ef59c6a

-------------------------------------------------------------------
Thu Jul 19 13:59:28 CEST 2012 - okir@suse.de

- ifup: allow policies to override the <control> information
- commit fe62c79

-------------------------------------------------------------------
Thu Jul 19 13:29:16 CEST 2012 - okir@suse.de

- ifup: allow policy selection based on control mode and on boot-stage
- commit 1eef86b

-------------------------------------------------------------------
Thu Jul 19 12:43:31 CEST 2012 - okir@suse.de

- client: replace two pairs of min/max variables with a uint_range_t
- commit da589ba

-------------------------------------------------------------------
Thu Jul 19 12:38:26 CEST 2012 - okir@suse.de

- client: s:ni_netif_action:ni_iftransition:g
- commit 7ebe832

-------------------------------------------------------------------
Thu Jul 19 12:35:08 CEST 2012 - okir@suse.de

- Get rid of obsolete ni_ifworker_children_t and ni_ifworker_edge_t
- commit 28972d2

-------------------------------------------------------------------
Thu Jul 19 12:28:04 CEST 2012 - okir@suse.de

- Get rid of ni_ifworker_edge_precondition
- commit d72f2c0

-------------------------------------------------------------------
Thu Jul 19 12:25:12 CEST 2012 - okir@suse.de

- Rewrite how we handle constraints on the state of a subordinate device
  Up to now, there was some special hackery in the ifworker children array.
  Change the implementation to use generic "ni_ifworker_req" requirements
  (well, mostly generic).
- commit 8fdbc2f

-------------------------------------------------------------------
Thu Jul 19 11:46:25 CEST 2012 - okir@suse.de

- When deleting an ifworker, free the list of requirements (action->require.list)
- commit 1f31548

-------------------------------------------------------------------
Thu Jul 19 11:09:14 CEST 2012 - okir@suse.de

- ifup: check for duplicate netif-reference/modem-reference meta nodes when building ifworker hierarchy
- commit ad540f9

-------------------------------------------------------------------
Thu Jul 19 10:56:26 CEST 2012 - okir@suse.de

- Remove obsolete node member from ni_ifworker_edge_t
- commit 46af36f

-------------------------------------------------------------------
Wed Jul 18 11:19:57 CEST 2012 - okir@suse.de

- In the XML interface definition, change <boot-label> to <mode>
- commit 7e9cfaf

-------------------------------------------------------------------
Wed Jul 18 11:14:36 CEST 2012 - okir@suse.de

- ifup: make link detection optional; handle link detect timeout etc
- commit f168488

-------------------------------------------------------------------
Tue Jul 17 11:26:17 CEST 2012 - okir@suse.de

- ifworker: move uuid and config_origin into struct ifworker.config {}
- commit 8932ba4

-------------------------------------------------------------------
Tue Jul 17 11:23:34 CEST 2012 - okir@suse.de

- Move ifworker->config into ifworker->config.node
- commit d38705c

-------------------------------------------------------------------
Mon Jul 16 15:59:35 CEST 2012 - okir@suse.de

- Add partial manpage wicked-config.5
- commit 907c994

-------------------------------------------------------------------
Mon Jul 16 14:14:21 CEST 2012 - okir@suse.de

- Make system-updater extensions a first-class animal in the config.xml zoo
- commit 28a0317

-------------------------------------------------------------------
Mon Jul 16 13:54:02 CEST 2012 - okir@suse.de

- config.xml: rename <extension> to <dbus-service>
- commit 4c48613

-------------------------------------------------------------------
Mon Jul 16 12:46:42 CEST 2012 - okir@suse.de

- Update manpages to somewhat reflect current status
- commit 1d2b540

-------------------------------------------------------------------
Mon Jul 16 12:46:18 CEST 2012 - okir@suse.de

- wickedd: kill obsolete --no-fork option
- commit 963096e

-------------------------------------------------------------------
Mon Jul 16 09:59:52 CEST 2012 - okir@suse.de

- Two fixes for package building
- commit 0116a36

-------------------------------------------------------------------
Mon Jul 16 08:58:52 CEST 2012 - okir@suse.de

- ifup: add new option --skip-origin
- commit adbc074

-------------------------------------------------------------------
Fri Jul 13 17:25:17 CEST 2012 - okir@suse.de

- ifup: when bringing up interfaces configured via firmware, mark config-origin as "firmware"
- commit 48c12ab

-------------------------------------------------------------------
Fri Jul 13 17:21:03 CEST 2012 - okir@suse.de

- ifdown: avoid calling netif.setClientState on a device we just destroyed
- commit 7a18790

-------------------------------------------------------------------
Fri Jul 13 17:15:00 CEST 2012 - mt@suse.com

- Just moved some functions to their places
- commit 4d43648

-------------------------------------------------------------------
Fri Jul 13 16:35:13 CEST 2012 - okir@suse.de

- Fix invalid access to free'd memory when processing the response of a DBus call
  ni_dbus_object_call_variant would deserialize the message and store everything
  in a (tree of) ni_dbus_variant's. However, we record the keys of a dict as just
  the pointer rather than doing a strdup of the string - which is kind of bad if
  ni_dbus_object_call_variant then throws aways the response message.
  For this reason, introduce a __message pointer in ni_dbus_variant where we keep
  a pointer to the message this variant came from. The message is dereferenced when
  we throw away the variant eventually.
  This is obviously a time/space tradeoff; we trade the extra pointer in the variant
  for the time it takes to copy lots of little strings. Not sure what's better;
  maybe at one point we need to revisit this.
- commit bf6aecc

-------------------------------------------------------------------
Fri Jul 13 16:34:40 CEST 2012 - okir@suse.de

- Fix a bug in how we convert a UUID to XML in dbus-xml.c
  Thanks to valgrind.
- commit 7c8654e

-------------------------------------------------------------------
Fri Jul 13 14:56:01 CEST 2012 - mt@suse.com

- Adopted to use ni_dhcp6_status pointer in addr-lease
- commit 44f5c38

-------------------------------------------------------------------
Fri Jul 13 14:50:18 CEST 2012 - okir@suse.de

- Redo the netif.setClientState() method
  Change from netif.setClientState(uuid, state) to
  netif.setClientInfo(dict), where the dict is a ni_netdev_clientinfo_t
  wrapped in dbus regalia. Change all the code that references this stuff
- commit d8f2662

-------------------------------------------------------------------
Fri Jul 13 14:27:06 CEST 2012 - okir@suse.de

- Introduce config_origin field to ni_netdev_clientinfo
- commit 0b3b28d

-------------------------------------------------------------------
Fri Jul 13 14:20:40 CEST 2012 - mt@suse.com

- Use a forward declaration of ni_dhcp6_status only
- commit 38c17b9

-------------------------------------------------------------------
Fri Jul 13 13:20:37 CEST 2012 - okir@suse.de

- ni_netdev_t: move client information into a struct of its own
  Until now, ni_netdev_t contained the fields client_state and uuid.
  Move these into a struct of their own (ni_netdev_clientinfo_t), and
  replace the the two ni_netdev members with a pointer to this struct.
  Adapt all dbus-object ifup code that uses this information.
- commit 06e3bbe

-------------------------------------------------------------------
Fri Jul 13 13:16:41 CEST 2012 - mt@suse.com

- Changed to use ni_timeout_param and functions
- commit 57e4699

-------------------------------------------------------------------
Fri Jul 13 13:14:37 CEST 2012 - mt@suse.com

- Return timeout in ni_timeout_arm for recalculations
  Added ni_timeout_arm_msec variant and a randomize utility.
- commit 4e17236

-------------------------------------------------------------------
Fri Jul 13 12:23:02 CEST 2012 - okir@suse.de

- Remove obsolete parent pointer from ni_netdev_t
- commit 5f36a1b

-------------------------------------------------------------------
Fri Jul 13 08:57:55 CEST 2012 - okir@suse.de

- ifup: use xml_node_hash() to generate a unique finger print of a configuration
- commit c4411f7

-------------------------------------------------------------------
Fri Jul 13 08:56:33 CEST 2012 - okir@suse.de

- ni_hashctx_get_digest should truncate the result rathern than error
  ni_hashctx_get_digest: if the return buffer is too small, truncate
  the result rathern than error out
- commit 4b07039

-------------------------------------------------------------------
Thu Jul 12 17:47:11 CEST 2012 - okir@suse.de

- Introduce xml_document_hash and xml_node_hash to fingerprint XML docs
- commit 109b674

-------------------------------------------------------------------
Thu Jul 12 17:46:55 CEST 2012 - okir@suse.de

- Introduce functions for hashing data with md5
- commit 7047479

-------------------------------------------------------------------
Wed Jul 11 18:52:05 CEST 2012 - mt@suse.com

- Disarm and cleanup the lease+dhcp6 additions after
  Current status is, that it sends SOLICIT, receives and parses
  responses (ADVERTISE) and discards the data again.
- commit e214443

-------------------------------------------------------------------
Wed Jul 11 18:15:28 CEST 2012 - okir@suse.de

- Move ni_timeout_* functions from capture.c to timer.c
- commit 800f095

-------------------------------------------------------------------
Wed Jul 11 18:12:07 CEST 2012 - okir@suse.de

- Make ni_timeout_param_t public and move prototypes into wicked/socket.h
  ni_timeout_recompute() can be used by send/retransmit code to adjust
  the current timeout. Depending on the settings in ni_timeout_param_t,
  it will do contant timeouts, linear backoff or exponential backoff,
  respectively.
  ni_timeout_arm() will compute a deadline (now + timeout) using the
  information from ni_timeout_param_t. If requested, it will apply some
  jitter in computing the timeout.
- commit 9a0d3e5

-------------------------------------------------------------------
Wed Jul 11 17:43:30 CEST 2012 - okir@suse.de

- Remove obsolete member from ni_timeout_param_t
- commit 1a9c074

-------------------------------------------------------------------
Wed Jul 11 17:15:32 CEST 2012 - okir@suse.de

- Add enablement for dhcp6-supplicant in a number of places
- Start/stop dhcp6-supplicant in wicked init script
- Add dhcp6-supplicant to the spec file
- Add a dbus config file for com.suse.Wicked.DHCP6
- commit 6c8ca35

-------------------------------------------------------------------
Wed Jul 11 17:13:06 CEST 2012 - okir@suse.de

- Move some purely dhcp6 related structs out of public wicked/addrconf.h file
- commit 9f6a142

-------------------------------------------------------------------
Wed Jul 11 16:45:24 CEST 2012 - okir@suse.de

- Create fake /sys tree for testing extension/ibft
- commit cf49318

-------------------------------------------------------------------
Wed Jul 11 16:43:28 CEST 2012 - okir@suse.de

- dbus-objectmodel: pass around the address cacheinfo when present
- commit 909cf4f

-------------------------------------------------------------------
Wed Jul 11 16:29:41 CEST 2012 - okir@suse.de

- Store IPv6 addr/route preferred and valid lifetime in ni_{address,route}_t
- commit 5d32823

-------------------------------------------------------------------
Wed Jul 11 16:16:54 CEST 2012 - mt@suse.com

- Adopted to use ni_config_statedir() for state dir
- commit d14184d

-------------------------------------------------------------------
Wed Jul 11 15:41:15 CEST 2012 - okir@suse.de

- Make __ni_netdev_add_autoconf_prefix static
- commit 86a313a

-------------------------------------------------------------------
Wed Jul 11 15:38:52 CEST 2012 - okir@suse.de

- Get rid of "expires" field in ni_{address,route}_t
- commit 81fa10c

-------------------------------------------------------------------
Wed Jul 11 15:38:35 CEST 2012 - okir@suse.de

- First stab at ibft firmware-discovery script
- commit 3084373

-------------------------------------------------------------------
Wed Jul 11 15:20:03 CEST 2012 - mt@suse.com

- Enabled build of dhcp6-supplicant
- commit 02628ea

-------------------------------------------------------------------
Wed Jul 11 15:18:34 CEST 2012 - mt@suse.com

- First step of a dhcp6 supplicant implementation
- commit 382ef4b

-------------------------------------------------------------------
Wed Jul 11 15:13:13 CEST 2012 - mt@suse.com

- First shot of /config/addrconf/dhcp6 node parsing
- commit 8d46775

-------------------------------------------------------------------
Wed Jul 11 15:11:41 CEST 2012 - mt@suse.com

- Added initial dhcp6 structure to ni_addrconf_lease
- commit ac2d215

-------------------------------------------------------------------
Wed Jul 11 15:10:20 CEST 2012 - mt@suse.com

- Declared dhcp6 dbus objectmodel constants
- commit 4a5e67d

-------------------------------------------------------------------
Wed Jul 11 14:11:25 CEST 2012 - okir@suse.de

- In <device> <identify>, support an <ifindex> lookup
- commit 5f392de

-------------------------------------------------------------------
Wed Jul 11 12:49:09 CEST 2012 - okir@suse.de

- gitignore: ignore autogen files in etc/init.d
- commit dcb4414

-------------------------------------------------------------------
Wed Jul 11 11:22:34 CEST 2012 - okir@suse.de

- Add dummy ibft extension script
- commit 6938613

-------------------------------------------------------------------
Wed Jul 11 11:20:15 CEST 2012 - okir@suse.de

- In make distclean, remove autogenerated files in etc/init.d
- commit 5e5dd2c

-------------------------------------------------------------------
Wed Jul 11 11:07:59 CEST 2012 - okir@suse.de

- Introduce two new init scripts for wicked
- commit 43f01c1

-------------------------------------------------------------------
Wed Jul 11 11:02:22 CEST 2012 - okir@suse.de

- Introduce ni_netconfig_firmware_discovery and use it in ifup/ifdown
  This is a client-side function which executes all scripts listed in
  the configuration file's <netif-firmware-discovery> extension.
  The scripts are expected to produce a list of <interface> elements,
  which we will then bring up or down as requested.
- commit 5c0059a

-------------------------------------------------------------------
Wed Jul 11 11:01:02 CEST 2012 - okir@suse.de

- In __ni_process_run, print meaningful error if argv[0] is not executable
- commit d861a1a

-------------------------------------------------------------------
Wed Jul 11 10:59:37 CEST 2012 - okir@suse.de

- Introduce xml_document_from_buffer()
  This function processes a ni_buffer_t and builds an XML document from its
  content.
- commit dca19f8

-------------------------------------------------------------------
Wed Jul 11 10:58:03 CEST 2012 - okir@suse.de

- Introduce several new buffer handling functions
  ni_buffer_new, ni_buffer_new_dynamic and ni_buffer_free
  for allocating and freeing buffer objects
  ni_buffer_ungetc for putting a character back into a buffer.
- commit c75587b

-------------------------------------------------------------------
Wed Jul 11 10:47:46 CEST 2012 - okir@suse.de

- Introduce ni_file_executable()
- commit cd07e81

-------------------------------------------------------------------
Tue Jul 10 19:22:11 CEST 2012 - okir@suse.de

- Implement new function ni_process_run_and_capture_output
- commit f08e31a

-------------------------------------------------------------------
Tue Jul 10 18:15:27 CEST 2012 - okir@suse.de

- Implement <netif-firmware-discovery> element in config.xml
  This element lets you specify one or more scripts that will
  discover interfaces configured by the firmware (such as iBFT)
- commit 4966b1c

-------------------------------------------------------------------
Tue Jul 10 18:07:41 CEST 2012 - okir@suse.de

- config.xml: unify handling of objectmodel extensions and netif-ns extensions
- commit c4ee6ef

-------------------------------------------------------------------
Tue Jul 10 17:07:57 CEST 2012 - mt@suse.com

- Added NULL checking ni_string_len utility function
- commit df94f5d

-------------------------------------------------------------------
Tue Jul 10 17:06:33 CEST 2012 - mt@suse.com

- Increased ni_opaque_t data size to 130 for duid
- commit 5cb3671

-------------------------------------------------------------------
Tue Jul 10 17:03:39 CEST 2012 - mt@suse.com

- Changed ni_buffer_put to return status
- commit a2a3f11

-------------------------------------------------------------------
Tue Jul 10 17:01:37 CEST 2012 - mt@suse.com

- Added ni_timer_get_time function to use for timers
- commit 517eaf4

-------------------------------------------------------------------
Tue Jul 10 16:55:06 CEST 2012 - mt@suse.com

- Avoid several poll timeout 0 loops for 1 millisecond
- commit 0ddf13c

-------------------------------------------------------------------
Mon Jul  9 10:57:48 CEST 2012 - okir@suse.de

- ifup: support new option --boot-stage
- commit bcfc68a

-------------------------------------------------------------------
Mon Jun 18 19:46:15 CEST 2012 - mt@suse.com

- Fixed install-bin target to depend on lib, not itself
- commit be9f78b

-------------------------------------------------------------------
Mon Jun 18 18:04:00 CEST 2012 - mt@suse.com

- Exclude all dot files from dist archive
- commit 7bdfdbd

-------------------------------------------------------------------
Mon Jun 18 12:34:53 CEST 2012 - okir@suse.de

- Remove two obsolete defines
- commit 8769cd7

-------------------------------------------------------------------
Mon Jun 18 12:34:33 CEST 2012 - okir@suse.de

- Propagate configure etcdir to NI_DEFAULT_CONFIG_PATH
- commit 10f9048

-------------------------------------------------------------------
Mon Jun 18 12:28:09 CEST 2012 - okir@suse.de

- Get rid of CONFIG_WICKED_STATEDIR
- commit 15e423f

-------------------------------------------------------------------
Mon Jun 18 11:29:40 CEST 2012 - okir@suse.de

- Introduce ni_mkdir_maybe
- commit b91cfd1

-------------------------------------------------------------------
Mon Jun 18 09:42:06 CEST 2012 - okir@suse.de

- Get rid of obsolete ni_afinfo_t and related functions
- commit 1141ba1

-------------------------------------------------------------------
Fri Jun 15 15:30:47 CEST 2012 - okir@suse.de

- Move ni_resolve_*_timed into libwicked
- commit 3383895

-------------------------------------------------------------------
Fri Jun 15 15:26:15 CEST 2012 - okir@suse.de

- Move ni_sockaddr_t definition out of netinfo.h
- commit 2757669

-------------------------------------------------------------------
Fri Jun 15 12:09:36 CEST 2012 - okir@suse.de

- Add <device-alias> conditional in policy elements
- commit 8247f22

-------------------------------------------------------------------
Fri Jun 15 11:32:02 CEST 2012 - okir@suse.de

- Introduce ni_ifworker_match_alias()
- commit 58e3765

-------------------------------------------------------------------
Fri Jun 15 10:51:04 CEST 2012 - okir@suse.de

- Minor coding style changes in sysfs ibft code
- commit 0989388

-------------------------------------------------------------------
Fri Jun 15 09:59:01 CEST 2012 - okir@suse.de

- In XML documents, format UUIDs a little more nicely
  Introduce a uuid notation, and make the uuid-type type use it.
- commit 220e034

-------------------------------------------------------------------
Fri Jun 15 09:44:57 CEST 2012 - okir@suse.de

- More work on xml policies
- commit 56e0b79

-------------------------------------------------------------------
Thu Jun 14 16:04:44 CEST 2012 - mt@suse.com

- Fixed binary & config file checks in init script
- commit a3d9d73

-------------------------------------------------------------------
Thu Jun 14 16:04:12 CEST 2012 - mt@suse.com

- Fixed double free (fclose) in __ni_pidfile_write
- commit e551c4f

-------------------------------------------------------------------
Thu Jun 14 15:22:46 CEST 2012 - mt@suse.com

- Added sockaddr sa to ni_sockaddr union to avoid casts
- commit db973f2

-------------------------------------------------------------------
Thu Jun 14 15:18:41 CEST 2012 - mt@suse.com

- Set underflow in _buffer_pull_head, not overflow
- commit 713cb29

-------------------------------------------------------------------
Thu Jun 14 15:08:34 CEST 2012 - mt@suse.com

- Use xmalloc, xrealloc in buffer_ensure_tailroom
- commit d89fd4e

-------------------------------------------------------------------
Thu Jun 14 15:05:41 CEST 2012 - mt@suse.com

- Added guards to for buffer_count and _tailroom
- commit e64474e

-------------------------------------------------------------------
Thu Jun 14 15:03:16 CEST 2012 - mt@suse.com

- Use xcalloc when allocating socket struct
- commit 7f14a64

-------------------------------------------------------------------
Thu Jun 14 15:02:37 CEST 2012 - mt@suse.com

- Added also xmalloc and xrealloc wrappers
- commit 6764b05

-------------------------------------------------------------------
Thu Jun 14 15:00:35 CEST 2012 - mt@suse.com

- Commented out unused variables causing build failure
- commit cf8b5e0

-------------------------------------------------------------------
Wed Jun 13 17:11:14 CEST 2012 - okir@suse.de

- Introduce xml_node_array type and functions
- commit 69f1d57

-------------------------------------------------------------------
Wed Jun 13 16:42:05 CEST 2012 - okir@suse.de

- Remove current policy merging code, and get rid of ni_ifpolicy_rebind_action
- commit 0244c0a

-------------------------------------------------------------------
Wed Jun 13 16:39:18 CEST 2012 - okir@suse.de

- Introduce xml_node_merge()
- commit 9794f33

-------------------------------------------------------------------
Wed Jun 13 15:25:57 CEST 2012 - okir@suse.de

- Make xml_node_clone() clone the location information as well
- commit d17f948

-------------------------------------------------------------------
Wed Jun 13 14:33:23 CEST 2012 - okir@suse.de

- Introduce a couple of __xml_node_list_* helper functions
- commit 142e7c6

-------------------------------------------------------------------
Wed Jun 13 13:08:00 CEST 2012 - okir@suse.de

- Introduce xml_node_clone_ref to increment a node's refcount
- commit c195913

-------------------------------------------------------------------
Wed Jun 13 13:01:14 CEST 2012 - okir@suse.de

- In xml_document_new(), use xml_node_new rather than calloc
  This makes sure the refcount of doc->root is correct, and we don't
  crash in xml_document_free later on
- commit 7298d26

-------------------------------------------------------------------
Mon Jun 11 13:15:13 CEST 2012 - okir@suse.de

- Fix compile error in com.suse.Wicked.Protocol.IPv6 property table
- commit 622e521

-------------------------------------------------------------------
Fri Jun  8 11:34:26 CEST 2012 - okir@suse.de

- Make xml_node_t a refcounted object
- commit dcbd15c

-------------------------------------------------------------------
Wed Jun  6 12:48:12 CEST 2012 - okir@suse.de

- Support IPv4 device attributes (forwarding, accept-redirects)
- commit 0329b00

-------------------------------------------------------------------
Wed Jun  6 12:35:56 CEST 2012 - okir@suse.de

- Move ipv6 devinfo get/set code into ipv6.c
- commit dd4baa8

-------------------------------------------------------------------
Wed Jun  6 12:10:35 CEST 2012 - okir@suse.de

- Make netdev.ipv4/6 a pointer rather than a struct
- commit 261548d

-------------------------------------------------------------------
Wed Jun  6 11:50:23 CEST 2012 - okir@suse.de

- mv src/dbus-objects/{protocol,ipv6}.c
- commit eff6592

-------------------------------------------------------------------
Wed Jun  6 11:34:46 CEST 2012 - okir@suse.de

- Changed netdev.ipv6 and ipv4 from ni_afinfo to a new ip*devinfo type
- commit eede1cf

-------------------------------------------------------------------
Wed Jun  6 08:29:38 CEST 2012 - okir@suse.de

- Add support for changeProtocol to change IPv4/6 device settings
- commit 43fd7b2

-------------------------------------------------------------------
Tue Jun  5 17:01:01 CEST 2012 - okir@suse.de

- Bigfux: ni_file_open(..., "w") failed to truncate the file
- commit 7efdfeb

-------------------------------------------------------------------
Tue Jun  5 16:29:10 CEST 2012 - okir@suse.de

- Introduce <enabled> element to dhcp and autoip4 configuration
- commit 96d2b15

-------------------------------------------------------------------
Tue Jun  5 16:19:23 CEST 2012 - okir@suse.de

- *blush* when parsing a dbus-bool, we actually mapped "false" to TRUE
- commit 9d16491

-------------------------------------------------------------------
Tue Jun  5 15:55:45 CEST 2012 - okir@suse.de

- fixed syntax error in policy.xml
- commit 7ec6f3b

-------------------------------------------------------------------
Tue Jun  5 15:54:13 CEST 2012 - okir@suse.de

- Fix interface matching in ifup case
- commit 3edbc73

-------------------------------------------------------------------
Tue Jun  5 15:49:21 CEST 2012 - okir@suse.de

- Introducing interface policy
- commit a56a1ee

-------------------------------------------------------------------
Tue Jun  5 15:49:00 CEST 2012 - okir@suse.de

- Some more miniscule README changes
- commit ca4052d

-------------------------------------------------------------------
Tue Jun  5 15:30:44 CEST 2012 - okir@suse.de

- dhcp4-supplicant: Fix a typo in the dhcp4-state.xml pathname
- commit 56ecc56

-------------------------------------------------------------------
Fri Apr 20 17:53:08 CEST 2012 - okir@suse.de

- Implement 'wicked ifcheck'
- commit a606d9e

-------------------------------------------------------------------
Fri Apr 20 11:54:07 CEST 2012 - okir@suse.de

- More cleanups of the <require> handling
- commit 1db2ed7

-------------------------------------------------------------------
Fri Apr 20 11:40:48 CEST 2012 - okir@suse.de

- ifup: move reachability checking into a separate .c file
- commit a2ffde2

-------------------------------------------------------------------
Fri Apr 20 11:10:30 CEST 2012 - okir@suse.de

- ifup: reduce the number of static variables - introduce fsm object
- commit 6317f94

-------------------------------------------------------------------
Fri Apr 20 09:36:14 CEST 2012 - okir@suse.de

- ifup: moved type defs etc into separate header file
- commit 83b385e

-------------------------------------------------------------------
Fri Apr 20 09:06:08 CEST 2012 - okir@suse.de

- ifup: minor cleanups
- commit 9361475

-------------------------------------------------------------------
Fri Apr 20 08:51:10 CEST 2012 - okir@suse.de

- ifup: clean up ifworker struct by grouping several members into a struct fsm {}
- commit b11cfb1

-------------------------------------------------------------------
Fri Apr 20 08:43:44 CEST 2012 - okir@suse.de

- ifup: make sure we free the list of FSM actions in ni_ifworker_free
- commit 9e32996

-------------------------------------------------------------------
Fri Apr 20 08:31:51 CEST 2012 - okir@suse.de

- ifup: for each netif worker, generate a UUID for the configuration
  This uuid is based on the file's meta information, such as size and
  inode number.
- commit e9eb589

-------------------------------------------------------------------
Fri Apr 20 08:31:47 CEST 2012 - okir@suse.de

- Introduce ni_uuid_for_file
- commit 542bb18

-------------------------------------------------------------------
Thu Apr 19 19:03:52 CEST 2012 - okir@suse.de

- wicked client: Update the server's client-state whenever it changes
- commit a624021

-------------------------------------------------------------------
Thu Apr 19 18:49:51 CEST 2012 - okir@suse.de

- Use ni_uuid_print instead of ni_print_hex where possible
- commit ab2fbda

-------------------------------------------------------------------
Thu Apr 19 18:45:41 CEST 2012 - okir@suse.de

- objectmodel: use dbus_variant_*uuid functions where possible
- commit b8bef69

-------------------------------------------------------------------
Thu Apr 19 18:40:09 CEST 2012 - okir@suse.de

- Introduce Interface.setClientState()
- commit cadb3d8

-------------------------------------------------------------------
Thu Apr 19 18:26:59 CEST 2012 - okir@suse.de

- netif: add client-state property
  Clean up the property table a little in the process.
- commit b0b0292

-------------------------------------------------------------------
Thu Apr 19 18:09:19 CEST 2012 - okir@suse.de

- Introduce generic support for uuid properties
- commit 77f8444

-------------------------------------------------------------------
Thu Apr 19 17:57:46 CEST 2012 - okir@suse.de

- Moved some stuff to separate wicked/dbus-service.h file
  This concerns much of the stuff needed to define properties
  of a DBus service.
- commit c590a49

-------------------------------------------------------------------
Thu Apr 19 17:57:24 CEST 2012 - okir@suse.de

- Fix compile problem introduced by recent change to ni_dbus_object_get_properties_as_dict
- commit 6f5bb4b

-------------------------------------------------------------------
Thu Apr 19 17:52:34 CEST 2012 - okir@suse.de

- Move ni_dbus_object_set_properties_from_dict to dbus-object.c
- commit a733386

-------------------------------------------------------------------
Thu Apr 19 16:20:54 CEST 2012 - okir@suse.de

- dhcp4: re-discover and re-acquire leases on restart
- commit f66319a

-------------------------------------------------------------------
Thu Apr 19 14:58:19 CEST 2012 - okir@suse.de

- DHCP4 properties now show the original request.
  This is needed if we want to preserve the state across restarts.
- commit 4467663

-------------------------------------------------------------------
Thu Apr 19 14:23:43 CEST 2012 - okir@suse.de

- Change ni_dbus_object_set_properties_from_dict to take an additional DBusError argument
- commit 53c0a40

-------------------------------------------------------------------
Thu Apr 19 14:20:55 CEST 2012 - okir@suse.de

- Change ni_dbus_object_get_properties_as_dict() to take an additional DBusError argument
- commit b8284ab

-------------------------------------------------------------------
Thu Apr 19 14:17:14 CEST 2012 - okir@suse.de

- Move NI_DBUS_DICT_SIGNATURE to wicked/dbus.h
- commit 5d00704

-------------------------------------------------------------------
Thu Apr 19 13:56:33 CEST 2012 - okir@suse.de

- dhcp4-supplicant: when extracting the dhcp device from a dbus object, follow the usual conventions
- commit 7fe53e8

-------------------------------------------------------------------
Thu Apr 19 13:52:15 CEST 2012 - okir@suse.de

- Make dhcp4-supplicant write its state to an xml file on exit
- commit f2a9e31

-------------------------------------------------------------------
Thu Apr 19 13:49:58 CEST 2012 - okir@suse.de

- Get rid of old opt_personality bogosity
- commit ae86bc7

-------------------------------------------------------------------
Thu Apr 19 13:41:44 CEST 2012 - okir@suse.de

- Move the code for handling state.xml files to ni_objectmodel space
- commit b21ed74

-------------------------------------------------------------------
Thu Apr 19 13:30:53 CEST 2012 - okir@suse.de

- Make __ni_objectmodel_schema available to all objectmodel functions
- commit 69a9ffc

-------------------------------------------------------------------
Thu Apr 19 13:23:12 CEST 2012 - okir@suse.de

- When saving the wicked server state, make sure the file is mode 0600
- commit 884b254

-------------------------------------------------------------------
Thu Apr 19 13:11:25 CEST 2012 - okir@suse.de

- Introduce ni_file_open and use it in __ni_pidfile_write
- commit 89eca0a

-------------------------------------------------------------------
Thu Apr 19 12:28:43 CEST 2012 - okir@suse.de

- wickedd: remove some dead code
- commit 17a621d

-------------------------------------------------------------------
Thu Apr 19 12:24:41 CEST 2012 - okir@suse.de

- Recover lease information on server restart
- commit 0401f3c

-------------------------------------------------------------------
Thu Apr 19 12:09:22 CEST 2012 - okir@suse.de

- On server exit, write the interface state to a file
- commit bda2f0c

-------------------------------------------------------------------
Thu Apr 19 11:52:13 CEST 2012 - okir@suse.de

- wickedd: catch termination signals (SIGINT, SIGTERM)
- commit 72f2200

-------------------------------------------------------------------
Thu Apr 19 11:47:16 CEST 2012 - okir@suse.de

- Introduce ni_dbus_xml_serialize_properties
- commit 3b5407b

-------------------------------------------------------------------
Thu Apr 19 11:46:44 CEST 2012 - okir@suse.de

- Minor fix to ni_objectmodel_object_by_path
  Only try to convert the given path to a relative path if it starts
  with a '/'
- commit 02f90b1

-------------------------------------------------------------------
Thu Apr 19 10:23:12 CEST 2012 - okir@suse.de

- Clean up function naming in the objectmodel source code wrt interface/netif/netdevice
  Try to standardize on the term "netif" and rename various functions that used
  "interface" or "netdevice" or simply "device" in their names.
- commit 4c863e6

-------------------------------------------------------------------
Thu Apr 19 09:51:25 CEST 2012 - okir@suse.de

- Emit a resolverUpdated event when we changed the resolver
- commit 8adf1a3

-------------------------------------------------------------------
Thu Apr 19 09:36:25 CEST 2012 - okir@suse.de

- Some changes to the interface event handler
  Rename ni_server_listen_events -> ni_server_listen_interface_events
  and remove the first argument (ni_netconfig_t *) from the list of arguments
  to the event handler.
- commit 9b5ed33

-------------------------------------------------------------------
Wed Apr 18 20:46:44 CEST 2012 - okir@suse.de

- sed -i 's:ni_objectmodel_modem_unwrap:ni_objectmodel_unwrap_modem:'
- commit 28aa807

-------------------------------------------------------------------
Wed Apr 18 15:46:48 CEST 2012 - okir@suse.de

- Implementing PPP support, stage 2
- commit b2702c0

-------------------------------------------------------------------
Wed Apr 18 13:22:19 CEST 2012 - okir@suse.de

- Try to make the tempdir handling used in openvpn a little more generic
- commit e044d96

-------------------------------------------------------------------
Wed Apr 18 13:01:20 CEST 2012 - okir@suse.de

- Implementing PPP support, stage 1
- commit dc8fd96

-------------------------------------------------------------------
Wed Apr 18 10:51:30 CEST 2012 - mt@suse.com

- Ignore all files beginning with a dot.
- commit 89ea490

-------------------------------------------------------------------
Mon Apr 16 09:04:44 CEST 2012 - okir@suse.de

- Don't loop forever in __ni_tuntap_create_tun if user specified netdev exists already
- commit 91e5dec

-------------------------------------------------------------------
Sun Apr 15 10:24:51 CEST 2012 - okir@suse.de

- Work a bit on ppp schema file
- commit 112e05f

-------------------------------------------------------------------
Sun Apr 15 10:09:03 CEST 2012 - okir@suse.de

- ifup: allow <interface> elements to specify a device class using link-type="..." attribute
- commit 3066f11

-------------------------------------------------------------------
Sun Apr 15 09:32:10 CEST 2012 - okir@suse.de

- Some updates to PPP schema
- commit eaea81e

-------------------------------------------------------------------
Sat Apr 14 10:38:27 CEST 2012 - okir@suse.de

- Update ChangeLog
- commit a8dc44f

-------------------------------------------------------------------
Sat Apr 14 10:35:33 CEST 2012 - okir@suse.de

- Rename NI_EVENT_LINK_{CREATE,DELETE} to NI_EVENT_DEVICE_{CREATE,DELETE}
- commit cd84988

-------------------------------------------------------------------
Sat Apr 14 10:30:39 CEST 2012 - okir@suse.de

- Change all NI_ERROR_INTERFACE_* errors to NI_ERROR_DEVICE_*
  Do this for the corresponding DBus errors as well.
- commit 5ddd790

-------------------------------------------------------------------
Sat Apr 14 10:25:40 CEST 2012 - okir@suse.de

- Clean up error code conventions a bit
  We used to have a few functions left that returned libc error codes rather
  than NI_ERROR_* values. Remove these last vestiges and straighten things out.
- commit d94a9e6

-------------------------------------------------------------------
Sat Apr 14 09:58:14 CEST 2012 - okir@suse.de

- Fixed an error message in the wicked client
- commit 60cb6d0

-------------------------------------------------------------------
Sat Apr 14 09:57:18 CEST 2012 - okir@suse.de

- Removed utms.xml
- commit ee72336

-------------------------------------------------------------------
Fri Apr 13 21:41:44 CEST 2012 - okir@suse.de

- modem schema: list the events wickedd emits
- commit 160517b

-------------------------------------------------------------------
Fri Apr 13 21:33:20 CEST 2012 - okir@suse.de

- modem naming service: allow matching by manufacturer, model and version
- commit feb6f07

-------------------------------------------------------------------
Fri Apr 13 21:32:54 CEST 2012 - okir@suse.de

- ModemManager: retrieve manuf/model/version info when discovering a modem
- commit 5e382df

-------------------------------------------------------------------
Fri Apr 13 21:32:02 CEST 2012 - okir@suse.de

- dbus: handle dbus structs in messages received
- commit bbc9634

-------------------------------------------------------------------
Fri Apr 13 21:31:32 CEST 2012 - okir@suse.de

- Introduce ni_string_eq_nocase
- commit 5cb5b98

-------------------------------------------------------------------
Fri Apr 13 20:48:12 CEST 2012 - okir@suse.de

- Implement modem Connect()
- commit 8851bf0

-------------------------------------------------------------------
Fri Apr 13 20:11:22 CEST 2012 - okir@suse.de

- Include sample utms config file
- commit 9e3b8c0

-------------------------------------------------------------------
Fri Apr 13 18:06:49 CEST 2012 - okir@suse.de

- Propagate modem deviceDown events to client
- commit 08638b3

-------------------------------------------------------------------
Fri Apr 13 17:52:13 CEST 2012 - okir@suse.de

- modem-manager: fix handling of MM events with more than one param
- commit c692105

-------------------------------------------------------------------
Fri Apr 13 17:52:11 CEST 2012 - okir@suse.de

- Small improvement to __ni_dbus_print_argument
- commit cc370f9

-------------------------------------------------------------------
Thu Apr 12 21:13:37 CEST 2012 - okir@suse.de

- wicked server: propagate modem link association signals to the clients
- commit bee23c2

-------------------------------------------------------------------
Thu Apr 12 21:13:04 CEST 2012 - okir@suse.de

- ModemManager: capture StateChange signals and emit Link Association signals as needed
- commit d82b83f

-------------------------------------------------------------------
Thu Apr 12 21:10:49 CEST 2012 - okir@suse.de

- Implement ni_objectmodel_modem_event()
- commit 052e906

-------------------------------------------------------------------
Thu Apr 12 18:38:02 CEST 2012 - okir@suse.de

- Support some ModemManager signals
- commit 5595726

-------------------------------------------------------------------
Thu Apr 12 17:37:48 CEST 2012 - okir@suse.de

- More work on modem support
- commit 7478223

-------------------------------------------------------------------
Thu Apr 12 17:36:23 CEST 2012 - okir@suse.de

- Add new function ni_string_empty()
- commit 695ca49

-------------------------------------------------------------------
Thu Apr 12 17:35:57 CEST 2012 - okir@suse.de

- More schema changes for ppp/umts
- commit a3e61ad

-------------------------------------------------------------------
Thu Apr 12 17:35:36 CEST 2012 - okir@suse.de

- Add naming service for modem devices
- commit 74e3a12

-------------------------------------------------------------------
Thu Apr 12 17:34:39 CEST 2012 - okir@suse.de

- xpath: handle a path of "/"
- commit 5d27c03

-------------------------------------------------------------------
Thu Apr 12 14:43:34 CEST 2012 - okir@suse.de

- More work on ModemManager stuff
- commit 68e9dbc

-------------------------------------------------------------------
Thu Apr 12 14:37:59 CEST 2012 - okir@suse.de

- ni_dbus_object_find_descendant_by_handle: make a formerly private helper function public
- commit e818588

-------------------------------------------------------------------
Thu Apr 12 14:34:03 CEST 2012 - okir@suse.de

- ni_dbus_generic_property_get_string: return NOT_PRESENT if the string is NULL
- commit 726eee6

-------------------------------------------------------------------
Thu Apr 12 11:16:08 CEST 2012 - okir@suse.de

- Add modem support to wicked
- commit 5e163c9

-------------------------------------------------------------------
Wed Apr 11 22:41:37 CEST 2012 - okir@suse.de

- Get rid of obsolete dbus_class->init_child method
- commit bc0087d

-------------------------------------------------------------------
Wed Apr 11 22:40:00 CEST 2012 - okir@suse.de

- Change how we create netif objects on the client side in GetManagedObjects
  We do this in two pieces. When we instantiate an object as a child of a
  list object (such as Wicked/Interfaces), we automatically assign the
  default list item class to the new child.
  In a second step, we check if the new child has an initialize member
  function, and if it has, we use that to create a local netdev object
  and assign that to the proxy object.
- commit 71d9bf0

-------------------------------------------------------------------
Wed Apr 11 11:00:46 CEST 2012 - okir@suse.de

- Change the naming service API from being netdev specific to generic dbus objects
- commit 2516d08

-------------------------------------------------------------------
Tue Apr 10 12:50:57 CEST 2012 - mt@suse.com

- Use LIBS and LDFLAGS variables to link binaries
- commit 77a2678

-------------------------------------------------------------------
Tue Apr 10 11:46:02 CEST 2012 - mt@suse.com

- Added testing make target using pattern rules
- commit 38a8574

-------------------------------------------------------------------
Tue Apr 10 11:19:58 CEST 2012 - okir@suse.de

- Make modem objects refcounted
- commit f9ca0cc

-------------------------------------------------------------------
Thu Apr  5 16:27:16 CEST 2012 - okir@suse.de

- More ModemManager changes
- commit f64d0cd

-------------------------------------------------------------------
Thu Apr  5 16:06:44 CEST 2012 - okir@suse.de

- Introduce ni_dbus_object_get_relative_path()
- commit cabf719

-------------------------------------------------------------------
Thu Apr  5 12:34:34 CEST 2012 - okir@suse.de

- Fix make depend
- commit 5b6704d

-------------------------------------------------------------------
Thu Apr  5 12:33:10 CEST 2012 - okir@suse.de

- Got rid of dbus-client.h, which was mostly empty anyway
- commit 491131f

-------------------------------------------------------------------
Thu Apr  5 11:21:23 CEST 2012 - okir@suse.de

- Next steps for modem manager support
- commit d96275e

-------------------------------------------------------------------
Thu Apr  5 10:15:20 CEST 2012 - okir@suse.de

- dbus: support generic bool properties
- commit 384fb78

-------------------------------------------------------------------
Thu Apr  5 10:10:26 CEST 2012 - okir@suse.de

- Add ni_dbus_object_refresh_properties which uses Properties.GetAll to refresh an object's properties
- commit fc2554e

-------------------------------------------------------------------
Wed Apr  4 16:24:00 CEST 2012 - okir@suse.de

- wickedd: initialize the modem manager client
- commit 4f69c21

-------------------------------------------------------------------
Wed Apr  4 16:22:56 CEST 2012 - okir@suse.de

- dbus: support array of object paths in a dbus message
- commit f3b089a

-------------------------------------------------------------------
Wed Apr  4 12:25:02 CEST 2012 - okir@suse.de

- Add modem-manager: WIP
- commit ab4ffb7

-------------------------------------------------------------------
Fri Mar 30 14:55:17 CEST 2012 - mt@suse.com

- Readed form feeds to COPYING file
- commit 6125679

-------------------------------------------------------------------
Fri Mar 30 14:28:35 CEST 2012 - mt@suse.com

- Require dbus and nl (pkgconfig) in devel package
- commit 4f1db8b

-------------------------------------------------------------------
Fri Mar 30 14:27:35 CEST 2012 - mt@suse.com

- Remove config.h in distclean, not in clean
- commit cda4c45

-------------------------------------------------------------------
Fri Mar 30 13:57:44 CEST 2012 - mt@suse.com

- Updated FSF address in GPL references
- commit fe067fb

-------------------------------------------------------------------
Wed Mar 28 19:50:54 CEST 2012 - mt@suse.com

- Fixed several rpm lint warnings
- commit ef09535

-------------------------------------------------------------------
Wed Mar 28 19:50:19 CEST 2012 - mt@suse.com

- Use compile flags provided by autoconf (rpm)
- commit 19164b3

-------------------------------------------------------------------
Wed Mar 28 19:49:06 CEST 2012 - mt@suse.com

- Create /var/run/wicked before we're starting
- commit 6052535

-------------------------------------------------------------------
Wed Mar 28 19:48:30 CEST 2012 - mt@suse.com

- Fixed some syntax typos - hope correctly
- commit e537910

-------------------------------------------------------------------
Wed Mar 28 19:11:51 CEST 2012 - mt@suse.com

- Added more autoconf variables to Makefile.vars.in
- commit 99aaf23

-------------------------------------------------------------------
Wed Mar 28 19:10:16 CEST 2012 - mt@suse.com

- Include config.h in sources when HAVE_CONFIG_H is defined
- commit c4c76e6

-------------------------------------------------------------------
Mon Mar 26 17:41:43 CEST 2012 - okir@suse.de

- ibft: we were reading some sysfs values twice
- commit 382d652

-------------------------------------------------------------------
Sun Mar 25 19:20:44 CEST 2012 - okir@suse.de

- Change the minimum autoconf version required to 2.63
  My workstation is running SLED 11 not openSUSE Factory :)
- commit 05ca350

-------------------------------------------------------------------
Fri Mar 23 18:22:16 CET 2012 - mt@suse.com

- Updated .gitignore
- commit 39293c0

-------------------------------------------------------------------
Fri Mar 23 18:21:27 CET 2012 - mt@suse.com

- Configure on demand and include Makefile.vars
- commit d405658

-------------------------------------------------------------------
Fri Mar 23 18:20:54 CET 2012 - mt@suse.com

- Renamed Makefile.in back to Makefile
- commit 5e252f5

-------------------------------------------------------------------
Fri Mar 23 18:19:59 CET 2012 - mt@suse.com

- Generate Makefile.vars instead of Makefile
- commit 34e5613

-------------------------------------------------------------------
Fri Mar 23 18:19:14 CET 2012 - mt@suse.com

- Added Makefile.vars.in that will be generated
- commit 36d61ca

-------------------------------------------------------------------
Fri Mar 23 14:49:24 CET 2012 - mt@suse.com

- Cleaned up forgotten & commented out stuff in spec
- commit 391050d

-------------------------------------------------------------------
Fri Mar 23 14:33:33 CET 2012 - mt@suse.com

- Added autoconf changes to ChangeLog
- commit 02512d0

-------------------------------------------------------------------
Fri Mar 23 14:30:49 CET 2012 - mt@suse.com

- Adopted .gitignore for autoconf
- commit 80f3a1c

-------------------------------------------------------------------
Fri Mar 23 13:56:01 CET 2012 - mt@suse.com

- Adopted Makefile.in to use config.h, fixed dist target
- commit 0b72a86

-------------------------------------------------------------------
Fri Mar 23 13:54:47 CET 2012 - mt@suse.com

- Generate config.h, wicked has appconfig.h now
- commit d4b2f06

-------------------------------------------------------------------
Fri Mar 23 13:39:49 CET 2012 - mt@suse.com

- Make -Werror happy on unused variables
- commit 1057ec4

-------------------------------------------------------------------
Fri Mar 23 13:22:38 CET 2012 - mt@suse.com

- configure.ac cleaup with version from VERSION file
- commit f72142e

-------------------------------------------------------------------
Fri Mar 23 13:21:40 CET 2012 - mt@suse.com

- Adopted for wicked.spec.in for autoconf, improved
- commit 97ee6ed

-------------------------------------------------------------------
Fri Mar 23 13:20:49 CET 2012 - mt@suse.com

- Use autoconf variables in Makefile.in, dist target
- commit aa48ea2

-------------------------------------------------------------------
Fri Mar 23 13:19:21 CET 2012 - mt@suse.com

- Added autogen.sh helper for autoreconf && configure <defaults>
- commit 00438d0

-------------------------------------------------------------------
Fri Mar 23 13:18:52 CET 2012 - mt@suse.com

- Added VERSION file that contains the version
- commit 45e0c93

-------------------------------------------------------------------
Thu Mar 22 17:33:46 CET 2012 - okir@suse.de

- Have fun, Marius :-)
- commit f5aaed2

-------------------------------------------------------------------
Thu Mar 22 17:25:25 CET 2012 - okir@suse.de

- Adding ChangeLog file
- commit 48db82e

-------------------------------------------------------------------
Thu Mar 22 17:18:12 CET 2012 - okir@suse.de

- move src/config.h to src/appconfig.h
- commit 37df490

-------------------------------------------------------------------
Thu Mar 22 17:13:38 CET 2012 - okir@suse.de

- Remove obsolete vlan extension script
- commit eb4b9d0

-------------------------------------------------------------------
Thu Mar 22 17:10:58 CET 2012 - okir@suse.de

- ni_dbus_xml_map_method_argument: only return skip_call=TRUE if we *really* want to skip the call
- commit 09e3c27

-------------------------------------------------------------------
Thu Mar 22 17:06:01 CET 2012 - okir@suse.de

- Fix a segfault when bringing up network interfaces that have no device config
  eg. we would crash if we tried to bring up an ethernet interface without
  having an <ether> element in your config. *blush*
- commit ab74b23

-------------------------------------------------------------------
Thu Mar 22 16:43:09 CET 2012 - okir@suse.de

- Add workaround for an issue with exiting subprocesses
  Sometimes, when running extensions scripts, we will see a POLLHUP on
  the socketpair even though the subprocess hasn't exited yet. As a
  work around, we do a blocking waitpid when we detect this.
- commit 0641cd9

-------------------------------------------------------------------
Thu Mar 22 16:19:41 CET 2012 - okir@suse.de

- ifup: some tracing improvements
- commit 0bc0ea3

-------------------------------------------------------------------
Thu Mar 22 16:10:08 CET 2012 - okir@suse.de

- objectmodel: treat all virtual interfaces alike in deleteDevice
  We now call ni_dbus_object_free() in all delete methods
- commit a987f81

-------------------------------------------------------------------
Thu Mar 22 16:06:43 CET 2012 - okir@suse.de

- Bugfix: in __ni_rtevent_dellink, we were clearing dev->next uncoditionally
  This made all devices following it in the list disappear
- commit 95b812a

-------------------------------------------------------------------
Thu Mar 22 16:04:55 CET 2012 - okir@suse.de

- ifup: When a parent device waits for a failed child device, fail the parent, too
- commit fa36d38

-------------------------------------------------------------------
Thu Mar 22 16:01:51 CET 2012 - okir@suse.de

- ifup/down: correctly count marked interfaces if more than one i/f is given on the command line
- commit fced6c4

-------------------------------------------------------------------
Thu Mar 22 15:22:02 CET 2012 - okir@suse.de

- ifup: fix sorting of ifworkers in ni_ifworkers_flatten
  Our compare function was buggy.
- commit 87e0e42

-------------------------------------------------------------------
Thu Mar 22 12:43:30 CET 2012 - okir@suse.de

- Renamed s/build_hierarchy/ni_ifworkers_&/
- commit 8eadf8f

-------------------------------------------------------------------
Thu Mar 22 12:42:51 CET 2012 - okir@suse.de

- ifup/ifdown: Allow to specify more than one device on the command line
- commit 089cb32

-------------------------------------------------------------------
Thu Mar 22 12:37:41 CET 2012 - okir@suse.de

- Fix the bridge creation case
- commit 478eb1b

-------------------------------------------------------------------
Thu Mar 22 12:01:03 CET 2012 - okir@suse.de

- ifup: get the debug logging straight
  All logging now uses a new log facility named "application"
- commit 4b7b1ea

-------------------------------------------------------------------
Thu Mar 22 11:54:33 CET 2012 - okir@suse.de

- ifup: unify <require> processing
- commit e6e774e

-------------------------------------------------------------------
Thu Mar 22 10:23:43 CET 2012 - okir@suse.de

- ifup: bind methods as early as possible
- commit a239e83

-------------------------------------------------------------------
Thu Mar 22 09:22:56 CET 2012 - okir@suse.de

- More ifup changes: separate binding and execution of FSM steps
- commit 5d17d70

-------------------------------------------------------------------
Thu Mar 22 09:22:40 CET 2012 - okir@suse.de

- objectmodel: fix the Ethernet.changeDevice method definition
- commit eca3636

-------------------------------------------------------------------
Thu Mar 22 08:21:21 CET 2012 - okir@suse.de

- ifup: some trace/debug cleanups
- commit 882acf4

-------------------------------------------------------------------
Thu Mar 22 08:04:42 CET 2012 - okir@suse.de

- ifup: fix reachability checking
- commit 04717df

-------------------------------------------------------------------
Wed Mar 21 19:15:40 CET 2012 - mt@suse.com

- Added wicked.pc generation to configure.ac
- commit d7d2f79

-------------------------------------------------------------------
Wed Mar 21 18:53:02 CET 2012 - okir@suse.de

- ifup: simplify the code in ni_ifworkers_refresh_state a little
- commit 99e67cf

-------------------------------------------------------------------
Wed Mar 21 18:43:26 CET 2012 - okir@suse.de

- Fix require handling in ifup/ifdown
- commit b4a1962

-------------------------------------------------------------------
Wed Mar 21 18:35:12 CET 2012 - mt@suse.com

- Added wicked.pc.in for pkg-config support
- commit a79f29a

-------------------------------------------------------------------
Wed Mar 21 18:14:48 CET 2012 - mt@suse.com

- Added autoconf configure.ac
- commit 09969e7

-------------------------------------------------------------------
Wed Mar 21 18:07:53 CET 2012 - mt@suse.com

- Renamed wicked.spec to wicked.spec.in
- commit 84a68fe

-------------------------------------------------------------------
Wed Mar 21 18:07:30 CET 2012 - mt@suse.com

- Renamed Makefile to Makefile.in
- commit c551f6b

-------------------------------------------------------------------
Wed Mar 21 12:58:29 CET 2012 - okir@suse.de

- ifup: support early prompting for xml elements
- commit 8947a7a

-------------------------------------------------------------------
Wed Mar 21 12:25:13 CET 2012 - okir@suse.de

- ifup: remove dead code
- commit 5702d22

-------------------------------------------------------------------
Wed Mar 21 12:02:07 CET 2012 - okir@suse.de

- More work on ifup/ifdown device ordering
- commit 84705f9

-------------------------------------------------------------------
Tue Mar 20 23:20:03 CET 2012 - okir@suse.de

- Fix an overly cautious check in ni_ifworker_bind_device_factory_api
- commit 1bff81f

-------------------------------------------------------------------
Tue Mar 20 23:12:25 CET 2012 - okir@suse.de

- ifup/down: record netif-child-state requires for a parent/child relation
- commit e255240

-------------------------------------------------------------------
Tue Mar 20 22:54:33 CET 2012 - okir@suse.de

- Enrich the ifworker children array
- commit e006dc0

-------------------------------------------------------------------
Tue Mar 20 22:44:40 CET 2012 - okir@suse.de

- More work on getting the wicked device ID'ing right
- commit d3c2ab0

-------------------------------------------------------------------
Tue Mar 20 22:43:55 CET 2012 - okir@suse.de

- xml-schema: fix two infinite loops introduced by fixing xml_node_reparent
- commit 2bcec87

-------------------------------------------------------------------
Tue Mar 20 22:43:18 CET 2012 - okir@suse.de

- Clear the node's next pointer in xml_node_reparent()
- commit 7c6fa0f

-------------------------------------------------------------------
Tue Mar 20 20:30:59 CET 2012 - okir@suse.de

- Work on handling requires better
- commit 6904616

-------------------------------------------------------------------
Tue Mar 20 12:45:36 CET 2012 - okir@suse.de

- dbus-xml: support <meta:mapping> shorthand of metadata elements
- commit 745e8c5

-------------------------------------------------------------------
Tue Mar 20 12:31:36 CET 2012 - okir@suse.de

- Introduce xml_node_reparent
- commit c70ddfa

-------------------------------------------------------------------
Tue Mar 20 12:29:54 CET 2012 - okir@suse.de

- ni_string_dup: fix the order of strdup/free
  After all, the string we assign to the variable may be a substring of
  the currently assigned string. Example:
  	if (!strncmp(var, "prefix:", 6))
  		ni_string_dup(&var, var + 6);
- commit 51d07cb

-------------------------------------------------------------------
Tue Mar 20 12:18:39 CET 2012 - okir@suse.de

- ifup: remove dead code
- commit 1210f08

-------------------------------------------------------------------
Tue Mar 20 10:35:02 CET 2012 - okir@suse.de

- ifup/ifdown would always return an error.
- commit cccebf4

-------------------------------------------------------------------
Tue Mar 20 10:28:49 CET 2012 - okir@suse.de

- ifup: we were not printing a success message when we were waiting for dhcp
- commit 81736cf

-------------------------------------------------------------------
Tue Mar 20 10:17:52 CET 2012 - okir@suse.de

- ifup: device creation now relies on the meta:mapping data to locate the right factory service
- commit 25446a1

-------------------------------------------------------------------
Tue Mar 20 10:17:01 CET 2012 - okir@suse.de

- Bugfix:ni_dbus_xml_map_method_argument was disregarding the index argument
- commit cad83fa

-------------------------------------------------------------------
Tue Mar 20 10:16:45 CET 2012 - okir@suse.de

- Introduce ni_objectmodel_compatible_services_for_class
- commit 6a8565c

-------------------------------------------------------------------
Tue Mar 20 10:16:04 CET 2012 - okir@suse.de

- Fix up the name of the Bonding service - schema and C code were using different names
- commit 1f42aa0

-------------------------------------------------------------------
Tue Mar 20 09:50:32 CET 2012 - okir@suse.de

- ifup/down: refactor the finite state machine, step 5
- commit c76245a

-------------------------------------------------------------------
Tue Mar 20 09:25:39 CET 2012 - okir@suse.de

- ifup/down: refactor the finite state machine, step 4
- commit 4a9edda

-------------------------------------------------------------------
Tue Mar 20 09:07:01 CET 2012 - okir@suse.de

- ifup/down: refactor the finite state machine, step 3
- commit 0e99214

-------------------------------------------------------------------
Tue Mar 20 09:06:29 CET 2012 - okir@suse.de

- dbus-xml: fix up several functions to not abort if a method has no schema definition
- commit c05354e

-------------------------------------------------------------------
Tue Mar 20 09:06:04 CET 2012 - okir@suse.de

- Introduce ni_dbus_object_get_all_services_for_method()
- commit 969f7f6

-------------------------------------------------------------------
Tue Mar 20 09:05:17 CET 2012 - okir@suse.de

- Disable ibft addrconf service for now
- commit a1ad6f7

-------------------------------------------------------------------
Tue Mar 20 09:04:56 CET 2012 - okir@suse.de

- shut up all xpath debug messages
- commit 04662be

-------------------------------------------------------------------
Tue Mar 20 09:04:24 CET 2012 - okir@suse.de

- Map two more generic dbus errors to a useful wicked error code
- commit e5308c2

-------------------------------------------------------------------
Tue Mar 20 08:09:10 CET 2012 - okir@suse.de

- ni_dbus_connection_open: print a proper error message when name is already registered
- commit bb652d2

-------------------------------------------------------------------
Mon Mar 19 17:07:01 CET 2012 - okir@suse.de

- ifup/down: refactor the finite state machine, step 2
- commit 4bb2835

-------------------------------------------------------------------
Mon Mar 19 17:06:23 CET 2012 - okir@suse.de

- dbus xml: support xml document validation, and <meta> tags in schema files
- commit d6a8e4d

-------------------------------------------------------------------
Mon Mar 19 17:05:57 CET 2012 - okir@suse.de

- Add NI_ERROR_DOCUMENT_ERROR error code
- commit 04b1569

-------------------------------------------------------------------
Mon Mar 19 11:22:38 CET 2012 - okir@suse.de

- ifup/down: refactor the finite state machine, step 1
- commit fd5512e

-------------------------------------------------------------------
Mon Mar 19 10:55:21 CET 2012 - mt@suse.com

- GPL from http://www.gnu.org/licenses/gpl-2.0.txt
  Updated to use actualized COPYING file (new address).
- commit 9110684

-------------------------------------------------------------------
Mon Mar 19 08:10:48 CET 2012 - okir@suse.de

- Use some new <meta> tags in openvpn schema
- commit bb5bb1a

-------------------------------------------------------------------
Mon Mar 19 01:05:48 CET 2012 - okir@suse.de

- schema: introduce <meta> node that allows us to attach application specific metadata to a schema node
  The idea here is to store information inside the schema that tells us how to prompt for
  certain values (user names, passwords), or how to resolve network device name references.
  For now, this is only allowed inside scalar types.
- commit 8775308

-------------------------------------------------------------------
Mon Mar 19 01:05:24 CET 2012 - okir@suse.de

- Introduce xml_node_detach for detaching xml nodes from a tree
- commit 8e184c3

-------------------------------------------------------------------
Mon Mar 19 00:51:06 CET 2012 - okir@suse.de

- schema: when building a type with constraints, make sure we always clone the underlying type
  Move xs_type cloning into a helper function
- commit 48042f2

-------------------------------------------------------------------
Fri Mar 16 17:18:56 CET 2012 - okir@suse.de

- schema: start to enhance schema with new require/conflict constraints
- commit 0fe3d11

-------------------------------------------------------------------
Thu Mar 15 09:44:04 CET 2012 - okir@suse.de

- xml schema: add constraints that require/conflict certain elements
- commit 679b1a2

-------------------------------------------------------------------
Thu Mar 15 09:13:54 CET 2012 - okir@suse.de

- xml schema: enable ni_xs_type_built_constraints to deal with multipl constraints
- commit ab92296

-------------------------------------------------------------------
Thu Mar 15 09:10:24 CET 2012 - okir@suse.de

- xml schema: refactor the constraint building code, take 2
- commit 61b783c

-------------------------------------------------------------------
Thu Mar 15 09:09:48 CET 2012 - okir@suse.de

- xml schema: refactor the constraint building code
- commit 038cca6

-------------------------------------------------------------------
Thu Mar 15 09:02:42 CET 2012 - okir@suse.de

- xml schema: implement range constraint
- commit 235f1b6

-------------------------------------------------------------------
Thu Mar 15 08:38:48 CET 2012 - okir@suse.de

- schema parser: unify handling of enum and bitmap constraints
  At the same time, make sure we clone these constraints when cloning a type.
- commit de7a95d

-------------------------------------------------------------------
Thu Mar 15 08:11:30 CET 2012 - okir@suse.de

- Fix type bug in bonding schema
- commit d7e28d6

-------------------------------------------------------------------
Thu Mar 15 07:45:37 CET 2012 - okir@suse.de

- xml schema: disallow anonymous (type-only) elements inside a <dict>
- commit 56d5b88

-------------------------------------------------------------------
Wed Mar 14 18:19:21 CET 2012 - okir@suse.de

- Update TODO
- commit 468129d

-------------------------------------------------------------------
Wed Mar 14 17:00:20 CET 2012 - okir@suse.de

- ifup: support explicit dependencies for interfaces
   You can specify explicit dependencies for an interface using
   <dependencies>
     <require check="name-of-check" action="ifup|ifdown" state="link-up">parameter(s)</require>
   </dependencies>
   The "action" attribute specifies whether this rule applies to ifup or ifdown.
   The "state" attribute marks this as a requirement for entering the indicated
   state (device-up, firewall-up, network-up, ...). If omitted, it defaults to
   link-up.
   The "check" attribute specifies the name of the built-in test to use.
   Currently implemented checks:
   reachable: the parameter is taken as a hostname, which must be resolvable
              and for which we have a route.
- commit 99ba8c2

-------------------------------------------------------------------
Wed Mar 14 16:48:46 CET 2012 - okir@suse.de

- Rewrite the async hostname resolution using getaddrinfo_a
- commit edeabe1

-------------------------------------------------------------------
Wed Mar 14 11:05:17 CET 2012 - okir@suse.de

- Revert "ifup: when linkUp returns an unreachable/unresolvable error, wait until things change"
  This reverts commit 81d482442cd20b32c74d5ac6655509d7ed8812eb.
- commit e5c3e3a

-------------------------------------------------------------------
Wed Mar 14 11:05:05 CET 2012 - okir@suse.de

- Revert "openvpn extension: make sure VPN gateway is reachable before trying to bring up tunnel"
  This reverts commit 0ae2bcd8b24e69494a4a30e1c9ee391be6f90f7a.
- commit 2e6f4f6

-------------------------------------------------------------------
Wed Mar 14 10:38:10 CET 2012 - okir@suse.de

- openvpn extension: make sure VPN gateway is reachable before trying to bring up tunnel
- commit 0ae2bcd

-------------------------------------------------------------------
Wed Mar 14 10:36:12 CET 2012 - okir@suse.de

- ifup: when linkUp returns an unreachable/unresolvable error, wait until things change
- commit 81d4824

-------------------------------------------------------------------
Wed Mar 14 09:55:58 CET 2012 - okir@suse.de

- ifup: Change all ni_call_* methods to return an int error code rather than a bool
- commit c10c40e

-------------------------------------------------------------------
Wed Mar 14 09:53:48 CET 2012 - okir@suse.de

- Introduce new error NI_ERROR_METHOD_NOT_SUPPORTED
- commit 346330a

-------------------------------------------------------------------
Wed Mar 14 09:25:56 CET 2012 - okir@suse.de

- openvpn extension: make sure the remote gateway is reachable; if not return an error
- commit 02ef101

-------------------------------------------------------------------
Wed Mar 14 09:21:18 CET 2012 - okir@suse.de

- Simplify openvpn extension script a little
- commit ad528fe

-------------------------------------------------------------------
Wed Mar 14 08:26:22 CET 2012 - okir@suse.de

- Introduce wicked check subcommand
- commit a788660

-------------------------------------------------------------------
Tue Mar 13 19:58:59 CET 2012 - okir@suse.de

- Remove obsolete argument from wicked_get_interface()
- commit 806fc0b

-------------------------------------------------------------------
Tue Mar 13 19:55:51 CET 2012 - okir@suse.de

- Introduce error codes for unreachable addr/unresolveable host
- commit 3b54eda

-------------------------------------------------------------------
Tue Mar 13 16:16:13 CET 2012 - okir@suse.de

- Update TODO
- commit f7d0140

-------------------------------------------------------------------
Tue Mar 13 16:10:24 CET 2012 - okir@suse.de

- Fix/improve the handling of XML entities like &lt;
  When reading an xml document, processing entities should be case
  insensitive.
  When writing an xml document, make sure we escape <>& properly
- commit 357a8ae

-------------------------------------------------------------------
Tue Mar 13 16:09:50 CET 2012 - okir@suse.de

- ni_stringbufs with a static buffer were broken and were segfaulting.
- commit da2837b

-------------------------------------------------------------------
Tue Mar 13 15:42:56 CET 2012 - okir@suse.de

- bonding: clean up how we invoke __ni_bonding_get_module_option_* functions
- commit 19d8bd5

-------------------------------------------------------------------
Tue Mar 13 15:25:33 CET 2012 - okir@suse.de

- bonding: support xmit-hash-policy
- commit 4149657

-------------------------------------------------------------------
Tue Mar 13 15:24:20 CET 2012 - okir@suse.de

- objectmodel for bonding: add missing "primary" property
- commit a257e4f

-------------------------------------------------------------------
Tue Mar 13 15:04:47 CET 2012 - okir@suse.de

- bonding: get rid of old module option handling code
- commit e8a6957

-------------------------------------------------------------------
Tue Mar 13 14:58:35 CET 2012 - okir@suse.de

- Add some comments to openvpn extension script explaining how it works
- commit b874715

-------------------------------------------------------------------
Tue Mar 13 14:45:35 CET 2012 - okir@suse.de

- Update TODO: we now support openvpn :-)
- commit 8e57b50

-------------------------------------------------------------------
Tue Mar 13 14:40:26 CET 2012 - okir@suse.de

- Fix some dubious code in __ni_address_new
  In this function, we always set the broadcast address for IPv4 addresses.
  This is a little fishy, but the worst part is it break point-to-point links.
  So at least refrain from messing around when prefixlen == 32.
- commit 3ff3e91

-------------------------------------------------------------------
Tue Mar 13 14:27:17 CET 2012 - okir@suse.de

- Fix up the bool return type of a bunch of ni_address_*/ni_route_* functions
  These functions return true/false, but were declared as returning an int.
  Fix the declarations to use ni_bool_t.
- commit 604c5ce

-------------------------------------------------------------------
Tue Mar 13 14:26:52 CET 2012 - okir@suse.de

- openvpn: install an addrconf lease for the addresses received from remote
- commit 7fd9d8f

-------------------------------------------------------------------
Tue Mar 13 14:10:32 CET 2012 - okir@suse.de

- objectmodel: add com.suse.Wicked.Interface.installLease()
  This is in support of intrinsice leases
- commit e6c6572

-------------------------------------------------------------------
Tue Mar 13 14:10:02 CET 2012 - okir@suse.de

- Improve "wicked lease" support: include a state element in the lease
- commit f2a3cb3

-------------------------------------------------------------------
Tue Mar 13 13:44:33 CET 2012 - okir@suse.de

- Add client-side support for "wicked lease install".
  This lets link layers such as ppp or openvpn inform wickedd about
  the leases they have acquired. This information is useful in reporting
  the right owner.
- commit 83243cf

-------------------------------------------------------------------
Tue Mar 13 13:15:27 CET 2012 - okir@suse.de

- schema: add deviceUp/Down signals to interface schema
- commit c5a3a2b

-------------------------------------------------------------------
Tue Mar 13 13:03:15 CET 2012 - okir@suse.de

- openvpn: link-up now blocks until the openvpn tunnel is up
  Build an addrconf lease file in the callback script.
  The next step will be to inform wickedd about this lease, so that it
  can track it properly.
- commit cc68d6c

-------------------------------------------------------------------
Tue Mar 13 13:03:04 CET 2012 - okir@suse.de

- wicked client: add lease command to build an addrconf lease file (xml)
- commit fe4ad5c

-------------------------------------------------------------------
Tue Mar 13 10:53:41 CET 2012 - okir@suse.de

- Next steps on getting the openvpn extension to work
- commit 6c322bc

-------------------------------------------------------------------
Tue Mar 13 10:50:55 CET 2012 - okir@suse.de

- objectmodel extensions: do not crash if extension script returns no xml
- commit b7e08ab

-------------------------------------------------------------------
Tue Mar 13 10:50:28 CET 2012 - okir@suse.de

- ni_process_reap: we failed to handle waitpid return code 0 correctly.
- commit 2221ce7

-------------------------------------------------------------------
Mon Mar 12 20:29:48 CET 2012 - okir@suse.de

- Try to make openvpn auth handling work properly
- commit 05b8e0d

-------------------------------------------------------------------
Mon Mar 12 20:29:00 CET 2012 - okir@suse.de

- ifup: track how often we get (distinct) AuthInfoMissing errors
- commit 4c4ae1d

-------------------------------------------------------------------
Mon Mar 12 18:44:11 CET 2012 - okir@suse.de

- openvpn: fix creation and removal of openvpn config dir
- commit a56712b

-------------------------------------------------------------------
Mon Mar 12 18:43:29 CET 2012 - okir@suse.de

- Fix ni_file_remove_recursively: we didn't remove the directory itself
- commit 1510782

-------------------------------------------------------------------
Mon Mar 12 17:32:49 CET 2012 - okir@suse.de

- More work on openvpn extension
- commit 5940cfd

-------------------------------------------------------------------
Mon Mar 12 17:14:39 CET 2012 - okir@suse.de

- ifup: after creating the device, refresh our knowledge of it
- commit d4fd026

-------------------------------------------------------------------
Mon Mar 12 17:14:22 CET 2012 - okir@suse.de

- Fixed an assert failure in xpath-fmt
- commit 65502fe

-------------------------------------------------------------------
Mon Mar 12 17:14:03 CET 2012 - okir@suse.de

- Fixed two bugs in external-file representation
- commit c6e978a

-------------------------------------------------------------------
Mon Mar 12 16:36:34 CET 2012 - okir@suse.de

- Introduce helper function ni_dbus_print_error
- commit 116e2c0

-------------------------------------------------------------------
Mon Mar 12 14:58:24 CET 2012 - okir@suse.de

- Fix an inconsistency in the device model (changeDevice vs deviceChange)
  This was not used consistently, and hence didn't work for all interface
  types. Fixed this; we're now using changeDevice throughout.
- commit bbaba8a

-------------------------------------------------------------------
Mon Mar 12 14:46:00 CET 2012 - okir@suse.de

- Add OpenVPN interfaces to wicked dbus ACLs
- commit 909d355

-------------------------------------------------------------------
Mon Mar 12 14:45:36 CET 2012 - okir@suse.de

- objectmodel: actually register openvpn device and factory interfaces
- commit c960282

-------------------------------------------------------------------
Mon Mar 12 14:45:18 CET 2012 - okir@suse.de

- openvpn: objectmodel supports tunnel-id property
- commit 8cdce66

-------------------------------------------------------------------
Mon Mar 12 14:28:12 CET 2012 - okir@suse.de

- Next steps on our way to openvpn support: add the necessary objectmodel plumbing
- commit f835deb

-------------------------------------------------------------------
Mon Mar 12 12:54:46 CET 2012 - okir@suse.de

- Introduce stubs for openvpn support
- commit 8a8cb15

-------------------------------------------------------------------
Mon Mar 12 12:36:16 CET 2012 - okir@suse.de

- Add ni_file_remove_recursively()
- commit 62f1c74

-------------------------------------------------------------------
Mon Mar 12 12:17:30 CET 2012 - okir@suse.de

- Add ni_string_strip_suffix()
- commit b415303

-------------------------------------------------------------------
Mon Mar 12 10:53:03 CET 2012 - okir@suse.de

- schema: support declaration of abstract object classes
  This is needed to support "functional" extensions for network
  devices, such as openvpn or ibft.
- commit a6683ca

-------------------------------------------------------------------
Sun Mar 11 13:15:02 CET 2012 - okir@suse.de

- ni_objectmodel_device_factory_result() takes an additional dbus_class argument
  This lets the factory function override the class of the newly created device.
- commit 6061252

-------------------------------------------------------------------
Sun Mar 11 13:10:49 CET 2012 - okir@suse.de

- objectmodel: when registering a a netdev, support overriding the object class
  Also rename __ni_objectmodel_build_interface_object to __ni_objectmodel_build_netdev_object
  in the process.
- commit 36e3bba

-------------------------------------------------------------------
Fri Mar  9 17:19:05 CET 2012 - okir@suse.de

- Fix a segfault that happens with interface schemas that have a bad classname
- commit 9a66e05

-------------------------------------------------------------------
Fri Mar  9 17:11:07 CET 2012 - okir@suse.de

- Add initial support for virtual tunnel devices (and some incomplete openvpn stuff)
- commit 954d503

-------------------------------------------------------------------
Fri Mar  9 17:09:02 CET 2012 - okir@suse.de

- add netif level support for tuntap creation/deletion
  add functions ni_system_tun_create and ni_system_tun_delete
- commit 671ae43

-------------------------------------------------------------------
Fri Mar  9 17:08:11 CET 2012 - okir@suse.de

- add low-level support to create/delete tuntap devices
- commit 4e4118a

-------------------------------------------------------------------
Fri Mar  9 15:45:11 CET 2012 - okir@suse.de

- ifdown would exit silently if /etc/sysconfig/network didn't contain any config files
- commit 8bbec89

-------------------------------------------------------------------
Fri Mar  9 15:44:48 CET 2012 - okir@suse.de

- dbus config files were missing new ibft addrconf services
- commit 0031134

-------------------------------------------------------------------
Fri Mar  9 14:48:14 CET 2012 - okir@suse.de

- objectmodel extensions: when converting message data to XML, save external-file data to tempfiles
  By tracking these tempfiles through an ni_tempstate_t, we make sure we get to delete these
  files once the extension script call returns.
- commit dbc7b31

-------------------------------------------------------------------
Fri Mar  9 13:44:17 CET 2012 - okir@suse.de

- objectmodel: use the process tempstate to track tempfiles when calling extension scripts
- commit 4ec380d

-------------------------------------------------------------------
Fri Mar  9 13:42:12 CET 2012 - okir@suse.de

- Add an ni_tempstate_t handle to each process, and clean it up when we destroy the process
- commit 08d3bf6

-------------------------------------------------------------------
Fri Mar  9 13:41:39 CET 2012 - okir@suse.de

- Add ni_tempstate_t for tracking (and cleanup) of resources like tempfiles
- commit c17b693

-------------------------------------------------------------------
Fri Mar  9 11:47:17 CET 2012 - okir@suse.de

- Change how we find the dbus services for a given netif.
  Rather than requiring the tag name to be a device type, we try to
  be a little more generic - use the tag to look up a service, using
  the tag name given in the schema; and use the new functions to look
  up the corresponding factory and auth services, if there are any.
- commit 24e79b5

-------------------------------------------------------------------
Fri Mar  9 11:13:08 CET 2012 - okir@suse.de

- objectmodel: introduce new functions for looking up services
  ni_objectmodel_service_by_tag(const char *tag);
  	Look up a service given an XML tag from the config file.
  	The tag must correspond to the <service name="..."> tag
  	in the schema file.
  ni_objectmodel_factory_service(const ni_dbus_service_t *);
  	Given a dbus service, look up the corresponding factory service.
  	By default, this will be $servicename.Factory, but you can
  	override this in the schema file by specifying the
  	factory="..." attribute in the service definition.
  ni_objectmodel_auth_service(const ni_dbus_service_t *);
  	Given a dbus service, look up the corresponding auth service.
  	By default, this will be $servicename.Auth, but you can
  	override this in the schema file by specifying the
  	auth="..." attribute in the service definition.
- commit 3cfe5e2

-------------------------------------------------------------------
Fri Mar  9 09:29:25 CET 2012 - okir@suse.de

- dbus-xml: introduce a notation "external-file" which sources the contents of a file
  Also define a global type "external-file" in the xml schema which uses this.
  This lets you use things in your schema like
   <key type="external-file"/>
  and have configuration files like
   <key>/etc/openvpn/my-key.pem</key>
  This will automatically read the key file and send that to the wicked
  server, rather than transferring the file path.
- commit a2b82a5

-------------------------------------------------------------------
Thu Mar  8 20:47:39 CET 2012 - okir@suse.de

- Add new function ni_file_read
- commit 70a9618

-------------------------------------------------------------------
Thu Mar  8 17:56:24 CET 2012 - okir@suse.de

- dbus-xml: change array notations to support sizes bigger than 128
- commit 9bfaf0f

-------------------------------------------------------------------
Thu Mar  8 12:16:22 CET 2012 - okir@suse.de

- objectmodel: simplify registration of addrconf services
- commit 2e16bf7

-------------------------------------------------------------------
Thu Mar  8 12:10:00 CET 2012 - okir@suse.de

- Add some comments on ibft to the TODO file
- commit 7a7c42c

-------------------------------------------------------------------
Thu Mar  8 08:55:40 CET 2012 - okir@suse.de

- Add deviceUp/Down signals to interface schema
- commit 8d75e62

-------------------------------------------------------------------
Wed Mar  7 15:59:15 CET 2012 - okir@suse.de

- Introduce ifup --skip-active
  This will cause wicked to not touch interfaces that are up already.
  Helps during boot, especially with interfaces configured in the initrd.
- commit ad11228

-------------------------------------------------------------------
Wed Mar  7 13:54:46 CET 2012 - okir@suse.de

- Introduce ibft addrconf service
- commit 297a727

-------------------------------------------------------------------
Wed Mar  7 13:19:30 CET 2012 - okir@suse.de

- Add a naming service that lets you identify ibft interfaces
  This allows interface descriptions like this:
   <interface>
   <identify>
     <ibft>ethernet0</ibft>
   </identify>
   <ipv4:ibft/>
   </interface>
  Note, ibft addrconf is not yet implemented.
- commit 81f9c0f

-------------------------------------------------------------------
Wed Mar  7 13:13:14 CET 2012 - okir@suse.de

- Introduce ibft support to netconfig_t and netdev_t
  In ni_global_state_handle(), do the ibft discovery (from sysfs).
  During interface state refresh, when we find a new network interface,
  check if there's an iBFT entry for it, and if so, attach it to the
  netdev_t as dev->ibft_nic.
- commit 6364341

-------------------------------------------------------------------
Wed Mar  7 12:59:53 CET 2012 - okir@suse.de

- Move ni_global_state_handle() from state.c to netinfo.c
- commit 32df7ae

-------------------------------------------------------------------
Wed Mar  7 12:50:14 CET 2012 - okir@suse.de

- Make ni_netconfig_t totally private, and fix the remaining users.
  Introduce several helper/accessor functions in the process
- commit 099884d

-------------------------------------------------------------------
Wed Mar  7 12:25:47 CET 2012 - okir@suse.de

- Makefile: fix make depend for shared lib case
- commit 1214c64

-------------------------------------------------------------------
Tue Mar  6 21:05:57 CET 2012 - mt@suse.com

- Updated library list for testing/*-test
- commit 441ba9b

-------------------------------------------------------------------
Tue Mar  6 20:50:01 CET 2012 - mt@suse.com

- Added testing/ibft-test.c printing ibft nic data
- commit 1c462ed

-------------------------------------------------------------------
Tue Mar  6 20:43:30 CET 2012 - mt@suse.com

- Added ni_sysfs_ibft_scan_nics function
- commit e0f8688

-------------------------------------------------------------------
Tue Mar  6 20:40:32 CET 2012 - mt@suse.com

- Added iBFT NIC structure and functions
- commit 188439c

-------------------------------------------------------------------
Tue Mar  6 20:04:37 CET 2012 - okir@suse.de

- dhcp: discard leases when device goes down (not when network goes down)
- Introduce device-up/down events when the entire device is taken up or down
- Clean up addrconf signal handling a bit
- Fix up some error returns in netdev object model
- Move linkstats into separate header file, and add support for ethtool stats
- Update TODO file
- {autoip4,dhcp4}-supplicant: disable wirelss scans on network interfaces.
  It's enough if wickedd does this.
- Introduce ni_wireless_set_scanning, which enables/disables AP scanning globally
- make clean: remove binaries in testing/
- commit 63fa38c

-------------------------------------------------------------------
Mon Mar  5 10:05:08 CET 2012 - okir@suse.de

- Update TODO
- commit b6334f7

-------------------------------------------------------------------
Mon Mar  5 10:04:46 CET 2012 - okir@suse.de

- Enable code for writing resolv.conf
- commit 139498d

-------------------------------------------------------------------
Mon Mar  5 09:58:46 CET 2012 - okir@suse.de

- Fixed a bug in ni_process_run_and_wait: don't use WNOHANG if you want to wait :-)
- commit 34418fe

-------------------------------------------------------------------
Mon Mar  5 09:35:55 CET 2012 - okir@suse.de

- Fix ni_link_address_length to return ETH_ALEN for wireless interfaces
- commit 35f7b3b

-------------------------------------------------------------------
Mon Mar  5 09:34:00 CET 2012 - okir@suse.de

- When forwarding an addrconf drop request, silently ignore AddrconfNoLease errors
  This makes sure we can shut down an interface that was brought up partially only.
- commit a2ee62d

-------------------------------------------------------------------
Mon Mar  5 09:32:42 CET 2012 - okir@suse.de

- When trying to drop a non-existant release, return AddrconfNoLease error.
  Otherwise, wickedd would wait forever for a LeaseDropped event that will not come.
- commit 4ac5bf1

-------------------------------------------------------------------
Mon Mar  5 09:32:08 CET 2012 - okir@suse.de

- Introduce ni_dbus_set_error_from_code()
- commit 63d6590

-------------------------------------------------------------------
Mon Mar  5 09:17:35 CET 2012 - okir@suse.de

- dhcp4-supplicant: refuse to send leaseAcquired event if the caller did not
  provide a lease pointer.
- commit 034fa94

-------------------------------------------------------------------
Mon Mar  5 09:14:01 CET 2012 - okir@suse.de

- Introduce NI_ERROR_ADDRCONF_NO_LEASE error
- commit 3f8cf85

-------------------------------------------------------------------
Mon Mar  5 07:54:33 CET 2012 - okir@suse.de

- Move from static to shared libwicked
- commit 8fa40c1

-------------------------------------------------------------------
Sun Mar  4 10:32:11 CET 2012 - okir@suse.de

- Work on fixing the addrconf update mechanism
- commit 0efed25

-------------------------------------------------------------------
Sat Mar  3 10:14:15 CET 2012 - okir@suse.de

- wireless: fix the <key-management> property
- commit 59d6080

-------------------------------------------------------------------
Sat Mar  3 10:11:29 CET 2012 - okir@suse.de

- schema: add definition for com.suse.Wicked.Addrconf.ipv4.auto
- commit 09bcfa3

-------------------------------------------------------------------
Sat Mar  3 10:07:08 CET 2012 - okir@suse.de

- Shut up one false error message in ni_dbus_serialize_xml
  When we do a GetManagedObject call, we receive a list of all interfaces
  registered for the object. Not all of these may actually contain properties,
  but we always tried to retrieve the "properties" type for each interface.
  If there was none defined, we would complain.
  We now check whether the dict returned by the server contains any
  properties at all - and if it doesn't, we return quietly and do not
  print any errors at all.
- commit d1695cd

-------------------------------------------------------------------
Sat Mar  3 09:58:43 CET 2012 - okir@suse.de

- schema: support kernel level route attibutes, such as scope, type, table
- commit fcfafdf

-------------------------------------------------------------------
Sat Mar  3 08:40:45 CET 2012 - okir@suse.de

- Add naming functions for route type/scope/table etc
- commit aec811b

-------------------------------------------------------------------
Sat Mar  3 08:40:21 CET 2012 - okir@suse.de

- Introduce ni_format_int_maybe_mapped
- commit 2e40823

-------------------------------------------------------------------
Fri Mar  2 12:31:39 CET 2012 - okir@suse.de

- Move definition of ni_route_t into new header file wicked/route.h
- commit 644e07f

-------------------------------------------------------------------
Fri Mar  2 12:26:42 CET 2012 - okir@suse.de

- Support more route attributes in ni_route_t
- commit b158153

-------------------------------------------------------------------
Fri Mar  2 09:43:20 CET 2012 - okir@suse.de

- Renamed lots of local netdev_t variables from ifp to dev
- commit a470d3e

-------------------------------------------------------------------
Fri Mar  2 09:42:52 CET 2012 - okir@suse.de

- wpa-supplicant.c: renamed lots of variables from ifp -> wpa_dev
- commit 2ddf6a8

-------------------------------------------------------------------
Fri Mar  2 09:21:29 CET 2012 - okir@suse.de

- Rename lots of ni_interface_* functions to ni_netdev_*
- commit f9f85b0

-------------------------------------------------------------------
Fri Mar  2 09:12:33 CET 2012 - okir@suse.de

- Remove obsolete ni_interface_array
- commit 7ff312b

-------------------------------------------------------------------
Fri Mar  2 09:07:05 CET 2012 - okir@suse.de

- Rename ni_interface_request_t -> ni_netdev_req_t
  Rename associated ctor/dtor functions as well
- commit 8fa02bf

-------------------------------------------------------------------
Fri Mar  2 09:04:03 CET 2012 - okir@suse.de

- Rename ni_interface_t -> ni_netdev_t
- commit b6f4d50

-------------------------------------------------------------------
Fri Mar  2 08:57:09 CET 2012 - okir@suse.de

- Remove some obsolete functions from netinfo.h
- commit e10edf6

-------------------------------------------------------------------
Thu Mar  1 22:17:51 CET 2012 - okir@suse.de

- schema: fix definition of bonding properties
- commit fa57c29

-------------------------------------------------------------------
Thu Mar  1 22:16:11 CET 2012 - okir@suse.de

- Makefile: remove one test related target
- commit af77e31

-------------------------------------------------------------------
Thu Mar  1 22:12:48 CET 2012 - okir@suse.de

- Remove obsolete outdir
- commit fc53599

-------------------------------------------------------------------
Thu Mar  1 22:11:32 CET 2012 - okir@suse.de

- Get rid of the silly old linkage.c hack
- commit 600a8bd

-------------------------------------------------------------------
Thu Mar  1 22:01:20 CET 2012 - okir@suse.de

- Update TODO file
- ifdown: fix the case where we didn't provide an ifconfig file on the cmdline
- Fix the wireless ifup case - when bringing up a device, wew checked for a linkChange method, but we should really look for deviceChange
- As we now load the config files after discovering all interfaces, we need to make sure that the interface config gets attached to the right ifworker.
- wicked: support loading network config from /etc/sysconfig/network
- wicked: made global opt_* variables globally visible
  Also renamed them to opt_global_* in the process
- wicked: fix some silly code in main() wrt handling optind.
- wicked: get rid of global --timeout option
- ifup/down: make the timeout configurable via cmdline option
- Move WICKED_DBUS_* macros out of dbus.h into objectmodel.h
  Rename them to NI_OBJECTMODEL_* in the process.
- Fix year in copyright headers
- Get rid of ifaction stuff and ifebehavior
- In wicked ifup, introduce support for <boot-label> element
- Small spec file update
- Fix make install to install all applications
- Cleanup: binaries are placed in bin/, not in top-level dir anymore
- Moved mkconst to new subdir util
- Cleanup: move all *.c files out of toplevel directory
- Remove some more obsolete files from source tree
- Moved config.xml from etc/wicked/config.xml to etc/config.xml in source tree
- Implement dynamically loaded naming services
- Introduce debug facility "objectmodel"
- config.xml: refactor how we process <extension> elements
- Update TODO
- Use new identifyDevice() to identify an interface in the config file
  Instead of specifying the interface name via a <name> element, you can
  also use the new <identify> element:
   <identify>
    <ethernet:permanent-address>00:24:E8:CD:34:9B</ethernet:permanent-address>
   </identify>
- Implement interface naming services
- Introduce InterfaceList.identifyDevice() call
- Introduce ni_dbus_dict_get_entry
  This function lets you iterate over a dict as an array of entries
- Rename some functions related to dbus message processing
  Two functions that are used to iterate over a dict in a dbus message
  had somewhat unfortunate names. Changed them to ni_dbus_message_*dict*
- Introduce permanent-address to ethernet devs, and retrieve it
- Remove obsolete dbus config file
- Small cleanups of bridge extension script
- commit 1a0d18c

-------------------------------------------------------------------
Thu Mar  1 17:28:16 CET 2012 - mt@suse.com

- Added ni_stringbuf_vprintf and grow functions
- commit fc94e29

-------------------------------------------------------------------
Thu Mar  1 17:12:30 CET 2012 - mt@suse.com

- Ignore .*project files (eclipse)
- commit d03fae6

-------------------------------------------------------------------
Mon Feb 27 18:21:30 CET 2012 - okir@suse.de

- Bridge services: use ni_bridge_port_new instead of calloc()
- commit a1d0203

-------------------------------------------------------------------
Mon Feb 27 18:19:43 CET 2012 - okir@suse.de

- Convert vlan factory to use ni_objectmodel_get_netif_argument() and simplify code
- commit 519bf31

-------------------------------------------------------------------
Mon Feb 27 18:17:38 CET 2012 - okir@suse.de

- Get rid of some dead code
- commit e7f7d8d

-------------------------------------------------------------------
Mon Feb 27 18:11:50 CET 2012 - okir@suse.de

- Update TODO file
- commit e92ba69

-------------------------------------------------------------------
Mon Feb 27 18:10:18 CET 2012 - okir@suse.de

- Rename remaining __wicked_dbus_* functions to __ni_objectmodel_*
- commit a5e3726

-------------------------------------------------------------------
Mon Feb 27 18:08:02 CET 2012 - okir@suse.de

- Introduce ni_objectmodel_get_netif_argument
  This is a useful helper argument to build a ni_interface_t from the properties given
  by the caller.
- commit 8b60fcf

-------------------------------------------------------------------
Mon Feb 27 18:00:08 CET 2012 - okir@suse.de

- Make all netif factory functions use ni_dbus_error_invalid_args instead of open-coding the same
- commit 1f08e72

-------------------------------------------------------------------
Mon Feb 27 17:57:06 CET 2012 - okir@suse.de

- Introduce ni_objectmodel_device_factory_result
  This function is shared by all device factory functions. It registers the
  new device with the dbus server, and returns its object path.
- commit 91d6281

-------------------------------------------------------------------
Mon Feb 27 17:44:07 CET 2012 - okir@suse.de

- Some cleanup in the bridge dbus interface code
- commit a098f5f

-------------------------------------------------------------------
Mon Feb 27 16:17:03 CET 2012 - okir@suse.de

- When serializing an xml node as an array of dbus dicts, we used to
  encode it as an array of variants, with each variant containing a dict
  (ie signature "av").
  Change this to actually encode it as an array of dicts, ie signature "aa{sv}".
- commit 2926760

-------------------------------------------------------------------
Mon Feb 27 16:16:18 CET 2012 - okir@suse.de

- In the Bridge/Bonding.newDevice implementation, we forgot to set
  the device name when the user requested an anonymous device.
- commit f227ac2

-------------------------------------------------------------------
Mon Feb 27 16:15:08 CET 2012 - okir@suse.de

- In __ni_objectmodel_bridge_device_arg, we created a bonding device, not a
  bridge device. Bad bad bad :)
- commit 46b3f8c

-------------------------------------------------------------------
Mon Feb 27 16:14:15 CET 2012 - okir@suse.de

- In Bridge.newDevice, don't try to handle duplicate create of a bridge
  device. It's useless, and it crashes :-)
- commit 27749c0

-------------------------------------------------------------------
Mon Feb 27 15:04:15 CET 2012 - okir@suse.de

- In __ni_dbus_object_set_properties_from_dict, clear the dbus_error (as we ignore it)
- commit 8f62306

-------------------------------------------------------------------
Mon Feb 27 09:45:27 CET 2012 - okir@suse.de

- Move NI_BRIDGE_VALUE_NOT_SET to bridge.h
- commit 75ccc6d

-------------------------------------------------------------------
Mon Feb 27 09:43:18 CET 2012 - okir@suse.de

- ni_bridge_add_port: take ownership of new port object
- commit 2f90662

-------------------------------------------------------------------
Mon Feb 27 09:36:06 CET 2012 - okir@suse.de

- Get rid of bridge_port.device member
- commit e057cfb

-------------------------------------------------------------------
Mon Feb 27 09:24:12 CET 2012 - okir@suse.de

- Get rid of some old bridge helper code
- commit 7505e46

-------------------------------------------------------------------
Mon Feb 27 09:19:40 CET 2012 - okir@suse.de

- Change how we treat bridge ports - primary identifier is the ifindex, not the name
- commit 9919926

-------------------------------------------------------------------
Mon Feb 27 08:34:24 CET 2012 - okir@suse.de

- Work on bridge create/setup/delete methods
- commit d681b59

-------------------------------------------------------------------
Sat Feb 25 13:56:08 CET 2012 - okir@suse.de

- Support changing ethtool attributes on an ethernet device
- commit a87b62b

-------------------------------------------------------------------
Sat Feb 25 13:19:37 CET 2012 - okir@suse.de

- Fixed Interface.linkDown() schema definition - this function doesn't take any args
- commit 1293d4d

-------------------------------------------------------------------
Sat Feb 25 13:15:21 CET 2012 - okir@suse.de

- ifup/down: Be less rigid with state dependencies
  For instance, if  we want to configure a vlan device, we can *create*
  the device as soon as the eth device is available.  And we should do
  this, even is the link is not there. Lack of link on the eth device
  should only prevent us from bringing up the vlan link
- commit 7094094

-------------------------------------------------------------------
Sat Feb 25 13:14:46 CET 2012 - okir@suse.de

- Introduced ni_uint_range_t and some functions to update it
- commit 9bc681c

-------------------------------------------------------------------
Sat Feb 25 13:07:41 CET 2012 - okir@suse.de

- Fixed the dbus call signatures for newDevice methods (vlan, bonding)
- commit 1c7c49b

-------------------------------------------------------------------
Sat Feb 25 12:28:46 CET 2012 - okir@suse.de

- Make the list of ifworker actions a copy, not just a reference
- commit f384130

-------------------------------------------------------------------
Sat Feb 25 12:13:43 CET 2012 - okir@suse.de

- Add note to TODO file
- commit aa64079

-------------------------------------------------------------------
Sat Feb 25 12:10:26 CET 2012 - okir@suse.de

- xml_node_get_child: don't segfault if the caller gives us a NULL node
- commit 9025124

-------------------------------------------------------------------
Sat Feb 25 12:09:53 CET 2012 - mt@suse.com

- Marked unfinished code with FIXME (satisfy -Werror)
- commit 0da32f0

-------------------------------------------------------------------
Sat Feb 25 12:08:54 CET 2012 - mt@suse.com

- Removed unused variables (satisfy -Werror)
- commit 5260222

-------------------------------------------------------------------
Sat Feb 25 12:07:20 CET 2012 - mt@suse.com

- Fixed debug watch verbose ifdefs
- commit 66fb303

-------------------------------------------------------------------
Sat Feb 25 12:04:21 CET 2012 - okir@suse.de

- Fix a segfault in ni_ifworker_find_link_properties: we crashed if the interface node was empty
- commit b97a76e

-------------------------------------------------------------------
Sat Feb 25 12:01:50 CET 2012 - okir@suse.de

- Extended ethernet schema: added offload info to properties
- commit 54f5463

-------------------------------------------------------------------
Sat Feb 25 11:45:42 CET 2012 - okir@suse.de

- xml names of bonding modes now use different names; the kernel ones aren't comptible with xml
- commit 0d238b4

-------------------------------------------------------------------
Sat Feb 25 11:40:22 CET 2012 - mt@suse.com

- Fixed ni_dbus_object_set_properties_from_dict return value
- commit e1110ff

-------------------------------------------------------------------
Sat Feb 25 11:38:51 CET 2012 - okir@suse.de

- For services with builtin properties or methods, register the services
  explicitly in ni_objectmodel_register_netif_classes rather than doing it
  in the config file.
- commit 1ccbd19

-------------------------------------------------------------------
Sat Feb 25 11:23:34 CET 2012 - okir@suse.de

- Fix the method names in config.xml: newLink -> newDevice, same for deleteDevice
- commit 5f6beab

-------------------------------------------------------------------
Sat Feb 25 11:21:33 CET 2012 - okir@suse.de

- Fix for bonding schema: we used different type names for the device configuration
- commit b0be067

-------------------------------------------------------------------
Sat Feb 25 11:17:03 CET 2012 - okir@suse.de

- Implement the changeDevice() method for bonding interfaces
- commit d2ed674

-------------------------------------------------------------------
Sat Feb 25 11:16:37 CET 2012 - okir@suse.de

- Add ni_system_bond_setup() function to (re)configure bonding device
- commit bb6c308

-------------------------------------------------------------------
Sat Feb 25 07:55:14 CET 2012 - okir@suse.de

- Introduce ni_dbus_error_invalid_args
- commit 10b3e9f

-------------------------------------------------------------------
Sat Feb 25 07:42:15 CET 2012 - okir@suse.de

- Register bonding services
- commit db658ee

-------------------------------------------------------------------
Sat Feb 25 05:55:56 CET 2012 - okir@suse.de

- Introduce ni_interface_make_name and use it
- commit 8952429

-------------------------------------------------------------------
Sat Feb 25 05:49:53 CET 2012 - okir@suse.de

- Add newDevice method to Bonding service
- commit 893281c

-------------------------------------------------------------------
Sat Feb 25 05:30:21 CET 2012 - okir@suse.de

- Rename linkChange method to deviceChange
- commit 8ed0a33

-------------------------------------------------------------------
Sat Feb 25 05:20:54 CET 2012 - okir@suse.de

- Change the device-up code to always call linkChange method if it exists
- commit bc9dd25

-------------------------------------------------------------------
Sat Feb 25 05:19:43 CET 2012 - okir@suse.de

- Add ni_string_strip_prefix()
- commit 90dd218

-------------------------------------------------------------------
Sat Feb 25 05:03:54 CET 2012 - okir@suse.de

- Renamed newLink method to newDevice
- commit 19d39f9

-------------------------------------------------------------------
Fri Feb 24 13:39:28 CET 2012 - okir@suse.de

- Introduce ni_dbus_error_property_not_present
- commit 953f455

-------------------------------------------------------------------
Fri Feb 24 13:30:56 CET 2012 - okir@suse.de

- Remove another obsolete member from ni_bonding
- commit 32b0282

-------------------------------------------------------------------
Fri Feb 24 13:29:37 CET 2012 - okir@suse.de

- fix bonding schema and properties handling
- commit 4c874e1

-------------------------------------------------------------------
Thu Feb 23 15:21:01 CET 2012 - okir@suse.de

- Remove some obsolete bonding struct members
- commit 23d3060

-------------------------------------------------------------------
Thu Feb 23 15:20:04 CET 2012 - okir@suse.de

- bonding schema: add enums for bonding mode and arp-validate
- commit c0ef8ab

-------------------------------------------------------------------
Thu Feb 23 14:59:30 CET 2012 - okir@suse.de

- README update
- commit 56788e2

-------------------------------------------------------------------
Sat Feb 18 14:06:38 CET 2012 - okir@suse.de

- Remove wicked-convert; we need to do this differently
- commit d242824

-------------------------------------------------------------------
Sat Feb 18 11:18:12 CET 2012 - okir@suse.de

- Remove lots of dead code
- commit 01acd5f

-------------------------------------------------------------------
Sat Feb 18 11:00:40 CET 2012 - okir@suse.de

- Moved some code out of wicked.c into client/*.c files
- commit f9fb2f2

-------------------------------------------------------------------
Sat Feb 18 10:53:32 CET 2012 - okir@suse.de

- wicked: remove some old code
- commit 90a3b27

-------------------------------------------------------------------
Sat Feb 18 10:51:13 CET 2012 - okir@suse.de

- ifup: make ni_call_link_method() more generic - check for the number of arguments to decide whether to encode a dict argument or not.
- commit a26fc9e

-------------------------------------------------------------------
Sat Feb 18 10:50:33 CET 2012 - okir@suse.de

- Provide functions to query whether a method takes any arguments/returns any data
- commit 4bbc9c5

-------------------------------------------------------------------
Sat Feb 18 10:43:34 CET 2012 - okir@suse.de

- Change ni_dbus_serialize_return to return -1, 0, or 1 rather than a boolean. We need to know this in case the method doesn't return anything at all.
- commit 2336800

-------------------------------------------------------------------
Sat Feb 18 10:35:14 CET 2012 - okir@suse.de

- Move extension scripts into /etc/wicked
- commit af5dd2d

-------------------------------------------------------------------
Sat Feb 18 10:27:08 CET 2012 - okir@suse.de

- Add firewall up/down
- commit 7aac6c3

-------------------------------------------------------------------
Sat Feb 18 10:25:28 CET 2012 - okir@suse.de

- config.xml: corrected a mistake in extension script shell vars: $property:ifindex -> $property:index
- commit 8bc1a6c

-------------------------------------------------------------------
Sat Feb 18 10:22:44 CET 2012 - okir@suse.de

- config.xml: remove __properties from com.suse.Wicked.Wireless
- commit ed0da0b

-------------------------------------------------------------------
Sat Feb 18 09:43:40 CET 2012 - okir@suse.de

- dbus config file was missing com.suse.Wicked.Wireless.Auth
- commit 3f56763

-------------------------------------------------------------------
Sat Feb 18 09:41:18 CET 2012 - okir@suse.de

- Add two more sample files
- commit d4546d2

-------------------------------------------------------------------
Thu Feb 16 18:47:53 CET 2012 - okir@suse.de

- Adapt README
- commit e57e09a

-------------------------------------------------------------------
Thu Feb 16 18:46:25 CET 2012 - okir@suse.de

- Remove obsolete netcf sample files
- commit 607d3dc

-------------------------------------------------------------------
Thu Feb 16 18:45:48 CET 2012 - okir@suse.de

- Remove some obsolete code
- commit 92c83b0

-------------------------------------------------------------------
Thu Feb 16 18:03:09 CET 2012 - okir@suse.de

- Fixed compile error on 64bit platforms
- commit a6da2c5

-------------------------------------------------------------------
Thu Feb 16 18:00:00 CET 2012 - okir@suse.de

- ni_ifworker_error_handler should actually return negative value (ie -RETRY_OPERATION).
- Translate more wicked errors to dbus errors
- Introduce machinery for wickedd to request additional auth info
- Add NI_ERROR_RETRY_OPERATION error
- Do some work on mapping dbus errors to wicked error codes
- Add sample config file for wlan interface
- Redid the wireless scan scheduling
- Move wireless_scan from device to device->wireless
- disable another useless debug message
- wireless: fix how we detect loss of association
- wpa-supplicant: fix how we encode/decode the signal level
- We were sending too many wireless scan requests to wpa-supplicant
- Debug improvements in ni_dbus_object_call and ni_dbus_call_simple
- Simplified a lot of code involving ni_wpa_interface_bind()
- Clean up some dead code
- Add wireless method setScanning()
- Some more fixes on how we retrieve wireless scan resultsW
- dbus-common: remove two stupid debug prints
- wpa_supplicant: when we get the initial state, propagate it to the master object
- Fix the entire wireless association code once more. - transmit essid/security info in linkChange call - do the actual association in linkUp
- Make all ni_interface_get_{bridge,bond,wireless,...} functions check for the device type, and return NULL if it doesn't match.
- Now tracking interface oper_state
- ni_call_link_method_xml: print an error if we cannot find the requested method
- Moved a debug message hen processing ifinfo messages from the kernel
- ifup: introduce link-auth support
- Introduce ni_wireless_disconnect
- ni_wireless_association_changed: use ifindex rather than ifname to identify the device
- wpa-supplicant: use ifindex internally to identify device
- More work on propagating wireless events
- Track wireless association
- Make __ni_interface_event available to internal users of the library
- Add new events NI_EVENT_LINK_ASSOCIATED/ASSOCIATION_LOST for wireless
- Managed to get the first successful wireless association with wpa-psk
- Some minor cleanups in ifup client code
- Introduce ni_wireless_associate and use it in the Wireless dbus service's linkChange routine
- wpa-supplicant: handle non-existant Network properties gracefully by returning a PropertyNotPresent error
- Introduce PropertyNotPresent dbus error and check for it in ni_dbus_object_properies_as_dict()
- wpa-supplicant: when processing the return value of an addInterface call, we actually need to use ni_dbus_variant_get_object_path, not _get_string
- ni_dbus_object_call_variant: if the caller doesn't specify an interface name, and we couldn't identify the interface based on the attached services, fall back to default interface
- Use uniform names for the dbus_services used in wpa-supplicant.c
- Publish ni_wpa_interface_associate
- ni_dbus_object_call_variant: don't segfault if object has no interfaces configured
- ni_dbus_object_call_simple: translate dbus error to error code when call returns an error
- Introduce helper function ni_wpa_client()
- Move ni_call_link_layer*service from wicked.c -> client/call.c
- Rename wicked_link_layer*service -> ni_call_link_layer*service
- Introduce ni_interface_get_wireless and use it in the wireless dbus service
- ifup: add linkChange support
- Next steps in adding wireless support
- Draft the code for configuring a wireless interface
- Got rid of ni_wpa_scan
- Allow to mix scanned networks and configured networks as children of a wpa_device
- Add proper .class information to wpa dbus interfaces
- Move some members of ni_wireless_network into scan_info struct
- Renamed wireless_network.auth_info to *.supported_auth_modes
- Indentation fixes in wireless.h
- Remove obsolete struct ni_wpa_interface_capabilities type
- s/ni_wpa_network_t/ni_wireless_network_t/g
- Fix up ni_wpa_network_properties_destroy properly
- Unify wireless network handling: make ni_wireless_network_t and ni_wpa_network_t the same type, and share objects between the two modules
- Introduce refcounting to ni_wireless_network_t
- Change the handling of wpa network scan results. We no longer keep a list of scanned networks inside the wpa_interface; instead we rely on the hierarchy of dbus objects to retrieve them. Also add a destroy function to the wpanet class so that destruction of the dbus object destroys the backing wpa_network object as well.
- ni_wpa_network: get rid of proxy pointer
- ni_wpa_network: stop using proxy pointer
- Fixed a bug introduced by the removal of wpa_bss_properties
- Publish ni_dbus_object_lookup
- Change wpa_network.last_seen to .expires
- Aligned some more variable names in wpa_network/wireless_network
- Aligned some variable names in wpa_network/wireless_network
- Get rid of wireless_network.bitrates array
- Introduce ni_wireless_ssid_t and use it in ni_wireless_network_t
- Get rid of ni_wpa_bss_properties
- wpa-supplicant: renamed a lot of variables and function for *bss* to *network*
- Renamed ni_wpa_bss_t -> ni_wpa_network_t
- Get rid of useless scan member in ni_wpa_bss_t
- wpa-supplicant: support more network properties
- Add more wireless properties
- Transport interface capabilities via dbus properties
- constants.xml: add a few more wireless related enums
- Change some meta names in constants.xml.in
- mkconst: Simplify the map generation code a bit
- Support some wireless related enums in the xml schema
- Add initial support for wireless objectmodel
- Add dbus_variant functions for setting/getting a double: ni_dbus_variant_{get,set}_double ni_dbus_dict_{get,add}_double
- Rename all ni_process_instance_* functions to ni_process_*
- Rename ni_process_instance_t to ni_process_t
- Rename ni_process_t to ni_shellcmd_t
- Add missing code in ni_system_updater_run
- Introduce ni_process_instance_run_and_wait
- Move the new system updater code into src/update.c
- Introduce ni_bool_t
- Makefile: clean up etc/mkconst.o as well
- Next steps with resolver/hostname updates
- In ni_objectmodel_addrconf_signal_handler, call the resolver updater before installing the addresses (sorry)
- When passing a lease via dbus, encode/decode the resolver information
- dhcp4-supplicant: print the list of DO_* flags when in debug mode
- Add first steps of updater framework
- Define new type builtin-ifflags and export the right constants to the schema
- Add xml schema for infiniband interfaces
- Define a changeLink method for Ethernet service
- Introduce infiniband-child and detect it in iflist.c
- Update config.xml
- config.xml: get rid of <socket> element
- Rearrange schema files a little
- Remove obsolete type from schema
- Update dbus config files
- Implement missing ethernet address property
- Allow device identification by <alias> element
- Add ifalias attributes to interface and ifreq properties
- dbus-message: do not crash when serializing NULL string
- Handle interface alias names (IFLA_IFALIAS)
- dhcp4-supplicant: if the network goes down on an interface, discard leases
- More client side fixes for addrconf event handling.
- When forwarding an addrconf drop release to an external service, the release event may actually arrive while we're processing the forwarded call. Thus, we have to re-check the lease after the call returns.
- Implement dropLease for dhcp4 and ipv4ll
- Fixing addrconf up case with dhcp
- Fixing ifdown --delete
- Rename wicked_link_change_* to ni_call_link_change_*
- More work on device_delete code in client
- client: move ni_call_create_client to calls.c, and use new ni_objectmodel_init() function
- Use ni_objectmodel_init in wickedd
- Introduce ni_objectmodel_init
- Rearrange some code in the wicked client
- First steps toward ifdown --delete
- ifdown: taking down an interface should not require a config file
- Remove two more obsolete functions
- Remove two obsolete function prototypes
- ni_system_interface_link_change no longer takes a netconfig_t as its first argument
- Moving code to new file client/calls.c
- client: rename some functions
- ifdown: implement the link-down step
- objectmodel: implement linkDown method
- ni_system_interface_link_change: allow function to be called with NULL interface_request, signalling link-down
- Get rid of obsolete ni_addrconf_request_t. Bye bye.
- dhcp4: make sure that the "update" field passed to us is retained while we try to acquire the lease, and sent back to the daemon when we have it.
- Include an "update" field in addrconf_lease_t
- Fix compile breakage due to changes in afinfo_t way back. Somehow .depend is broken :-(
- When requesting a DHCP lease, install a temporary lease with state "requesting". Exception: there already is an active lease on this interface; in which case we're not changing anything.
- In the addrconf forwarder, make sure we delete the temporary dbus object after use
- Shut down one particulary annoying debug message in ni_dbus_object_get_properties_as_dict
- Try to improve __ni_interface_update_addrs/routes: when we come across a route or address that's not managed by wicked, it's possible that the only reason we don't know about it is because the daemon has been restarted. In this case, we want to "adopt" the route or address.
- Implement network_down in the wicked client
- objectmodel addrconf: implement set part of all lease properties
- objectmodel: all addrconf services now return the lease state, if there is a lease
- xml schema: introduce new enum type "builtin-addrconf-state", which corresponds to the NI_ADDRCONF_STATE enums
- Fix a bug introduced by the recent change in ni_system_lease_update: When receiving static addrconf, we need to set the lease state to "GRANTED" or ni_system_lease_update will ignore the addresses.
- ni_objectmodel_addrconf_ipv6_static_service used the wrong service name
- Fix the schema
- Add ni_call_drop_lease
- Introduce dropLease functions to static ipv4/ipv6 addrconf service
- Rename addrconf method from configure to requestLease
- Starting to get ifdown working
- objectmodel addrconf forwarder: simplify the event processing
- Change default ifup timeout to 20sec
- Small cleanup in ifconfig.c
- When calling __ni_interface_update_routes, pass in ni_route_t *, not **
- Change __ni_interface_update_routes to pass in the old_lease rather than addrfamily and config mode
- When calling __ni_interface_update_addrs, pass in ni_address_t *, not **
- Change __ni_interface_update_addrs to pass in the old_lease rather than addrfamily and config mode
- Renamed some variables in __ni_interface_update_{addrs,routes}
- Get rid of ni_route.config_method
- Get rid of ni_address.config_method
- Add config_lease field to ni_route_t
- iflist: improve tracking of ipv6 autoconf addresses
- Add config_lease field to all addresses assigned to an interface
- Introduce named lease owner
- Get rid of afinfo.request[] array
- Changed how leases are associated with an ni_interface_t. PReviously, we used an array of lease_t's in ni_afinfo; now we're just using an open list directly in ni_interface_t
- Rename ni_ifworker_name -> ni_ifworker_state_name
- ifup: implement timeout of interface operations
- More debug message improvements
- ifup: Fix a bug in the linkUp code
- Fix two typos in schema
- ifup: change link_up code to work with recent linkUp changes on the server side
- ni_dbus_object_get_service_for_{method,signal}: when dealing with overloaded methods/signals, return the most specific service rather than the first one we come across.
- Change how we take the link up/down. Replace a single linkChange method with linkUp/linkDown. If the link change was immediate, these methods return nothing. If we have to wait for the link to come up, these functions will return a callback_info with uuid/event name.
- dhcp4-supplicant: seed the RNG, so that we get truly random timeouts
- Fix vlan-dhcp example to set a shorter acquire timeout
- ifup: misc debug enhancements
- When forwarding an addrconf to dhcp, pass along the uuid, and fix dhcp4-supplicant to record it, too :)
- Fix a stack memory overwrite in dhcp4 supplicant
- Remove two obsolete fields from ni_dhcp4_request
- dhcp4-supplicant: send request uuid if we have it
- improve one error msg in addrconf
- ifup: debug mode - show events we're waiting for
- Removed some obsolete code
- Moved objectmodel callback_info classes into misc.c
- Make addrconf methods return callback info and record that in the client, so we can later match the events we receive and see if what we expected to happen is actually taking place.
- addrconf methods for dhcp and ipv4ll now return callback_info
- Introduce __ni_objectmodel_event_to_signal()
- Get rid of event_id in ni_addrconf_request. addrconf methods now return the uuid directly, rather than an integer event_id
- Introduce ni_dbus_message_append_uuid
- Change last argument of ni_objectmodel_interface_event from integer event_id to uuid
- Change ni_interface_get_addrconf_request to look up requests by uuid
- Change the addrconf_request type to no longer contain mode/family to identify the addrconf service, but an "owner" string
- addrconf forwarder: put all supplicant related information into a sub-struct of ni_addrconf_forwarder
- dhcp4-supplicant acquire method now accepts an optional req uuid
- Change ni_objectmodel_unwrap_interface() to take an additional (optional) DBusError argument, and adjust all code that calls it.
- ni_objectmodel_addrconf_signal_handler: process optional uuid argument
- Introduce NI_UUID_INIT
- In the addrconf signal handler, when we find that we've acquired or dropped a lease and there's a corresponding addrconf request with an event_id, broadcast that event_id.
- Added ni_interface_get_addrconf_request
- Add an event_id argument to ni_objectmodel_interface_event
- Pass the addrconf forwarder info to ni_objectmodel_addrconf_signal_handler
- Simplify the objectmodel addrconf forwarders
- Introduce ni_interface_set_addrconf_request()
- Move uuid generation into ni_uuid_generate
- Drop some old cruft from ni_addrconf_request
- Get rid of a lot of the old addrconf mechanics
- Fix ni_objectmodel_addrconf_signal_handler to correctly handle ipv4ll leases, too
- Get rid of some of the old addrconf cruft in objectmodel
- autoip4-supplicant: get rid of all addrconf request processing for now; we're currently not using any parameters anyway
- Fix "make depend" for autoip4/*.c files
- Get rid of old ipv4ll addrconf forwarder code
- Get rid of old dhcp4 addrconf forwarder code
- Introduce addrconf forwarder for ipv4 zeroconf
- Removed dhcp and statik members from ni_addrconf_request_t. We no longer need to transport that stuff through the generic request struct; instead we forward the client's request unchanged.
- wicked: remove two obsolete functions
- Get rid of ni_dhcp_lease_matches_request()
- Change the dhcp4 supplicant to use its own dhcp4_request_t rather than the generic ni_addrconf_request_t, which is scheduled to go away.
- sample vlan-dhcp: show some uses of dhcp options
- wicked schema: sometimes we were mixing dict member names and type names. Pry them apart.
- xml-schema: avoid segfaults when using <nodename type="sometype"> where both nodename and sometype are valid type names.
- Change the dbus static addrconf services to use leases rather than addrconf requests
- Next steps on addrconf: support dhcp4 forwarding
- dhcp4 supplicant: only register interfaces that have ARPHRD_ETHER
- Cleanup some in dbus-objects/addrconf.c
- Set the compatible class when registering Addrconf.ipv4/ip6.static interface
- Define schema for dhcp4 addrconf request
- Removed lots of old, dead code
- Cleaned up some error reporting in wicked client
- Added new sample file for vlan with dhcp config
- Fix up schema for ipv4/6 static addrconf; add dhcp4 service
- static addrconf service doesn't propvide any properties for now
- Add sample config files
- Fixed ni_dbus_dict_get_next once more. Hopefully correct now
- Use new ni_dbus_message_append_uint32() helper to return an int32 value to the caller
- Introduce helper functions to append simple types to a reply message
- Implement IPv6 static addrconf alongside IPv4
- Implementing addconf handling in the wicked client
- Implement Addrconf.ipv4.static
- Fix up the schema - include path names were wrong - change how we represent addrconf requests; this makes for less xml clutter in the config files
- Add new function ni_system_interface_addrconf to handle built-in addrconf modes
- Fixed ni_dbus_dict_get_next(), that was completely broken :-(
- Improved error messages in ni_dbus_serialize_xml_array
- Fixed two bugs in __ni_notation_netaddr_prefix_parse
- Some ifup function renames
- ifup: get rid of is_slave field in ifworker
- ifup: move some code around
- Update .gitignore
- ifup: forgot to include new include file wicked-client.h
- Various updates to extension scripts
- Moved schema files from etc/wicked/schema to schema/ subdir
- Make sure bind all objectmodel methods *and properties* in the client as well
- Now take the link_change info from a <device> subnode rather than the <interface> element itself
- minor spelling changes in debug messages
- ifup: introduce link bring-up
- When ni_objectmodel_vlan_newlink encountered an existing object, it would remove it from the objectmodel :-(
- Try to clean up the dbus/interface_request mess a little by introducing ni_objectmodel_{,un}marshal_interface_request
- When calling ni_dbus_object_call_variant without an interface name, choose the most specific interface that provides the requested method. In case of derived types, methods attached to the most specific type win.
- Make __ni_dbus_object_set_properties_from_dict static; it's not used anywhere
- Introduce ni_dbus_class_is_subclass
- Clean up some old cruft in ifconfig.c
- Rename wicked_dbus_interface_service -> ni_objectmodel_netif_service
- More progress with ifup - but slow
- Re-introduce builtin VLAN newLink function. I preferred this over using the script approach, as it allows me to make it idempotent without having to just through hoops.
- Replace ni_interface_by_vlan_tag - this makes no sense; the only reasonable thing is to match by tag and physdev name: introduce ni_interface_by_vlan_name_and_tag and use that instead.
- ni_assert() uses abort(), include <stdlib.h> for his
- Return better error codes in ni_system_vlan_create()
- Publish functions for message marshaling/unmarshaling of variants
- ifup: more work
- ifup: more fsm work
- ifup: renamed ni_interface_state stuff to ni_ifworker
- ifup: continue to work on fsm code
- Add a clarifying comment in the GetManagedObject() code
- Reinstate some older code: we do have to refresh the global interface when a client calls GetManagedObjects on the netif-list
- Moved ni_objectmodel_netif_list_{class,service} from model.c to interface.c
- Fix some debug messages displayed during get_managed_objects
- Fixed an erroneous error thrown during get_managed_objects
- Improve and enable the build_hierarchy stuff in ifup
- Introduce xml_node_get_next_named to traverse an xml subtree in search of all nodes with a given name.
- Begin to work on next iteration of ifup FSM
- wicked: listen for server signals during ifup
- Define the signals emitted by an Interface. Change method name linkUp -> linkChange
- When building dbus service objects from the xml schema, we did not pay attention to services that have built-in methods. We just overwrote what was there. In addition, add support for signal definitions via the schema
- Fix a small issue in ni_dbus_server_send_signal - when checking for valid signals, we actually tried to look up the named signal in the method table instead of the signal table. This triggered incorrect warnings
- dbus xml schema now lets you define the signals emitted by a service
- Redo ni_objectmodel_interface_event(), and make sure we send events for all link/network up/down changes
- Yet another iteration at implementing ifup
- When doing extension calls, process the return/error XML generated by the script
- Introduce ni_dbus_xml_serialize_{return,error}
- Work on objectmodel's interface methods
- More schema refinements
- Some more schema enhancements
- Improve the xml output for interface
- Get rid of silly debug printf
- Bring properties for interface and bridge in line
- Introduce ni_dbus_dict_get_next to loop over all dict entries with the same name
- Fixed a bug in ni_sockaddr_prefix_{pack,unpack}, which resulted in incorrect on the wire representation
- wicked show-xml: introduce capability to display "cooked", i.e. schema-conformant xml. The original behavior is still available through wicked show-xml --raw
- Introduce ni_dbus_xml_deserialize_properties - bugfix: avoid re-registering dbus interfaces if they already exist. Instead, amend the existing dbus_service object.
- Bring interface properties in synch with the xml schema
- Clean up interface schema definition a little - we used two contexts with the name "interface"
- Publish ni_xs_scope_lookup_* functions
- ni_objectmodel_register_service() registers non-const service objects now. They need to be modifiable in order to register the schema information.
- Use new network-address notation in xml schema
- One more round of rearrangements for new sockaddr pack/unpack code
- Change names of existing notation print/parse functions slightly and make them more uniform
- Introduce new notations for netaddr and netaddr-prefix
- For existing dbus-xml notations, add matching print() functions
- Remove silly debug messages
- Improve output of show-xml a little
- Make bridge properties and schema conform with each other
- xml schema: for now, change all object-path elements to string, as we have no sane way to map interface names to strings.
- Fix use of dbus system vs session bus
- Remove old cruft from wicked/wicked.h
- Introduce ability to specify session instead of system bus
- Fix the dbus watch handling code. There's a strange spurious connection hangup being triggered by large messages, and it's not clear whether this is related to our handling of watches or not. At any right, it's good to fix the bugs :)
- re-enable bridge properties
- publish ni_socket_hold
- Get rid of obsolete function ni_objectmodel_interface_port_service
- re-enable ethernet properties
- ni_objectmodel_bind_extensions: do not segfault when service has no methods
- Add a hack to enforce linkage of property table(s) and such
- Remove some obsolete code
- Do proper object type checking in ni_objectmodel_get_vlan
- Move some objectmodel specific stuff into wicked/objectmodel.h
- Work on C binding for properties
- __ni_dbus_object_new_child: use init_child only in special circumstances. This is still b0rked
- First shot at allowing builtin method definition via config file
- Introduce new wicked verb: show-xml
- Fix up ni_dbus_object_call_variant: we didn't return the server's response, ever :-(
- Make sure the GetManagedObject client side code understands standard interfaces, too
- Made the netif-list class ready for client side operations, and got rid of the proxy-netif-list in the wicked client.
- More changes on the IPv4 addrconf xml schema
- In GetManagedObject client side: Fix up the class compatibility check once more
- Change where inside the <interface> element we keep the linkinfo
- Add extension scripts for vlan and bridge devices
- Minor changes to the bridge schema
- xpath-fmt: allow expansions to expand to the empty string.
- Properly handle DBUS_TYPE_DOUBLE in ni_dbus_variant_sprint
- add deleteLink method to bridge schema
- For async server calls, we used to remember the object for which this call arrived. If the object was deleted in the middle of the call, we would cancel the pending call. However, this happens regularly during the deletion of virtual interfaces such as vlans, bridges or bonds - hence we would never reply to such dbus calls.
  Change the implementation by no longer remembering the object, and ignoring any
  object unregistrations as far as async calls are concerned.
- Some objectmodel cleanup
- Implement expansion of extension environment
- Delete some old extension handling code
- Change how we handle the environment of an extension. Instead of generic xpath expressions, now use
  <extension>
     <putenv name="..." value="..."/>
   </extension>
  We want to support different meta-values such as
   value="$object-path"
   value="$property:name"
- Add ifindex property to Interface methods
- Fix deleteLink handling
- Fix client side of GetManagedObjects - type system was getting in the way here.
- ni_objectmodel_unwrap_interface: type check mustn't check for type equality, but inheritance. Use ni_dbus_object_isa() for this.
- Debug fix: changed a ni_warn() call into NI_TRACE_ENTER_ARGS
- Add simple schema for bridges
- dbus-xml: properly handle and array of dicts
- dbus-message: fix deserialization of an array of variants
- xml schema: properly implement scalar constraint="enum"
- Pass PATH variable to subprocess
- Clean up temporary files after an extension script exited
- Introduce ni_process_instance_getenv
- Increase the default timeout on a dbus call to 10 seconds
- Change the extension code to write out the argument list as XML document.
- Implement ni_dbus_xml_deserialize_arguments() to represent a variant as XML
- Implement missing function ni_dbus_variant_is_byte_array
- Introduce ni_dbus_variant_array_print_element
- ni_dbus_variant_array_parse_and_append_string: forgot to increment the array len if we managed to append the value successfully.
- More improvements to wicked create code
- Introduce ni_dbus_xml_serialize_arg to serialize a single xml argument, and use it in wicked to pass the newLink argument.
- wicked: work on the link creation code
- in the vlan schema, rename vlan.id to vlan.tag
- Add some tracing when calling extension scripts
- Fix two issues in process.c: - use a socketpair instead of a pipe to capture the output of the child process. Our mainloop i/o code only handles sockets right now - install a SIGCHLD handler
- When processing a dbus call message through an async method handler, we do not have a reply. But we tried to send that NULL reply back to the client, causing libdbus to barf.
- Fix two issues in socket code: - the code set SIGCHLD to SIG_IGN - we tried to access a closed socket after handling POLLHUP
- Introduce ni_error_extra
- Changing the wicked create code to use the new Factory services
- ni_dbus_xml_register_services no longer needs a dbus_server argument
- Introduce ni_objectmodel_register_all and call it to register all objectmodel classes and services prior to creating the server.
- ni_objectmodel_register_netif_classes() no longer needs a dbus_server argument
- Introduce ni_objectmodel_service_by_class
- ni_objectmodel_bind_compatible_interfaces() no longer needs server as its first argument
- Publish ni_objectmodel_link_classname()
- Introduce type object-path as valid scalar type in xml schema
- Minor code cleanup in dbus-objects/model.c
- Move class registry from dbus_server level to no_objectmodel and make it global
- Get rid of the notion of special link layer services in ni_objectmodel
- Moved code from dbus-objects/model.c -> interface.c
- Use ni_objectmodel_bind_compatible_interfaces when creating netif objects
- In ni_dbus_object_register_service(), fix the check for object/interface compatibility
  Instead of just checking for class equality, check superclasses as well.
- Implement ni_objectmodel_bind_compatible_interfaces()
- Introduce a superclass field in dbus_class. Create link-layer specific subclasses of netif
- Split ni_objectmodel_create_service and create ni_objectmodel_create_initial_objects
- Add a vlan-factory service definition
- Introduce new object-class attribute for xml service definition
  This attribute defines which object class this service is compatible
  with.
- Change the way how we deal with additional attrbutes of a <service> defintion.
  We used to identify the link layer using provides="link:foobar". Change that
  to link-layer="foobar" and introduce a more generic attribute handling for
  xml service definitions. Rather than parsing all these attributes immediately
  in the xml-schema code, we parse them later in dbus-xml when creating the
  dbus services
- More cleanups of ni_objectmodel stuff
- Allow registration of certain dbus classes with a dbus_server instance
- Clean up dhcp4 dbus interface a little
- Introduce NI_OBJECTMODEL_*_CLASS macros for dbus class names
- dbus: introduce the notion of interface compatibility.
  For dbus services, introduce a "compatible class" pointer, and make sure
  we never register a service with an object it is not compatible with
- Missed to fix up the autoip4 code for new dbus class interface
- Rename wicked_dbus_interface_destroy -> ni_objectmodel_netif_destroy
- Introduce ni_objectmodel_unwrap_interface to extract n_interface_t from a dbus object. Check the object's class for type safety.
- Get rid of ni_dbus_object_functions; move these functions in ni_dbus_class
- Introduce ni_dbus_class
- wickedd: use ni_objectmodel_bind_extensions() to bind all extensions
- dbus-xml: when registering all methods for a dbus service, avoid registering methods with NULL name
- wickedd: call ni_dbus_xml_register_services to tie the schema information to all services
- When parsing <service> element in a schema file, register the service in the enclosing scope, not in the subscope defined by the XML element.
- make install: Install dbus policy files
- Define debug facility for xml
- Fix dbus policy for dhcp4 service
- Next steps at implementing extension dbus services
- __ni_dbus_object_message: perform a few sanity checks on the message
- Avoid segfault in ni_dbus_object_get_service_for_method by checking for NULL method name
- Changed argument list method->async_completion() once more
- Introduce a few more process functions
- Introduce ni_file_write
- Change ni_mkstemp so it optionally returns the name of the tempfile
- dbus_server: changed first argument of async call handler from dbus_server to dbus_connection
- Introduce ni_dbus_connection_send_error()
- Removed some old ni_extension_* crap
- Change the extension script handling code to use ni_process_t's
- Introduce ni_dbus_async_server_call_run_command()
- Moved definition of ni_process_{,instance_}t from "process.h" to <wicked/types.h>
- ni_dbus_server_t: introduce server side support for processing calls asynchronously
- Export ni_process_instance_free
- More variable renames of async_client_calls
- Introduce ni_dbus_async_server_call support (partial)
- Renamed ni_dbus_pending* to ni_dbus_async_client_call*
- Rearrange the dbus server side method call code a little
- Introduce ni_objectmodel_bind_extensions
- Get rid of old warn() error() compat macros
- Wholesale fixup of obsolete error() calls to ni_error()
- Revive the extension handling to support DBUS script backends
- Introduce ni_dbus_xml_register_methods to register interfaces and methods defined in xml schema
- Move ni_objectmodel_register_*service decls into include file
- Introduce a better model for registering and looking up dbus objectmodel services
- Add NI_LAYER_NONE constant
- Some more schema enhancements
- xml schema cleanups
- Introduce ni_server_dbus_xml_schema(), used to parse the dbus xml schema
- %s/\<typedict\>/scope/g
- Add support for defining constants
- Fix refcounting issues in schema analysis
- Introduce <dict extends="base-type"> Change the syntax of bitmap constraints Improve the documentation in xml-schema.h a little
- Rollup of messed-up previous 2 commits
- Add bonding schema
- Support <include> element in schema
- Allow scoped type lookups as in vlan:linkinfo
- Introduce <service> definition into schema
- Rename ni_xs_typedict -> ni_xs_scope
- Keep typedict created in a named context
- Introduce bitmap constraint in xml schema
- Remove some dead code
- Move all xml schema *_info pointers into a union inside ni_xs_type_t.
- Introduce ni_xs_scalar_info_t
- Introduce ni_xs_{dict,struct}_info_t
- First shot at writing the dbus-xml conversion
- Introduce types and functions for handling subprocesses
- Hold a temporary reference to a socket pointer while calling all recv/send/hangup callbacks
- Change ni_string_array_append to accept NULL string pointers
- Introduce ni_buffer_ensure_tailroom. This is the first out-of-line buffer handling function, so we need a new buffer.c source file, too.
- Disable broken client-side ifup code for now
- commit b0267f4

-------------------------------------------------------------------
Thu Dec 22 17:38:05 CET 2011 - okir@suse.de

- Fix make depend target
- commit 6fbcdee

-------------------------------------------------------------------
Thu Dec 22 17:31:00 CET 2011 - okir@suse.de

- Get rid of interface.up_requesters
- commit 50825c8

-------------------------------------------------------------------
Thu Dec 22 17:29:06 CET 2011 - okir@suse.de

- Emit appropriate signals after interface shutdown
- commit 2458fe4

-------------------------------------------------------------------
Thu Dec 22 17:21:29 CET 2011 - okir@suse.de

- Support signal generation after having acquired a lease
- commit b743422

-------------------------------------------------------------------
Thu Dec 22 17:12:55 CET 2011 - okir@suse.de

- Simplified objectmodel service creation a little bit
- commit 465bee1

-------------------------------------------------------------------
Thu Dec 22 16:59:48 CET 2011 - okir@suse.de

- Working on ifup/ifdown events
- commit 57d8257

-------------------------------------------------------------------
Thu Dec 22 16:45:01 CET 2011 - okir@suse.de

- s:ni_server_send_signal:ni_dbus_server_send_signal:
- commit 69a0c24

-------------------------------------------------------------------
Thu Dec 22 16:26:46 CET 2011 - okir@suse.de

- Fix a crash in the lease handling code
- commit b78bfa9

-------------------------------------------------------------------
Thu Dec 22 16:00:47 CET 2011 - okir@suse.de

- When deserializing a route, we forgot to set config_method
- commit 1400e43

-------------------------------------------------------------------
Thu Dec 22 16:00:32 CET 2011 - okir@suse.de

- Fixed a compile error in some (usually disabled) debugging code
- commit 78398db

-------------------------------------------------------------------
Thu Dec 22 15:46:29 CET 2011 - okir@suse.de

- Add wicked-convert to .gitignore
- commit 118dd9c

-------------------------------------------------------------------
Thu Dec 22 14:47:09 CET 2011 - okir@suse.de

- Get rid of the syntax abstraction, step 3
- commit 828fa25

-------------------------------------------------------------------
Thu Dec 22 14:38:40 CET 2011 - okir@suse.de

- Get rid of the syntax abstraction, step 2
- commit 475e25b

-------------------------------------------------------------------
Thu Dec 22 14:29:29 CET 2011 - okir@suse.de

- Get rid of the syntax abstraction, step 1
- commit 4f7c825

-------------------------------------------------------------------
Thu Dec 22 13:38:15 CET 2011 - okir@suse.de

- Add two functions to netcf backend for direct access
- commit 7345c88

-------------------------------------------------------------------
Thu Dec 22 13:19:26 CET 2011 - okir@suse.de

- Added ni_netinfo_{new,free}
- commit c8c46b8

-------------------------------------------------------------------
Thu Dec 22 13:14:23 CET 2011 - okir@suse.de

- Changed ni_sysconfig_scandir to take a simplistic glob pattern rather than a match prefix
- commit 6f1cc6a

-------------------------------------------------------------------
Thu Dec 22 13:05:22 CET 2011 - okir@suse.de

- Changed ni_scandir from straightforward prefix matching to some simple globbing pattern
- commit 44183f1

-------------------------------------------------------------------
Thu Dec 22 12:51:23 CET 2011 - okir@suse.de

- Make wicked client transmit the right ifflags in ifup
- commit 7b91693

-------------------------------------------------------------------
Thu Dec 22 12:46:08 CET 2011 - okir@suse.de

- wicked ifup understands just netcf files going forward
- commit 8867e9d

-------------------------------------------------------------------
Thu Dec 22 12:27:39 CET 2011 - okir@suse.de

- Put the sysconfig->xml conversion into a separate utility program
- commit 1f4c3e1

-------------------------------------------------------------------
Thu Dec 22 12:25:44 CET 2011 - okir@suse.de

- Implemented missing ctor/dtor for netconfig
- commit 5860664

-------------------------------------------------------------------
Thu Dec 22 12:25:26 CET 2011 - okir@suse.de

- Added FIXME
- commit c6d9d5b

-------------------------------------------------------------------
Thu Dec 22 12:24:53 CET 2011 - okir@suse.de

- Use ni_error instead of perror() when unable to open sysconfig file
- commit a872278

-------------------------------------------------------------------
Thu Dec 22 12:24:26 CET 2011 - okir@suse.de

- Use the interface index for object path rather than the interface name
- commit 6fa0f2a

-------------------------------------------------------------------
Wed Dec 21 21:53:02 CET 2011 - okir@suse.de

- Missed a conversion earlier on when changing DBUS_PROPERTY macro
- commit 35a1299

-------------------------------------------------------------------
Wed Dec 21 17:18:26 CET 2011 - okir@suse.de

- Convert VLAN objects to use generic properties
- commit d925021

-------------------------------------------------------------------
Wed Dec 21 17:17:43 CET 2011 - okir@suse.de

- Add support for generic uint16 property
- commit 2cd8bc7

-------------------------------------------------------------------
Wed Dec 21 17:08:52 CET 2011 - okir@suse.de

- Changed Interface dbus objects to use generic properties as much as possible
- commit 4e83c70

-------------------------------------------------------------------
Wed Dec 21 16:55:38 CET 2011 - okir@suse.de

- Added support functions for generic string properties
- commit 38dcd62

-------------------------------------------------------------------
Wed Dec 21 15:00:25 CET 2011 - okir@suse.de

- Deliver LeaseLost signal if dhcp was unable to obtain a lease
- commit da7f418

-------------------------------------------------------------------
Wed Dec 21 14:52:48 CET 2011 - okir@suse.de

- __ni_dbus_object_set_properties_from_dict assumed that all dict properties
  were generic; but that is not the case.
- commit ac38291

-------------------------------------------------------------------
Tue Dec 20 10:57:58 CET 2011 - okir@suse.de

- Convert dbus Bridge interface to use generic properties
- commit a34d84c

-------------------------------------------------------------------
Tue Dec 20 10:57:23 CET 2011 - okir@suse.de

- Convert Ethernet interface to use generic attributes
- commit 9bd5142

-------------------------------------------------------------------
Tue Dec 20 10:56:43 CET 2011 - okir@suse.de

- Introduce ni_objectmodel_interface_path() helper
- commit a3c4e7a

-------------------------------------------------------------------
Tue Dec 20 10:50:47 CET 2011 - okir@suse.de

- Add dbus_variant_get/set functions for object_path types.
  Fix the dbus_dict_get_* functions to do proper type checking.
- commit 0bbf6b2

-------------------------------------------------------------------
Mon Dec 19 17:28:20 CET 2011 - okir@suse.de

- Add dbus config files
- commit 23e63c8

-------------------------------------------------------------------
Mon Dec 19 17:17:41 CET 2011 - okir@suse.de

- Introduce ni_dbus_dict_delete_entry, and use it in interface creation code to fix up the args dict
- commit ab64d8f

-------------------------------------------------------------------
Mon Dec 19 15:02:14 CET 2011 - okir@suse.de

- Improve interface creation support
- commit af7ec8c

-------------------------------------------------------------------
Mon Dec 19 11:43:18 CET 2011 - okir@suse.de

- ni_system_bond_create didn't return the newly created interface - fixed
- commit 8e13c5b

-------------------------------------------------------------------
Sun Dec 18 21:12:19 CET 2011 - okir@suse.de

- Enable ni_dbus_object_{get,set}_properties_{as,from}_dict to deal with dict properties
- commit 220008a

-------------------------------------------------------------------
Sun Dec 18 21:11:55 CET 2011 - okir@suse.de

- Minor fixlets
- commit f6dae98

-------------------------------------------------------------------
Sun Dec 18 12:34:56 CET 2011 - okir@suse.de

- More work on generic properties
- commit f667500

-------------------------------------------------------------------
Fri Dec 16 05:02:37 CET 2011 - okir@suse.de

- Reduce dependencies on dbus.h
- commit 7d7ccd9

-------------------------------------------------------------------
Fri Dec 16 04:27:21 CET 2011 - okir@suse.de

- Introduce support for generic properties
- commit e77f7b5

-------------------------------------------------------------------
Thu Dec 15 11:23:09 CET 2011 - okir@suse.de

- Get rid of obsolete .id field in dbus_property
- commit 206f7d9

-------------------------------------------------------------------
Thu Dec 15 11:16:49 CET 2011 - okir@suse.de

- Actually enable bonding support in dbus interface factory
- commit 462b973

-------------------------------------------------------------------
Thu Dec 15 11:16:34 CET 2011 - okir@suse.de

- Move NI_DBUS_DICT_SIGNATURE into dbus-common.h
- commit edfdebc

-------------------------------------------------------------------
Thu Dec 15 11:15:37 CET 2011 - okir@suse.de

- Introduce ni_bonding_new() constructor
- commit e5c3a68

-------------------------------------------------------------------
Thu Dec 15 10:16:42 CET 2011 - okir@suse.de

- Implement dbus support fr bonding devices
- commit b4fa67d

-------------------------------------------------------------------
Thu Dec 15 04:14:49 CET 2011 - okir@suse.de

- Fix a problem in wickedd - we did no enumerate all devices at startup
- commit 9370414

-------------------------------------------------------------------
Thu Dec 15 02:30:23 CET 2011 - okir@suse.de

- Move all ni_interface handling functions from netinfo.c to their own file
- commit 12b26ba

-------------------------------------------------------------------
Thu Dec 15 02:15:46 CET 2011 - okir@suse.de

- Clean out some old crap
- commit bd5c9b1

-------------------------------------------------------------------
Thu Dec 15 02:09:45 CET 2011 - okir@suse.de

- Moved name/constant mapping functions to names.c
- commit 1da293a

-------------------------------------------------------------------
Thu Dec 15 02:06:32 CET 2011 - okir@suse.de

- Moved some system related function declrs to separate header file
- commit 0af5a5a

-------------------------------------------------------------------
Thu Dec 15 02:03:08 CET 2011 - okir@suse.de

- Move vlan definitions into separate header file
- commit 478402c

-------------------------------------------------------------------
Thu Dec 15 01:58:21 CET 2011 - okir@suse.de

- Renamed a bunch of functions
- commit 59a5ac6

-------------------------------------------------------------------
Thu Dec 15 01:44:21 CET 2011 - okir@suse.de

- Clean up ifconfig.c
- commit 2b1cd8a

-------------------------------------------------------------------
Thu Dec 15 01:27:28 CET 2011 - okir@suse.de

- Get rid of ni_interface_by_name temp hack
- commit 4dc0ec6

-------------------------------------------------------------------
Thu Dec 15 01:16:03 CET 2011 - okir@suse.de

- More ni_handle cleanups
- commit e6c3bc9

-------------------------------------------------------------------
Thu Dec 15 01:12:41 CET 2011 - okir@suse.de

- Got rid of ni_handle type
- commit 9ee68d3

-------------------------------------------------------------------
Thu Dec 15 01:11:04 CET 2011 - okir@suse.de

- More ni_handle changes
- commit b4dc81a

-------------------------------------------------------------------
Wed Dec 14 12:53:57 CET 2011 - okir@suse.de

- Remove more obsolete ni_handle arguments in ifevent code
- commit cd0757b

-------------------------------------------------------------------
Wed Dec 14 12:44:40 CET 2011 - okir@suse.de

- Remove more obsolete ni_handle arguments
- commit 1e7627f

-------------------------------------------------------------------
Wed Dec 14 12:33:07 CET 2011 - okir@suse.de

- Removed some more useless ni_handle arguments from some wireless calls
- commit 6366bc2

-------------------------------------------------------------------
Wed Dec 14 12:32:16 CET 2011 - okir@suse.de

- Removed some more useless ni_handl arguments from some rtnetlink calls
- commit f1a752d

-------------------------------------------------------------------
Wed Dec 14 12:28:29 CET 2011 - okir@suse.de

- Removed netlink from ni_handle and made it a global var
- commit 5486780

-------------------------------------------------------------------
Wed Dec 14 12:25:22 CET 2011 - okir@suse.de

- Removed iocfd from ni_handle and made it a global var
- commit f156d12

-------------------------------------------------------------------
Wed Dec 14 12:13:20 CET 2011 - okir@suse.de

- Change all ni_interface_by* functions to take a netconfig rather than a handle
- commit 70a47f5

-------------------------------------------------------------------
Wed Dec 14 12:03:02 CET 2011 - okir@suse.de

- Get rid of some obsolete functions
- commit e18ca58

-------------------------------------------------------------------
Wed Dec 14 11:59:38 CET 2011 - okir@suse.de

- Get rid of temporary hack nc_interface_new
- commit da6512b

-------------------------------------------------------------------
Wed Dec 14 11:58:31 CET 2011 - okir@suse.de

- Change ni_interface_new to take a netconfig ptr rather than a handle
- commit 83a2bc9

-------------------------------------------------------------------
Wed Dec 14 11:56:11 CET 2011 - okir@suse.de

- Encapsulate ni_handle.iflist and routes in a netconfig struct
- commit ba248a1

-------------------------------------------------------------------
Wed Dec 14 11:53:12 CET 2011 - okir@suse.de

- Move the definition of __ni_netlink to kernel.h
- commit ff37199

-------------------------------------------------------------------
Wed Dec 14 11:50:25 CET 2011 - okir@suse.de

- Get rid of ni_handle.ops
- commit d7196ea

-------------------------------------------------------------------
Wed Dec 14 11:48:56 CET 2011 - okir@suse.de

- Get rid of ni_handle.seqno
- commit 42878ad

-------------------------------------------------------------------
Wed Dec 14 11:47:19 CET 2011 - okir@suse.de

- Removed some obsolete files
- commit c605a47

-------------------------------------------------------------------
Wed Dec 14 11:46:41 CET 2011 - okir@suse.de

- Get rid of a few generic ni_handle related functions
- commit 5761a41

-------------------------------------------------------------------
Wed Dec 14 11:35:02 CET 2011 - okir@suse.de

- Get rid of rest api functions
- commit c43398e

-------------------------------------------------------------------
Wed Dec 14 11:32:08 CET 2011 - okir@suse.de

- Get rid of some ni_handle opening functions
- commit 7d1848c

-------------------------------------------------------------------
Wed Dec 14 11:17:05 CET 2011 - okir@suse.de

- Get rid of ni_ops.refresh and ni_refresh
- commit 8834b43

-------------------------------------------------------------------
Wed Dec 14 10:41:34 CET 2011 - okir@suse.de

- Got rid of ni_ops.interface_refresh_one
- commit 6ea66a6

-------------------------------------------------------------------
Wed Dec 14 10:40:45 CET 2011 - okir@suse.de

- Got rid of ni_ops.delete_interface
- commit b15afd6

-------------------------------------------------------------------
Wed Dec 14 10:38:51 CET 2011 - okir@suse.de

- Got rid of ni_ops.configure_interface
- commit bfe3da8

-------------------------------------------------------------------
Wed Dec 14 10:36:13 CET 2011 - okir@suse.de

- Get rid of ni_ops.interface_stats_refresh
- commit a8abe15

-------------------------------------------------------------------
Wed Dec 14 10:34:25 CET 2011 - okir@suse.de

- Get rid of ni_ops.update_lease
- commit 066a7db

-------------------------------------------------------------------
Wed Dec 14 10:29:05 CET 2011 - okir@suse.de

- Removed some more functions from ni_ops
- commit 104ede6

-------------------------------------------------------------------
Wed Dec 14 10:05:53 CET 2011 - okir@suse.de

- Next step in ni_handle removal: moved update functions for system files out of ni_ops
- commit daf2d79

-------------------------------------------------------------------
Wed Dec 14 09:41:17 CET 2011 - okir@suse.de

- More cleanups - get rid of old netinfo handle
- commit 1e68a88

-------------------------------------------------------------------
Wed Dec 14 09:29:46 CET 2011 - okir@suse.de

- Get rid of indirect handle for now
- commit 9316714

-------------------------------------------------------------------
Wed Dec 14 09:28:25 CET 2011 - okir@suse.de

- Got rid of a bunch of dummy netconfig structs in backend-netcf.c
- commit 4e614f4

-------------------------------------------------------------------
Wed Dec 14 09:18:43 CET 2011 - okir@suse.de

- Getting rid of netconfig.seqno
- commit 7647fc4

-------------------------------------------------------------------
Wed Dec 14 09:11:52 CET 2011 - okir@suse.de

- First pass for ni_handle->ni_backend conversion
- commit 9b3dd33

-------------------------------------------------------------------
Wed Dec 14 08:14:34 CET 2011 - okir@suse.de

- Starting to rewrite backend handling
- commit d6f92ce

-------------------------------------------------------------------
Wed Dec 14 07:21:38 CET 2011 - okir@suse.de

- Remove obsolete ipv6 RA listener
- commit 60e28b0

-------------------------------------------------------------------
Wed Dec 14 06:19:09 CET 2011 - okir@suse.de

- First stab at handling RTM_NEWPREFIX messages to detect IPv6 autoconf prefixes
- commit 677c554

-------------------------------------------------------------------
Wed Dec 14 05:20:48 CET 2011 - okir@suse.de

- Renamed the last remaining file out of src/dhcp subdir
- commit 688e501

-------------------------------------------------------------------
Wed Dec 14 05:18:04 CET 2011 - okir@suse.de

- Got rid of xml_node argument in ni_addrconf_acquire_lease()
- commit f9349d3

-------------------------------------------------------------------
Wed Dec 14 05:06:19 CET 2011 - okir@suse.de

- Fix a bug when parsing the device path in a Lease event
- commit 6968fdb

-------------------------------------------------------------------
Wed Dec 14 05:05:40 CET 2011 - okir@suse.de

- In the dhcp supplicant, after releasing a lease, set its state to
  NI_ADDRCONF_STATE_RELEASED before sending the event.
- commit d141f36

-------------------------------------------------------------------
Wed Dec 14 04:59:37 CET 2011 - okir@suse.de

- Fixed definition of WICKED_DBUS_AUTO4_INTERFACE
- commit 74d3a9c

-------------------------------------------------------------------
Wed Dec 14 04:57:34 CET 2011 - okir@suse.de

- Simplified definition of ipv4ll addrconf hooks
- commit 387700d

-------------------------------------------------------------------
Wed Dec 14 04:50:57 CET 2011 - okir@suse.de

- Add some extra checks in ni_objectmodel_dhcp4_acquire
- commit 2fb12ee

-------------------------------------------------------------------
Wed Dec 14 04:47:29 CET 2011 - okir@suse.de

- Made the dhcp addrconf flow of control a little less confusing
- commit 16da459

-------------------------------------------------------------------
Wed Dec 14 04:24:20 CET 2011 - okir@suse.de

- Simplified addrconf acquire/release objectmodel interface
- commit 10189f2

-------------------------------------------------------------------
Wed Dec 14 04:05:59 CET 2011 - okir@suse.de

- Introduce ni_dbus_object_translate_error
- commit 980d255

-------------------------------------------------------------------
Wed Dec 14 03:54:46 CET 2011 - okir@suse.de

- Work on enabling IPv4ll dbus bindings
- commit f77a11a

-------------------------------------------------------------------
Wed Dec 14 03:19:01 CET 2011 - okir@suse.de

- Moved generic part of dhcp acquire/release client functionality into
  new file dbus-objects/addrconf.c
- commit 980d72d

-------------------------------------------------------------------
Wed Dec 14 02:44:42 CET 2011 - okir@suse.de

- Redo the timer handling for IPv4ll
- commit 62635b0

-------------------------------------------------------------------
Wed Dec 14 02:43:39 CET 2011 - okir@suse.de

- In ni_dhcp_device_close, make sure we always cancel the timer
- commit c4334f2

-------------------------------------------------------------------
Wed Dec 14 01:39:10 CET 2011 - okir@suse.de

- Moving IPv4ll code around
- commit d52926e

-------------------------------------------------------------------
Wed Dec 14 01:05:48 CET 2011 - okir@suse.de

- Introduce ni_capture_devinfo_init and ni_capture_devinfo_refresh and start using
  it in DHCP.
- commit 25b7350

-------------------------------------------------------------------
Tue Dec 13 17:21:49 CET 2011 - okir@suse.de

- Introduce generic ni_lease_event
- commit 5a4acca

-------------------------------------------------------------------
Tue Dec 13 12:55:51 CET 2011 - okir@suse.de

- In dbus synchronous calls, make sure we process any signals that were queued up
  while we waited for the reply.
- commit 1b1e87b

-------------------------------------------------------------------
Tue Dec 13 12:55:23 CET 2011 - okir@suse.de

- Process LeaseReleased signal emitted by DHCP4 supplicant when dropping a lease
- commit 0337912

-------------------------------------------------------------------
Tue Dec 13 12:31:50 CET 2011 - okir@suse.de

- Implement dhcp_release
- commit 74e7dcf

-------------------------------------------------------------------
Tue Dec 13 12:25:49 CET 2011 - okir@suse.de

- Add two debug statements in dbus code
- commit 102f9bc

-------------------------------------------------------------------
Tue Dec 13 12:25:18 CET 2011 - okir@suse.de

- Begin to fix dhcp4 release functionality
- commit 9e78f4d

-------------------------------------------------------------------
Tue Dec 13 12:24:38 CET 2011 - okir@suse.de

- Fix a bug in dhcp4_drop() client side function. It used the wrong object path
- commit a033d21

-------------------------------------------------------------------
Tue Dec 13 12:24:00 CET 2011 - okir@suse.de

- When shutting down an interface, first undo all the addrconf stuff,
  then take the interface down
- commit 211c48e

-------------------------------------------------------------------
Tue Dec 13 11:39:48 CET 2011 - okir@suse.de

- Implement dhcp4_device_destroy
- commit ebbfa88

-------------------------------------------------------------------
Tue Dec 13 11:34:06 CET 2011 - okir@suse.de

- Got rid of some FIXMEs in dhcp code
- commit 1cf7a51

-------------------------------------------------------------------
Tue Dec 13 11:18:00 CET 2011 - okir@suse.de

- Try to handle failure mode of quick renewal correctly
- commit a4c1077

-------------------------------------------------------------------
Tue Dec 13 11:17:31 CET 2011 - okir@suse.de

- Fixed make depend for dhcp4 stuff
- commit 680e0c9

-------------------------------------------------------------------
Tue Dec 13 10:55:41 CET 2011 - okir@suse.de

- Made some dhcp FSM functions private
- commit 457864e

-------------------------------------------------------------------
Tue Dec 13 10:52:13 CET 2011 - okir@suse.de

- dhcp: remove two obsolete functions
- commit 814b782

-------------------------------------------------------------------
Tue Dec 13 10:49:34 CET 2011 - okir@suse.de

- Make __ni_system_interface_update_lease use standard functions
  __ni_interface_update_{addrs,routes}
- commit 16a49ba

-------------------------------------------------------------------
Tue Dec 13 10:27:46 CET 2011 - okir@suse.de

- In the dhcp4 supplicant, try to handle link loss/reattach in a smart way
- commit a91863d

-------------------------------------------------------------------
Tue Dec 13 10:27:31 CET 2011 - okir@suse.de

- Fix mistake in ni_dbus_service_get_signal
- commit f33f420

-------------------------------------------------------------------
Tue Dec 13 08:35:26 CET 2011 - okir@suse.de

- Improve dhcp4 event support
- commit 1cf7be0

-------------------------------------------------------------------
Tue Dec 13 02:51:28 CET 2011 - okir@suse.de

- When passing an addrconf request around via dbus, include its uuid
- commit 94ddc13

-------------------------------------------------------------------
Tue Dec 13 02:49:27 CET 2011 - okir@suse.de

- When passing a lease around via dbus, include state and uuid
- commit 53a5451

-------------------------------------------------------------------
Tue Dec 13 02:37:51 CET 2011 - okir@suse.de

- Introduce ni_route_print
- commit 42ee088

-------------------------------------------------------------------
Tue Dec 13 02:06:38 CET 2011 - okir@suse.de

- Next steps in making DHCP4 work again
- commit b84c580

-------------------------------------------------------------------
Mon Dec 12 09:43:45 CET 2011 - okir@suse.de

- Fix a bug in representing addrconf requests introduced by previous patch
- commit bcbdae4

-------------------------------------------------------------------
Mon Dec 12 09:25:18 CET 2011 - okir@suse.de

- Removed dead files
- commit 09fe701

-------------------------------------------------------------------
Mon Dec 12 09:09:06 CET 2011 - okir@suse.de

- Work on representing addrconf leases as dbus objects
- commit cfa5052

-------------------------------------------------------------------
Mon Dec 12 09:08:24 CET 2011 - okir@suse.de

- Add missing implementation of ni_dbus_variant_is_string_array
- commit 38a0c6a

-------------------------------------------------------------------
Mon Dec 12 08:57:23 CET 2011 - okir@suse.de

- Code cleanup
- commit 11ab4c1

-------------------------------------------------------------------
Mon Dec 12 08:17:40 CET 2011 - okir@suse.de

- Removed obsolete file
- commit 19cf3ee

-------------------------------------------------------------------
Mon Dec 12 08:16:44 CET 2011 - okir@suse.de

- Properly handle creation and deletion of network interfaces in wickedd
- commit 9109241

-------------------------------------------------------------------
Mon Dec 12 07:30:50 CET 2011 - okir@suse.de

- Moved dhcp4 supplicant code into its own source directory
- commit 81968a4

-------------------------------------------------------------------
Mon Dec 12 05:47:11 CET 2011 - okir@suse.de

- Removed some dead code
- commit a17b68e

-------------------------------------------------------------------
Mon Dec 12 05:45:55 CET 2011 - okir@suse.de

- Removed obsolete rest api for dhcp4 service
- commit 4f8f9fa

-------------------------------------------------------------------
Mon Dec 12 05:44:09 CET 2011 - okir@suse.de

- Further work to get dhcp4 support working
- commit e504e1b

-------------------------------------------------------------------
Mon Dec 12 05:40:53 CET 2011 - okir@suse.de

- Added ni_timer_rearm()
- commit 50d201e

-------------------------------------------------------------------
Mon Dec 12 04:11:49 CET 2011 - okir@suse.de

- First steps towards making dhcp4 a dbus service
- commit b4b65be

-------------------------------------------------------------------
Mon Dec  5 11:19:49 CET 2011 - mt@suse.com

- Do not strip binaries in install by default
  Allow rpm to generate debuginfo package and strip itself.
  The new install-strip target can be used to force a strip.
- commit a1b055f

-------------------------------------------------------------------
Mon Dec  5 10:52:18 CET 2011 - mt@suse.com

- Accept single quoted strings in xml_get_token_tag
- commit e768048

-------------------------------------------------------------------
Mon Dec  5 10:14:37 CET 2011 - mt@suse.com

- Fixed xml_node_print return value
- commit 60ed66d

-------------------------------------------------------------------
Mon Dec  5 10:11:28 CET 2011 - mt@suse.com

- Removed unused retval variable
- commit cd0ee4b

-------------------------------------------------------------------
Wed Nov 30 06:19:40 CET 2011 - okir@suse.de

- Added ni_dbus_variant_{get,set}_uuid helpers
- commit a8ebba5

-------------------------------------------------------------------
Fri Oct 14 18:07:24 CEST 2011 - okir@suse.de

- Introduce __ni_device_refresh_link_info
- commit 40700bb

-------------------------------------------------------------------
Fri Oct 14 17:55:25 CEST 2011 - okir@suse.de

- Renamed vlan.interface_name to physdev_name
- commit 1a38153

-------------------------------------------------------------------
Fri Oct 14 17:50:26 CEST 2011 - okir@suse.de

- Renamed vlan.link to vlan.physdev_index
- commit d0776b8

-------------------------------------------------------------------
Fri Oct 14 17:47:32 CEST 2011 - okir@suse.de

- Fix vlan processing in __ni_process_ifinfomsg
- commit ded03b9

-------------------------------------------------------------------
Fri Oct 14 17:43:11 CEST 2011 - okir@suse.de

- Introduce __ni_vlan_new
- commit e1f2ec1

-------------------------------------------------------------------
Fri Oct 14 17:40:44 CEST 2011 - okir@suse.de

- Move the vlan pointer into struct link
- commit 36eb18a

-------------------------------------------------------------------
Fri Oct 14 17:37:39 CEST 2011 - okir@suse.de

- Change all ethtool functions to pass an ifname rather than the interface handle
- commit 0079196

-------------------------------------------------------------------
Mon Oct 10 15:31:21 CEST 2011 - okir@suse.de

- Consolidated a bunch of link-layer related members from ni_interface_t
  into a new ni_linkinfo_t struct.
- commit a99e334

-------------------------------------------------------------------
Wed Oct  5 13:26:43 CEST 2011 - okir@suse.de

- Begin to work on dbus based dhcp4 client
- commit cebc7e1

-------------------------------------------------------------------
Wed Oct  5 13:17:47 CEST 2011 - okir@suse.de

- Introduce ni_basename
- commit 123d993

-------------------------------------------------------------------
Tue Oct  4 21:58:17 CEST 2011 - okir@suse.de

- Remove some crud from wickedd
- commit 6512c6c

-------------------------------------------------------------------
Tue Oct  4 21:32:29 CEST 2011 - okir@suse.de

- Introduce uuid field in addrconf request and lease, so that we can match them
- commit 6f86d17

-------------------------------------------------------------------
Tue Oct  4 21:28:41 CEST 2011 - okir@suse.de

- Add ni_uuid_equal
- commit ee847cd

-------------------------------------------------------------------
Tue Oct  4 20:51:30 CEST 2011 - okir@suse.de

- Refactored addrconf code and fully support ifdown
- commit 9593936

-------------------------------------------------------------------
Tue Oct  4 08:57:10 CEST 2011 - okir@suse.de

- Support interface shutdown
- commit 7b3411d

-------------------------------------------------------------------
Tue Oct  4 08:31:24 CEST 2011 - okir@suse.de

- Worked on ifup implementation
- commit 31d8228

-------------------------------------------------------------------
Tue Oct  4 08:31:06 CEST 2011 - okir@suse.de

- Added two more error codes
- commit 5ecc39e

-------------------------------------------------------------------
Tue Oct  4 08:30:47 CEST 2011 - okir@suse.de

- Added another sample file for testing addr config
- commit 4e348aa

-------------------------------------------------------------------
Sat Oct  1 20:29:32 CEST 2011 - okir@suse.de

- For static addrconf, we should move the list of addrs and routes into an addrconf_request
- commit 31565a9

-------------------------------------------------------------------
Sat Oct  1 20:28:46 CEST 2011 - okir@suse.de

- Added new error code INVALID_ARGUMENTS
- commit c1f3b1b

-------------------------------------------------------------------
Sat Oct  1 18:58:05 CEST 2011 - okir@suse.de

- Introduce ni_addrconf_request_equal to compare addrconf requests
- commit f932d71

-------------------------------------------------------------------
Sat Oct  1 10:17:08 CEST 2011 - okir@suse.de

- Improve error reporting for dbus_objectmodel_new_* functions
- commit 9f94671

-------------------------------------------------------------------
Sat Oct  1 10:05:22 CEST 2011 - okir@suse.de

- Moved TRACE_DEBUG macros to new include file debug.h
- commit 773b92e

-------------------------------------------------------------------
Sat Oct  1 09:59:14 CEST 2011 - okir@suse.de

- Introduce macros WICKED_DBUS_${FOOBAR}_INTERFACE for VLAN, ETHERNET etc
- commit 8900af5

-------------------------------------------------------------------
Fri Sep 30 17:20:06 CEST 2011 - okir@suse.de

- Begin working on wicked ifup
- commit bdc0ea1

-------------------------------------------------------------------
Fri Sep 30 17:14:07 CEST 2011 - okir@suse.de

- Got rid of some old dict handling code no longer needed
- commit edf0c2e

-------------------------------------------------------------------
Fri Sep 30 17:10:02 CEST 2011 - okir@suse.de

- Introduce ni_dbus_object_get_properties_as_dict, and simplify Property.GetAll
  and ObjectManager.GetManagedObjects methods
- commit 75935c1

-------------------------------------------------------------------
Fri Sep 30 16:25:29 CEST 2011 - okir@suse.de

- Move ni_dbus_message_serialize_variants from dbus-client.c to dbus-message.c
- commit d6a071e

-------------------------------------------------------------------
Fri Sep 30 16:13:53 CEST 2011 - okir@suse.de

- Introduce ni_objectmodel_wrap_interface(), used to wrap a network device in a dummy dbus object
- commit b90cc57

-------------------------------------------------------------------
Fri Sep 30 08:31:54 CEST 2011 - okir@suse.de

- Introduce Interfac.up stub function
- commit 5bc13c8

-------------------------------------------------------------------
Fri Sep 30 08:31:07 CEST 2011 - okir@suse.de

- Introduce new pseudo syntax "sysconfig", which instructs the library
  to guess whether we're using suse or redhat sysconfig files
- commit 182ad5d

-------------------------------------------------------------------
Thu Sep 29 17:19:25 CEST 2011 - okir@suse.de

- Implement port removal for bridges
- commit 86d0ad6

-------------------------------------------------------------------
Thu Sep 29 17:16:36 CEST 2011 - okir@suse.de

- For now, ignore {NEW,DEL}LINK events coming from the bridge layer.
  These signal addition/removal of a port.
- commit 2858124

-------------------------------------------------------------------
Thu Sep 29 16:19:59 CEST 2011 - okir@suse.de

- Add missing support to parse port properties
- commit 04a7f47

-------------------------------------------------------------------
Thu Sep 29 15:48:36 CEST 2011 - okir@suse.de

- Trying to support port attributes when creating bridge ports
- commit 5bceacd

-------------------------------------------------------------------
Thu Sep 29 14:25:57 CEST 2011 - okir@suse.de

- Working on Bridge.addPort
- commit d54a46d

-------------------------------------------------------------------
Thu Sep 29 14:24:57 CEST 2011 - okir@suse.de

- Begin to introduce error codes and ni_strerror
- commit 1d498f6

-------------------------------------------------------------------
Thu Sep 29 11:33:22 CEST 2011 - okir@suse.de

- Add a parent pointer to ni_dbus_object
- commit 86df4b0

-------------------------------------------------------------------
Thu Sep 29 09:22:28 CEST 2011 - okir@suse.de

- Support per-property parser functions, and use it in wicked create
- commit 3ddb3e4

-------------------------------------------------------------------
Thu Sep 29 09:07:18 CEST 2011 - okir@suse.de

- Bridge creation and deletion is working now
- commit c84d5e9

-------------------------------------------------------------------
Thu Sep 29 08:41:49 CEST 2011 - okir@suse.de

- Support updating the configuration of a bridge interface
- commit 670d20d

-------------------------------------------------------------------
Wed Sep 28 21:05:36 CEST 2011 - okir@suse.de

- Some more work on bridge dbus interface
- commit 2e2f51b

-------------------------------------------------------------------
Wed Sep 28 21:05:08 CEST 2011 - okir@suse.de

- Some more cleanup of bridge types
- commit a642896

-------------------------------------------------------------------
Wed Sep 28 15:52:04 CEST 2011 - okir@suse.de

- First steps towards adding bridge support
- commit 51abde5

-------------------------------------------------------------------
Wed Sep 28 15:20:17 CEST 2011 - okir@suse.de

- In wicked_dbus_netif_refresh, now obtain the list of deleted interfaces
  directly from ni_refresh()
- commit cf4872e

-------------------------------------------------------------------
Wed Sep 28 15:15:43 CEST 2011 - okir@suse.de

- Optionally, allow ni_refresh to return list of deleted interfaces
- commit 51f1e65

-------------------------------------------------------------------
Wed Sep 28 15:15:34 CEST 2011 - okir@suse.de

- Update
- commit a065d1d

-------------------------------------------------------------------
Wed Sep 28 15:02:35 CEST 2011 - okir@suse.de

- Fix a typo in the vlan property macros
- commit 139e7fd

-------------------------------------------------------------------
Wed Sep 28 09:35:32 CEST 2011 - okir@suse.de

- Use new __ni_objectmodel_{get,set}_property helpers for interfaces, too
- commit 91cea49

-------------------------------------------------------------------
Wed Sep 28 09:27:37 CEST 2011 - okir@suse.de

- Introduce a set of generic dbus property get/set helpers, and use them in the ethernet interface
- commit 697faf4

-------------------------------------------------------------------
Wed Sep 28 09:05:29 CEST 2011 - okir@suse.de

- Introduce ni_dbus_variant_set_{int,uint,long,ulong}
  These functions cast the provided C type to the variant's type.
  The variant must be initialized to the desired type prior to this call.
- commit 5c5ddde

-------------------------------------------------------------------
Wed Sep 28 09:01:36 CEST 2011 - okir@suse.de

- When calling the get() function, initialize the variant prior to the call
- commit b7720ef

-------------------------------------------------------------------
Wed Sep 28 08:55:35 CEST 2011 - okir@suse.de

- Introduce ni_dbus_variant_init_signature to init an empty variant of a specific type
- commit dd6e90c

-------------------------------------------------------------------
Wed Sep 28 08:32:13 CEST 2011 - okir@suse.de

- Introduce helper functions ni_dbus_variant_get_{int,uint,long,ulong}
  These functions work off any integer compatible DBus type, casting
  the value to the requested destination type.
- commit 50c04fb

-------------------------------------------------------------------
Wed Sep 28 08:24:20 CEST 2011 - okir@suse.de

- Rename dbus_service member object_interface to name
- commit f909585

-------------------------------------------------------------------
Wed Sep 28 08:10:21 CEST 2011 - okir@suse.de

- Use dbus_variant_init_$foo_array instead of _set_array(NULL, 0)
- commit 4ff025d

-------------------------------------------------------------------
Wed Sep 28 07:59:15 CEST 2011 - okir@suse.de

- Fix a segfault when building the response to a Factory.create() call
- commit e74740e

-------------------------------------------------------------------
Tue Sep 27 23:23:58 CEST 2011 - okir@suse.de

- Move the interface refresh function from the individual Interfaces to
  the parent object.
- commit f66dff7

-------------------------------------------------------------------
Tue Sep 27 23:23:27 CEST 2011 - okir@suse.de

- Add __ni_dbus_object_unlink helper function (inline)
- commit 80f0c93

-------------------------------------------------------------------
Tue Sep 27 23:22:41 CEST 2011 - okir@suse.de

- In ni_dbus_object_free, defer destruction of active objects
- commit ac001b7

-------------------------------------------------------------------
Tue Sep 27 22:22:41 CEST 2011 - okir@suse.de

- Introduce helper function ni_handle_seq
- commit 991ba23

-------------------------------------------------------------------
Tue Sep 27 22:14:26 CEST 2011 - okir@suse.de

- Implement delete method
- commit 0ae021f

-------------------------------------------------------------------
Tue Sep 27 22:12:42 CEST 2011 - okir@suse.de

- DBus doesn't support dots in an object path, which happens eg with vlan interfaces
  We work around this by sanitizing the name we use for the object path. At the
  same time, we add a "name" property to the interface service.
- commit b0d16ea

-------------------------------------------------------------------
Tue Sep 27 19:02:39 CEST 2011 - okir@suse.de

- Factory.create now returns the object path of the newly created object
- commit fee8b63

-------------------------------------------------------------------
Tue Sep 27 16:12:36 CEST 2011 - okir@suse.de

- Get rid of old ni_dbus_process_properties functionaity
- commit 1fc5d2d

-------------------------------------------------------------------
Tue Sep 27 16:10:39 CEST 2011 - okir@suse.de

- Remove some dead code in wpa-supplicant client
- commit 1940f05

-------------------------------------------------------------------
Tue Sep 27 16:08:50 CEST 2011 - okir@suse.de

- Change the wpa interface capabilities code to use object_properties_from_dict
- commit 7c7cf34

-------------------------------------------------------------------
Tue Sep 27 15:39:17 CEST 2011 - okir@suse.de

- Change the wpa-bss code to use ni_dbus_object_set_properties_from_dict
- commit 7087f23

-------------------------------------------------------------------
Tue Sep 27 15:38:25 CEST 2011 - okir@suse.de

- Fix a few bugs in dbus_variant_get_int*, and introduce ni_dbus_object_set_properties_from_dict
- commit fd28b5c

-------------------------------------------------------------------
Tue Sep 27 14:36:25 CEST 2011 - okir@suse.de

- Removed some dead code from dbus_dict.c
- commit ce76592

-------------------------------------------------------------------
Tue Sep 27 14:29:25 CEST 2011 - okir@suse.de

- Minor cleanup in wpa-supplicant client
- commit b5118b8

-------------------------------------------------------------------
Tue Sep 27 14:10:39 CEST 2011 - okir@suse.de

- Make do_show actually show something!
- commit 756dd77

-------------------------------------------------------------------
Tue Sep 27 14:10:22 CEST 2011 - okir@suse.de

- Add missing code to set routes property
- commit cf233fb

-------------------------------------------------------------------
Tue Sep 27 14:09:43 CEST 2011 - okir@suse.de

- Implement missing ni_dbus_dict_get_string
- commit c6aded0

-------------------------------------------------------------------
Tue Sep 27 14:08:57 CEST 2011 - okir@suse.de

- In ni_route_free, free additional route nexthops
- commit 749ff61

-------------------------------------------------------------------
Tue Sep 27 13:43:07 CEST 2011 - okir@suse.de

- Refresh all objects in GetManagedObjects (this still needs some work though)
- commit 5b5b18f

-------------------------------------------------------------------
Tue Sep 27 13:25:20 CEST 2011 - okir@suse.de

- Fix decoding of address property
- commit 2c58ca0

-------------------------------------------------------------------
Tue Sep 27 10:19:27 CEST 2011 - okir@suse.de

- Add some missing pieces to 'show' command
- commit f967b2f

-------------------------------------------------------------------
Tue Sep 27 09:54:06 CEST 2011 - okir@suse.de

- Get rid of ni_dbus_proxy_t and use dbus_object instead
- commit c8aceea

-------------------------------------------------------------------
Tue Sep 27 08:52:41 CEST 2011 - okir@suse.de

- When creating a dbus child object, optionally call the parent's init_child() method
- commit d333561

-------------------------------------------------------------------
Tue Sep 27 08:47:22 CEST 2011 - okir@suse.de

- Get rid of __ni_dbus_server_object_new
- commit 2389b59

-------------------------------------------------------------------
Tue Sep 27 08:44:40 CEST 2011 - okir@suse.de

- Change __ni_dbus_object_new to alloc a copy of the path argument
- commit 4d35f32

-------------------------------------------------------------------
Tue Sep 27 08:39:35 CEST 2011 - okir@suse.de

- Introduce helper function dbus_object_insert
- commit 1c58c60

-------------------------------------------------------------------
Tue Sep 27 08:34:14 CEST 2011 - okir@suse.de

- Got rid of ni_dbus_server_create_anonymous_object
- commit c5a7fc2

-------------------------------------------------------------------
Tue Sep 27 08:14:19 CEST 2011 - okir@suse.de

- Move some generic dbus_object functions to dbus-object.[hc]
- commit 23d920a

-------------------------------------------------------------------
Tue Sep 27 07:10:29 CEST 2011 - okir@suse.de

- Split off the server-specific part of dbus_object
- commit ebfb8fa

-------------------------------------------------------------------
Mon Sep 26 18:04:29 CEST 2011 - okir@suse.de

- Work on refreshing client state via GetManagedObjects
- commit 809af6d

-------------------------------------------------------------------
Mon Sep 26 12:34:49 CEST 2011 - okir@suse.de

- Add DBus config file for wicked
- commit 1af6c3a

-------------------------------------------------------------------
Thu Sep 22 21:35:06 CEST 2011 - okir@suse.de

- Try to support create method in the wicked client
- commit 9374edb

-------------------------------------------------------------------
Wed Sep 21 22:19:52 CEST 2011 - okir@suse.de

- Introduce Interface.create() method
- commit 5719237

-------------------------------------------------------------------
Wed Sep 21 13:54:57 CEST 2011 - okir@suse.de

- Introduce ni_interface_create_vlan
- commit d086ca5

-------------------------------------------------------------------
Wed Sep 21 13:53:33 CEST 2011 - okir@suse.de

- Introduce ni_interface_by_vlan_tag
- commit dc947d7

-------------------------------------------------------------------
Wed Sep 21 11:55:16 CEST 2011 - okir@suse.de

- Added DBus interface for VLAN netifs
- commit 18ba7b8

-------------------------------------------------------------------
Wed Sep 21 11:48:06 CEST 2011 - okir@suse.de

- Add ni_dbus_variant_get_{string,byte}
- commit fe41003

-------------------------------------------------------------------
Wed Sep 21 11:35:42 CEST 2011 - okir@suse.de

- Remove some dead code
- commit 169cc2c

-------------------------------------------------------------------
Wed Sep 21 11:33:41 CEST 2011 - okir@suse.de

- Handle dbus property.set differently
- commit f215c67

-------------------------------------------------------------------
Wed Sep 21 08:18:43 CEST 2011 - okir@suse.de

- When processing Properties.Set call, identify the field(s) modified by the call
- commit f9755c7

-------------------------------------------------------------------
Wed Sep 21 08:18:18 CEST 2011 - okir@suse.de

- Implement ni_bitfield type
- commit 75d0930

-------------------------------------------------------------------
Tue Sep 20 22:37:34 CEST 2011 - okir@suse.de

- Explicitly mark read-only properties (presence of a set() function
  no longer means anything)
- commit 658cb71

-------------------------------------------------------------------
Tue Sep 20 22:32:28 CEST 2011 - okir@suse.de

- Implement more set() functions for interface and ethernet properties
- commit 5a0e6a6

-------------------------------------------------------------------
Tue Sep 20 15:45:20 CEST 2011 - okir@suse.de

- When libdbus unregisters an object, it invokes our __ni_dbus_object_unregister
  callback. Make this callback call the object's destuctor, if there is one.
- commit 448b2ce

-------------------------------------------------------------------
Tue Sep 20 15:42:33 CEST 2011 - okir@suse.de

- Implement basic object functions for netifs
- commit b65202c

-------------------------------------------------------------------
Tue Sep 20 15:35:13 CEST 2011 - okir@suse.de

- Introduce generic per-object functions
- commit 3edb2ae

-------------------------------------------------------------------
Tue Sep 20 15:23:12 CEST 2011 - okir@suse.de

- Fix make depend target to also work with src/dbus-objects/*.c
- commit 2a9b5b9

-------------------------------------------------------------------
Tue Sep 20 13:16:34 CEST 2011 - okir@suse.de

- Made ni_dbus_service_t a public object
- commit 3f70f32

-------------------------------------------------------------------
Tue Sep 20 11:41:08 CEST 2011 - okir@suse.de

- Register ObjectManager interface for all dbus server objects
- commit 9ec4201

-------------------------------------------------------------------
Tue Sep 20 11:30:48 CEST 2011 - okir@suse.de

- wpa-supplicant client: do not unbind from dbus when we lose the association
  with the access point
- commit 55e013d

-------------------------------------------------------------------
Tue Sep 20 11:00:08 CEST 2011 - okir@suse.de

- Avoid nested variant array/variant array in routes and addresses properties
- commit 721fc8c

-------------------------------------------------------------------
Tue Sep 20 10:37:16 CEST 2011 - okir@suse.de

- Trying to send an invalid message make libdbus close the connection.
  Avoid that if we can...
- commit 0bffbfa

-------------------------------------------------------------------
Tue Sep 20 10:04:42 CEST 2011 - okir@suse.de

- Implement interface routes property
- commit 6255959

-------------------------------------------------------------------
Tue Sep 20 09:49:52 CEST 2011 - okir@suse.de

- Add Interface.addresses property
- commit 5a05364

-------------------------------------------------------------------
Tue Sep 20 09:49:20 CEST 2011 - okir@suse.de

- Support arrays of variants
- commit a67af5c

-------------------------------------------------------------------
Tue Sep 20 08:09:42 CEST 2011 - okir@suse.de

- Support dbus dicts in message building
- commit f7e52b8

-------------------------------------------------------------------
Mon Sep 19 23:10:53 CEST 2011 - okir@suse.de

- Add dict support to ni_dbus_variant
- commit d76c35c

-------------------------------------------------------------------
Mon Sep 19 15:23:45 CEST 2011 - okir@suse.de

- Some more dbus cleanups
- commit b40563f

-------------------------------------------------------------------
Mon Sep 19 15:22:04 CEST 2011 - okir@suse.de

- Fixed a newly introduced bug in ni_dbus_variant_set_*_array
- commit 1b97610

-------------------------------------------------------------------
Mon Sep 19 15:13:33 CEST 2011 - okir@suse.de

- Moved some more dbus code around
- commit f596957

-------------------------------------------------------------------
Mon Sep 19 15:04:29 CEST 2011 - okir@suse.de

- Yet another cleanup of ni_dbus_message_iter_get_variant
- commit dd15069

-------------------------------------------------------------------
Mon Sep 19 14:53:15 CEST 2011 - okir@suse.de

- More dbus-dict simplifications
- commit 8bde950

-------------------------------------------------------------------
Mon Sep 19 14:33:10 CEST 2011 - okir@suse.de

- Further simplified some of the dbus-dict code
- commit 699f4f9

-------------------------------------------------------------------
Mon Sep 19 08:34:38 CEST 2011 - okir@suse.de

- Fix incomplete ni_variant destructor
- commit d8ed84f

-------------------------------------------------------------------
Mon Sep 19 08:32:31 CEST 2011 - okir@suse.de

- Fix Ethernet interface to reflect API changes for properties and methods
- commit 0da6330

-------------------------------------------------------------------
Mon Sep 19 08:30:44 CEST 2011 - okir@suse.de

- Fix GetManagedObjects method
- commit bf84e64

-------------------------------------------------------------------
Mon Sep 19 08:29:14 CEST 2011 - okir@suse.de

- Add support for string arrays in dbus properties
- commit 66a4b17

-------------------------------------------------------------------
Sun Sep 18 18:33:57 CEST 2011 - okir@suse.de

- Rewrite dbus method handling
- commit 0afebab

-------------------------------------------------------------------
Sun Sep 18 17:32:53 CEST 2011 - okir@suse.de

- Introduce generic dbus method argument parsing
- commit bdc6453

-------------------------------------------------------------------
Sun Sep 18 12:32:47 CEST 2011 - okir@suse.de

- Add dbus interface.addresses property (incomplete)
- commit 7b641b8

-------------------------------------------------------------------
Sat Sep 17 23:00:41 CEST 2011 - okir@suse.de

- Fix another set of get/set/handler return types
- commit 93a9311

-------------------------------------------------------------------
Sat Sep 17 22:52:04 CEST 2011 - okir@suse.de

- ni_dbus_object_register_service: always register new services at the end
- commit f500807

-------------------------------------------------------------------
Sat Sep 17 22:49:22 CEST 2011 - okir@suse.de

- Change the return type of method handler from int to bool
- commit 5473dcf

-------------------------------------------------------------------
Sat Sep 17 22:38:27 CEST 2011 - okir@suse.de

- Moved dbus objects into their own files
- commit c3242fd

-------------------------------------------------------------------
Sat Sep 17 19:03:03 CEST 2011 - okir@suse.de

- Fixed the dbus type of data returned by Properties.GetAll
- commit 7afee27

-------------------------------------------------------------------
Sat Sep 17 18:54:06 CEST 2011 - okir@suse.de

- Add Interface.Ethernet, and export some ethernet properties
- commit b96a536

-------------------------------------------------------------------
Sat Sep 17 18:51:25 CEST 2011 - okir@suse.de

- Fix a segfault in Properties.Get() if interface is ""
- commit 4729a08

-------------------------------------------------------------------
Sat Sep 17 18:18:05 CEST 2011 - okir@suse.de

- Add more properties to Interface
- commit 190c2f8

-------------------------------------------------------------------
Sat Sep 17 18:15:59 CEST 2011 - okir@suse.de

- Fix Properties.* methods to support an empty interface name as wildcard
- commit 84c6743

-------------------------------------------------------------------
Sat Sep 17 18:05:46 CEST 2011 - okir@suse.de

- Fix ni_dbus_dict_append_variant to also handle complex types
- commit 9c149c4

-------------------------------------------------------------------
Sat Sep 17 17:56:45 CEST 2011 - okir@suse.de

- Support byte_array properties (get only)
- commit ef09a01

-------------------------------------------------------------------
Sat Sep 17 16:41:44 CEST 2011 - okir@suse.de

- Add ni_dbus_variant_sprint helper function
- commit df87398

-------------------------------------------------------------------
Sat Sep 17 16:32:28 CEST 2011 - okir@suse.de

- Add ni_dbus_message_iter_get_variant() and complete Properties.Set implementation
- commit b2a84dd

-------------------------------------------------------------------
Sat Sep 17 16:11:46 CEST 2011 - okir@suse.de

- Simplify the code for adding variants to a message
- commit 5af3e5e

-------------------------------------------------------------------
Sat Sep 17 11:09:20 CEST 2011 - okir@suse.de

- Introduce dbus properties and some support functions
- commit f18cd37

-------------------------------------------------------------------
Fri Sep 16 17:41:36 CEST 2011 - okir@suse.de

- rename DBus Interface
- commit 06d74cd

-------------------------------------------------------------------
Fri Sep 16 17:41:23 CEST 2011 - okir@suse.de

- Added missing dbus.h include
- commit 486a629

-------------------------------------------------------------------
Fri Sep 16 17:38:23 CEST 2011 - okir@suse.de

- Add primitive dbus interface to network interfaces
- commit 3fc28b4

-------------------------------------------------------------------
Fri Sep 16 17:12:29 CEST 2011 - okir@suse.de

- Make all dbus objects support ObjectManager interface by default
- commit b7a1d1a

-------------------------------------------------------------------
Fri Sep 16 13:12:45 CEST 2011 - okir@suse.de

- Avoid warning when offsetof is already defined
- commit a1a5883

-------------------------------------------------------------------
Fri Sep 16 13:11:19 CEST 2011 - okir@suse.de

- First steps towards adding a dbus interface
- commit f86c25d

-------------------------------------------------------------------
Thu Sep 15 19:51:20 CEST 2011 - okir@suse.de

- Introduced dbus_server stuff
- commit 0fbc551

-------------------------------------------------------------------
Thu Sep 15 14:38:05 CEST 2011 - okir@suse.de

- Some more dbus connection rework
- commit fca79fb

-------------------------------------------------------------------
Thu Sep 15 08:53:50 CEST 2011 - okir@suse.de

- Moved some stuff from dbus-client.c to dbus-common.c
- commit 1c6d76c

-------------------------------------------------------------------
Thu Sep 15 08:44:59 CEST 2011 - okir@suse.de

- Removed superfluous client arg from ni_dbus_method_call_new(), and
  renamed all ni_dbus_method_call_* functions to ni_dbus_proxy_call_*
- commit 2a66596

-------------------------------------------------------------------
Thu Sep 15 08:43:25 CEST 2011 - okir@suse.de

- Moved generic connection related code from dbus-client.c to dbus-common.c
- commit 535e69c

-------------------------------------------------------------------
Thu Sep 15 08:23:36 CEST 2011 - okir@suse.de

- Allow ni_dbus_add_signal_handler getting called with NULL sender
- commit cbe3294

-------------------------------------------------------------------
Wed Sep 14 17:16:19 CEST 2011 - okir@suse.de

- Rename ni_dbus_call_simple to ni_dbus_proxy_call_simple and omit
  client argument
- commit c50acc6

-------------------------------------------------------------------
Wed Sep 14 17:12:22 CEST 2011 - okir@suse.de

- Get rid of ni_dbus_client_application_data stuff
- commit 886d853

-------------------------------------------------------------------
Wed Sep 14 17:08:07 CEST 2011 - okir@suse.de

- Simplify the ni_dbus_proxy_async_call interface
- commit 27567cd

-------------------------------------------------------------------
Wed Sep 14 16:59:30 CEST 2011 - okir@suse.de

- Change ni_dbus_call_async to use new proxy data (client)
- commit cf55315

-------------------------------------------------------------------
Wed Sep 14 16:53:37 CEST 2011 - okir@suse.de

- Rename ni_dbus_object to ni_dbus_proxy, and introduce new fields for client and local_data
- commit 645bc16

-------------------------------------------------------------------
Wed Sep 14 16:44:39 CEST 2011 - okir@suse.de

- Make an attempt to split dbus client/connection specific functionality
- commit eeef793

-------------------------------------------------------------------
Wed Sep 14 15:05:31 CEST 2011 - okir@suse.de

- Moved bus_name string from ni_dbus_client to ni_dbus_object
- commit 56ae5ad

-------------------------------------------------------------------
Tue Sep 13 16:40:53 CEST 2011 - okir@suse.de

- Started to move some common dbus code into dbus-common.[hc]
- commit b34e66b

-------------------------------------------------------------------
Tue Sep 13 15:19:22 CEST 2011 - okir@suse.de

- Som more changes related to wpa-supplicant
- commit 209634c

-------------------------------------------------------------------
Tue Apr  5 17:02:34 CEST 2011 - mt@suse.de

- Added dbus-1-devel to build require in spec file
- commit 15f5c83

-------------------------------------------------------------------
Tue Mar 29 13:09:08 CEST 2011 - okir@suse.de

- Replace unchecked strdup calls with calls to xstrdup
- commit a459ee6

-------------------------------------------------------------------
Tue Mar 29 12:56:59 CEST 2011 - okir@suse.de

- Encode interface wireless capabilities as part of <interface> XML
- commit f2b7621

-------------------------------------------------------------------
Tue Mar 29 12:55:38 CEST 2011 - okir@suse.de

- Rewrite wireless code to use wpa_supplicant instead of homegrown IW handling
- commit c65f4ba

-------------------------------------------------------------------
Tue Mar 29 12:30:46 CEST 2011 - okir@suse.de

- Minor update to TODO file
- commit 42eeea5

-------------------------------------------------------------------
Tue Mar 29 12:28:18 CEST 2011 - okir@suse.de

- Get rid of more dead code in wireless handling
- commit 42da4a5

-------------------------------------------------------------------
Tue Mar 29 12:21:09 CEST 2011 - okir@suse.de

- Remove useless notes from dbus-client.c
- commit b3d513d

-------------------------------------------------------------------
Tue Mar 29 11:38:00 CEST 2011 - okir@suse.de

- Add wpa_supplicant client
- commit a445b0b

-------------------------------------------------------------------
Tue Mar 29 11:31:17 CEST 2011 - okir@suse.de

- Add dbus client functions
- commit f57b85e

-------------------------------------------------------------------
Tue Mar 29 11:28:55 CEST 2011 - okir@suse.de

- socket API: allow abstraction of other send/recv functionality
  This is in preparation of integrating dbus sockets into this framework
- commit 7af9725

-------------------------------------------------------------------
Tue Mar 29 11:27:43 CEST 2011 - okir@suse.de

- buffer.h: need to include util_priv.h for xcalloc decl
- commit 9f67de0

-------------------------------------------------------------------
Tue Mar 29 11:23:36 CEST 2011 - okir@suse.de

- Clean up netlink code a little
  Move __ni_netlink_open/close from state.c to kernel.c
  Simplify some code in nl_talk.
- commit e0d12a6

-------------------------------------------------------------------
Tue Mar 29 11:04:15 CEST 2011 - okir@suse.de

- Introduce new ni_opaque_new/free functions
- commit a2f1666

-------------------------------------------------------------------
Tue Mar 29 11:03:27 CEST 2011 - okir@suse.de

- Introduce new helper function ni_link_address_set
- commit f695d65

-------------------------------------------------------------------
Tue Mar 29 11:01:42 CEST 2011 - okir@suse.de

- Introduce debug facilities for wireless and DBus client
- commit 77d2040

-------------------------------------------------------------------
Tue Mar 29 10:55:01 CEST 2011 - okir@suse.de

- Add new header file util_priv.h
- commit d8ef8df

-------------------------------------------------------------------
Tue Mar 29 07:14:03 CEST 2011 - okir@suse.de

- Add new macro ni_assert
- commit ad22567

-------------------------------------------------------------------
Wed Feb 23 21:21:34 CET 2011 - mt@suse.de

- Corrected ni_server_listen failure check
- commit 80e28f3

-------------------------------------------------------------------
Wed Feb 23 21:20:40 CET 2011 - mt@suse.de

- Write resolv.conf search domains as one line
- commit 0de3509

-------------------------------------------------------------------
Mon Dec  6 11:45:57 CET 2010 - okir@suse.de

- Encapsulate netlink handle, cache and cb into one object
- commit 73c71b3

-------------------------------------------------------------------
Mon Dec  6 09:33:50 CET 2010 - okir@suse.de

- spec file should build-require libnl-devel
- commit 6288068

-------------------------------------------------------------------
Mon Dec  6 09:16:05 CET 2010 - okir@suse.de

- Note to self: handle IFLA_OPERSTATE
- ethernet: handle "n/a" ethtool duplex/speed values when link is down
  When the link is down, ethtool reports a speed of 65535 and a duplex
  value of 255. Handle these gracefully.
- netcf: add missing code to parse <wireless-scan> element
- Implement a bunch of missing wireless helper functions
- Introduce generic ni_{format,parse}_hex functions
  These functions represent arbitrary hex data as a sequence of
  octets, separated by a colon.
  Also changes ni_link_address_{format,parse} functions to use these
  new helpers.
- Make ni_parse_{int,double.*} return error if the argument string is NULL
- Add ni_parse_double
- When processing wireless events, we were looking at a freed ni_interface
- Bugfixes and enhancements to previous change set
- Clean up code after netlink changes
- Replace low-level handling of rtnetlink sockets with libnl functions
- Begin to support wireless events
- wicked.7: document scanning interface
- REST API: delete /system/interface/<ifname>/scan will purge scan results
- wireless: do not try to retrieve scan results more than once
- REST API: split wireless scan into two operations: put requests a scan, get retrieves results
- wireless: in get_scan_results, do not query wireless unless a scan is active
- Intoduce ni_interface_check_activity
- wireless: if we had to bring up the interface for scanning; take it down again after some time
- Add timer support to wickedd
- Got wireless scanning to work
- First steps towards wireless scanning
- commit 9ea709e

-------------------------------------------------------------------
Mon Nov 29 16:40:20 CET 2010 - okir@suse.de

- Some cleanup of socket code
- commit 8114f6b

-------------------------------------------------------------------
Mon Nov 29 16:36:26 CET 2010 - okir@suse.de

- Converted remaining socket services to new process_request callback
  Get rid of all ni_socket_pull() functionality in the process
- commit af96ced

-------------------------------------------------------------------
Mon Nov 29 16:27:59 CET 2010 - okir@suse.de

- Do not complain all the time about unsupported ETHTOOL options
- commit 6b79f16

-------------------------------------------------------------------
Mon Nov 29 16:27:34 CET 2010 - okir@suse.de

- ni_socket: Fix up dgram receive handling after most recent changes
- commit ce90445

-------------------------------------------------------------------
Mon Nov 29 15:30:57 CET 2010 - okir@suse.de

- TODO update
- commit b26d933

-------------------------------------------------------------------
Mon Nov 29 15:30:21 CET 2010 - okir@suse.de

- ni_socket: redo all I/O processing to no longer require forking a subprocess
- commit 7fee1ab

-------------------------------------------------------------------
Mon Nov 29 14:27:09 CET 2010 - okir@suse.de

- ignore SIGPIPE - needed for socket code
- commit 3701fdd

-------------------------------------------------------------------
Mon Nov 29 14:26:24 CET 2010 - okir@suse.de

- ni_buffer: add support for dynamically allocated buffers
- commit 5ab5561

-------------------------------------------------------------------
Sun Nov 28 20:34:52 CET 2010 - okir@suse.de

- Move ni_socket_t type into private header file
- commit f84ed7d

-------------------------------------------------------------------
Sun Nov 28 20:23:46 CET 2010 - okir@suse.de

- config: add recv_max field to ni_config
- commit b5bc39c

-------------------------------------------------------------------
Fri Nov 26 17:54:26 CET 2010 - okir@suse.de

- Worked on ethtool support: we can now retrieve general settings, and offload info.
  Next: support changing these settings.
  Other settings, such as jumbo frames or ring sizes, to follow later.
- commit cd5e22a

-------------------------------------------------------------------
Fri Nov 26 14:57:20 CET 2010 - okir@suse.de

- Introduce /system/interface/<ifname>/stats node, and related functions
- commit 6c63042

-------------------------------------------------------------------
Fri Nov 26 14:34:14 CET 2010 - okir@suse.de

- Add xml_node_{add,get}_attr_ulong and supporting helpers
- commit 6e5d357

-------------------------------------------------------------------
Fri Nov 26 13:41:44 CET 2010 - okir@suse.de

- Changed the call signature of REST API functions
  Up to now, all REST functions took an additional "path" argument,
  providing the unparsed portion of the path. This was used to
  access individual interfaces, which of course do not have a
  static representation in the REST namespace.
  The new approach is to have an optional "wildcard" child in each
  node of the REST namespace. When encountering a path element that
  we don't know, and there is such a "wildcard" child, we simply store
  the path component in a new argv[] vector in the ni_wicked_request,
  and proceed with the wildcard node.
- commit 976cb0f

-------------------------------------------------------------------
Fri Nov 26 11:53:55 CET 2010 - okir@suse.de

- DHCP, IPv4LL: log an error message if we failed to send the event to our master
- commit 869661c

-------------------------------------------------------------------
Fri Nov 26 11:51:33 CET 2010 - okir@suse.de

- src/util.c needs sys/stat.h on some code bases
- commit 9f789e3

-------------------------------------------------------------------
Fri Nov 26 11:40:53 CET 2010 - okir@suse.de

- Get rid of obsolete xstreq function
- commit 9ae9776

-------------------------------------------------------------------
Fri Nov 26 11:36:46 CET 2010 - okir@suse.de

- Add spec file contributed by Matthias Eckermann
- commit de928f8

-------------------------------------------------------------------
Fri Nov 26 11:34:37 CET 2010 - okir@suse.de

- No longer install wicked-dhcp4
- commit 9856515

-------------------------------------------------------------------
Fri Nov 26 11:29:56 CET 2010 - okir@suse.de

- init script: Should-stop should use $null rather than empty string
- commit d8e33a9

-------------------------------------------------------------------
Fri Nov 26 11:24:15 CET 2010 - okir@suse.de

- Replace all (struct sockaddr_foobar) casts with accesses to ni_sockaddr_t union members
  This is to avoid gcc warnings/errors triggered by strict aliasing checks
- commit 27cc7c4

-------------------------------------------------------------------
Fri Nov 26 11:14:24 CET 2010 - okir@suse.de

- Introduce ni_sockaddr_set_ipv{4,6}() functions
- commit d1707f9

-------------------------------------------------------------------
Fri Nov 26 11:10:28 CET 2010 - okir@suse.de

- Replace all uses of sockaddr_storage with ni_sockaddr_t
- commit c1c5507

-------------------------------------------------------------------
Fri Nov 26 10:59:47 CET 2010 - okir@suse.de

- Introduce ni_sockaddr union
- commit d7ca536

-------------------------------------------------------------------
Fri Nov 26 10:53:19 CET 2010 - okir@suse.de

- Fix gcc complaint about unitialized variable
- commit eeeb28c

-------------------------------------------------------------------
Thu Nov 25 15:48:53 CET 2010 - okir@suse.de

- Add ANNOUNCE file
- commit 836375f

-------------------------------------------------------------------
Thu Nov 25 14:30:55 CET 2010 - okir@suse.de

- Fix header comment in dhcp/rest-api.c
- commit 48b728d

-------------------------------------------------------------------
Thu Nov 25 14:30:19 CET 2010 - okir@suse.de

- DHCP protocol processing: fix an error goto
- commit b060be3

-------------------------------------------------------------------
Thu Nov 25 14:08:06 CET 2010 - okir@suse.de

- Update README
- commit ef69071

-------------------------------------------------------------------
Thu Nov 25 12:23:02 CET 2010 - okir@suse.de

- Update manpages
- commit 9a4491b

-------------------------------------------------------------------
Thu Nov 25 10:05:13 CET 2010 - okir@suse.de

- Updated wicked(7) manpage
- commit 11d186e

-------------------------------------------------------------------
Thu Nov 25 09:00:00 CET 2010 - okir@suse.de

- get rid of obsolete <start mode> elements in sample netcf files
- wicked: when bringing down vlans etc, ignore children we're not asked to shut down
- ni_interface_new: provide reasonable defaults for manual ifup/ifdown behavior
- ni_syntax_parse_file: fix a long-standing bug when trying to parse relative path names
- wicked: worked on ifdown/delete
- Update
- wicked: remove some dead code
- wicked: handle interfaces with only-if-link flag
- wicked: do not show animated progress meter when piping to a file
- wicked: more fixes to the finite state machine
- wicked: some code cleanups
- wicked: next rewrite
- wicked: try to simplify code again
- wickedd: when configuring an interface in response to a link-up event, use ni_interface_configure2
- Fix an overoptimization in ni_policy_match_interface
- wicked: some more fixes to bring-up code
- When installing a new policy, check whether we should apply it right away
  For instance, a typical case when we would need to apply a new policy would
  be if it's attached to a link-up event, and the device(s) matching this
  policy have their link up already.
- Introduce ni_debug_wicked_xml to display arguments and results of wicked calls
- Fix ambiguity in interface XML represenation of addrconf info
- Added consistency check in ni_dhcp_addrconf_{request,release}
- In __ni_interface_addrconf, properly copy/clone addrconf_requests
  Prior to this change, we just used to move the request from the
  config interface to the current interface; this no longer works when
  we use policies.
- Change ni_interface_configure interface
  New interface:
  extern int   ni_interface_configure(ni_handle_t *, const ni_interface_t *);
  extern int   ni_interface_configure2(ni_handle_t *, ni_interface_t *, const ni_interface_t *);
  This removes the xml_node argument, and adds an optional interface
  to be replaced/modified. The "config" interface has now been constified,
  requiring constifying all over the place.
- Remove xml_node argument from ni_interface_configure API
- Misc changes for auto interface bringup - we're getting closer
- Add NI_INTERFACE_ARRAY_INIT macro
- SUSE backend: disable IPv6 autoconf on loopback interfaces
- Introduce __ni_address_list_find; do not mark all IPv6 addrs as "autoconf"
- Make sure we always assign 127.0.0.1 and ::1/128 to loopback - SUSE expects this
- ni_indirect: previous changes introduced bug in interface_configure
- Implement policy_update function for ni_indirect handles
- Fix a bug in __ni_policy_list_destroy
- Introduce __ni_policy_clone
- Streamline code in ni_indirect
- Renamed some policy related functions
- Move policy info into ni_handle, and get rid of global policy info
- Allow unnamed interfaces in <policy> elements
- Added xcalloc/xstrdup allocation helpers
- Next stage for the big policy rewrite
- Rewrite interface policy - we want to do it all differently
  Instead of defining all policy in a global policy.xml file,
  we want the wicked client to feed us policy info at boot time,
  taking that from the interface definitions found in the usual
  /etc/sysconfig locations. At a later time, we can add new config
  storage.
- wicked: try to handle link down, when we're supposed to do automatic bring-up
  For now, we just request wickedd to bring up the network anyway, and
  hope that all addrconf modes can cope with a link down situation
- In ni_dhcp_addrconf_request, don't require that link is up.
- fix how we determine whether the master wants us to start or stop DHCP
  DHCP: fix how we determine whether the master wants us to start or stop DHCP
  on a given interface. We used to check for IFF_UP, but that's broken. Instead,
  check whether ADDRCONF_DHCP is enabled for IPv4
- ni_scandir: get rid of annoying debug printf
- wicked: more code cleanup
- wicked: removed some dead code
- wicked: move some code about
- Introduce ni_interface_set_{ethernet,vlan,bridge,bonding}
  Introduce ni_interface_set_{ethernet,vlan,bridge,bonding} and
  obsolete __ni_interface_clear_{ethernet,vlan,bridge,bonding}.
- wicked: progress meter fixes
- wicked: convert ifdown to new approach, and remove dead code
- wicked: cleaned out some old code
- wicked: fixing new ifup logic, allowing for separate link-up/network-up steps.
- socket: fixed up connect error reporting
- wicked: another rewrite of the interface bring-up machinery
- Introduced dummy ni_ethernet type (will carry device config like link speed)
- Abstract ni_interface.flags
  We used to use standard IFF_* flags in the ni_interface object.
  Which is bad, because not all IFF_* bits are public, and we don't want
  to pollute the namespace of all application .c files
- commit 7b00dd9

-------------------------------------------------------------------
Mon Nov  8 09:00:00 CET 2010 - okir@suse.de

- Update TODO
- wicked: build_partial_topology semantics cleanup
- Add only-if-link flag to startmode definition
- When configuring an interface, weed out duplicates from ifp->addrs first
- Two bugfixes for __ni_interface_address_exists
  a) AF_INET6 case was completely broken
  b) In AF_INET case, compare full local_addr, not just the prefix
- Introduce __ni_address_list_dedup (to weed out dupes in ifp->addrs)
- Implement and export ni_address_clone, ni_address_list_append
- Update wicked manpage to reflect latest changes
- wicked: make "wicked ifdown shutdown" work
- ifconfig: when asked to configure an interface type we don't know, do not error out
  Previously, we returned an error if the user tried to configure an interface
  type we didn't know about, such as wireless. This patch changes this to a
  debug message, and proceeds with configuring the device. This allows us to
  at least take down eg a wireless interface, even if we cannot configure it
  properly.
- Split NI_IFACTIO_MANUAL into MANUAL_UP/DOWN
- wicked: implement client side support to wait for interfaces when they come up
- Clean up interface behavior definitions for suse/redhat
- ifconfig: when IPv6 autoconf is enabled, create a dummy lease on the fly
- Redid the ifbehavior struct completely to be more uniform
- wicked: move around code a little bit
- ni_indirect: fix two bugs in ni_indirect_interface_refresh_one
  The interface returned by ni_syntax_xml_to_interface has already
  been added to the ni_handle's iflist.
  In success case, make sure we return 0, not -1.
- Misc fixes to wicked client
  In ifup, we need to mark all interfaces as "up" first, before
  we do the device hierarchy fandango.
  Also, add some debug blubber for --dry-run
- ni_indirect: strip off bogus top-level XML node when handling responses
- Massive rewrite of wicked client: no longer manipulate XML directly, but convert to interface objects and use these
- Make __ni_interface_clear_{addresses,routes} public
- Implement ni_interface_refresh_one for indirect handles
- Add ni_interface_refresh_one(ni_handle_t *nih, const char *name)
  This function can be used to refresh the status of a single interface
- Add ni_interface_array_index()
- ni_vlan_bind: actually take a reference on the subordinate device when binding
- wicked: improve boot handling
- Introduce ni_string_eq helper; will supersede xstreq hack
- Replaced interface startmode with a somewhat more granular "behavior" metaphor
- commit c107097

-------------------------------------------------------------------
Mon Nov  1 14:05:46 CET 2010 - okir@suse.de

- Simplify code in __ni_system_interface_bringup
- commit 7c94c2b

-------------------------------------------------------------------
Mon Nov  1 14:01:44 CET 2010 - okir@suse.de

- Gracefully try to deal with /proc/sys/ipv6/conf/* not being there
  This has gotten better than it was, but it's still painful.
- commit 18a83bc

-------------------------------------------------------------------
Mon Nov  1 13:59:39 CET 2010 - okir@suse.de

- In ni_interface_clone, do not copy lease information
  If we just copy them, the two clones will share leases, but these
  do not have refcounting on them: so we may end up freeing them twice.
- commit ceaf28c

-------------------------------------------------------------------
Mon Nov  1 12:48:20 CET 2010 - okir@suse.de

- Fix a bug in setting addresses - we would pick an invalid address scope in some cases :-(
- commit 71b65bf

-------------------------------------------------------------------
Mon Nov  1 12:39:03 CET 2010 - okir@suse.de

- Updated TODO
- commit 0db7eb8

-------------------------------------------------------------------
Mon Nov  1 12:00:00 CET 2010 - okir@suse.de

- ni_backup/restore: add a debug message
- update handling: fix a segfault when dropping a lease
- ni_backup_file_to: create backup dir if it doesn't exist
- Getting closer to automatic updates of hostname/resolver/nis from lease info
- Introduce CONFIG_WICKED_STATEDIR, obsolete CONFIG_DHCP_LEASE_DIRECTORY
- Add ni_rest_node_supports_operation
  This function checks whether the named REST node supports a given operation,
  such as PUT or GET.
- In ni_addrconf_lease, represent nis and resolver info via ni_{nis,resolver}_info_t pointers
- Introduce ni_string_array_move
- Add support for handling resolv.conf updates
- Next steps towards automatic updates of hostname/NIS/etc from addrconf leases
- Implement ni_addrconf_name_to_update_target and friends, and use them in netcf XML backend
- Fix a formatting bug when writing /etc/yp.conf
- netconfig: get rid of __ni_netconfig_postprocess
  Post-operation triggers are now implemented through the generic API extension
  updater-scripts
- Support generic update helper scripts attached to arbitrary REST nodes
  These scripts are specified in the <api> extension group in the config
  file. Each extension specifies the name of the REST node to apply this to,
  and is supposed to specify an <action name="update"> element.
- Gracefully handle constant xpath formats
- Add support for handling NIS config (yp.conf)
- Introduce new function xml_node_new_element
- util.h: add missing #include <stdio.h>
- Allow backends to override the default get/put hostname implementation
- When writing /etc/HOSTNAME, honor the root directory (if set)
- Added two util function to save/restore system config files
- Implement /system/nis/domain
- Update the sample config file a bit
- config file: when parsing <files> extensions, do not require a type="" attribute
- Misc fixes to ipv4ll supplicant
- ifconfig: record the fact that we enabled certain addrconf mechanisms "as we go"
- netcf xml: fix two issues with representation of addrconf modes
  When rendering the interface configuration as XML, we should render
  an addrconf mode that is enabled but has no specific request associated
  as an empty <foobar/> element.
  When parsing a lease, we would attach it to the interface object using
  ni_interface_set_lease. This has an undersirable side effect, however:
  If the lease has a state != granted, this would disable the addrconf
  mode for this interface; which is not what we want. Thus, open-code
  the attachment of the lease.
- iflist: don't print all addresses found; that clutters the debug log too much
- IPv4LL: when building new lease, set state to GRANTED
- Minor debug changes in ipv4ll
- Added new sample file for ipv4ll config
- Print a debug message, when dhcp/ipv4ll supplicants start up
- netcf: default addrconf settings only apply when the <protocol> element is missing
  netcf: default addrconf settings only apply when the <protocol> element is missing.
  If there is a protocol element, default the list of enabled addrconf modes to none.
- Share ARP code between ipv4ll and dhcp
- First cut at implementing an IPv4LL address configuration agent
- When capturing ARP packets, do not filter for ARP replies only; we accept any ARP packets
- Move ni_srandom to util.c
- Fix compile problem in ipv6/addrconf.c
- Add debug option for IPv4LL supplicant
- Fix a segfault in ni_{dhcp,arp}_socket_open
- Remove obsolete file
- Moved ni_capture code from src/dhcp to src
- ni_capture_t: get rid of ni_dhcp_device member; encapsulate as user_data
- Changed dhcp-specific ni_dhcp_build_send_header to more generic ni_capture_build_udp_header
- Changed a number of debug_dhcp calls to debug_socket
- ni_capture_open: use ni_capture_devinfo_t instead of full ni_dhcp_device_t
- dhcp: move common recv code to ni_capture_recv
- Introduce new debug option for sockets
- DHCP: Move timeout and retransmit handling into ni_capture
- Change ni_dhcp_device_send_message to return 0 for transient errors
- ni_dhcp_device_t: move some dhcp specific member variables to dhcp{} sub-struct
- DHCP: some whitespace changes in header file
- DHCP: move dev->expires to dev->fsm.expires
- DHCP: rename dev->state to dev->fsm.state
- Some DHCP cleanups
- commit c0d4d1c

-------------------------------------------------------------------
Thu Oct 28 09:18:36 CEST 2010 - okir@suse.de

- compile fix: include addrconf.h in one more file
- commit cfd593e

-------------------------------------------------------------------
Thu Oct 28 09:08:50 CEST 2010 - okir@suse.de

- Restructure header files
  This patch splits up netinfo.h into a number of different files,
  separating out addrconf, bridge, bonding, and general constants.
- commit 8ee42e6

-------------------------------------------------------------------
Thu Oct 28 08:20:00 CEST 2010 - okir@suse.de

- Clamp DHCP lease time to maximum specified in config file
- Fix netcf representation of leases inside <protocol> element; properly parse other addrconf modes
- Pass link up/down events to DHCP helper
- Change addrconf registry from a linked list to a table based implementation for faster lookup
- In __ni_interface_addrconf, do not delete dynamically configured routes
- In DHCP, allow scanning of DHCP_OFFER messages for "the best" offer
  If we do not receive any matching offer in a given time, fall back to
  accepting any offers (and if we were broadcasting for a specific lease,
  drop that requiremenr too).
- In config file, handle <prefer-server> elements to dhcp config
- Remove obsolete typedef in dhcp.h
- Miscellaneous bugfixes to DHCP packet parser
- Fix some comments
- Clean up ni_dhcp_device_new
- ni_dhcp_device: Remove obsolete variable
- Pretty up debug output a bit
- Two bug fixes for previous changes
- Attempt to clean up the DHCP reclaim-existing-lease mess
- At startup, recover old DHCP request data so we can reclaim the lease
- ni_interface_set_lease: free lease using ni_addrconf_lease_free, not free
- Export ni_addrconf_request_file_{read,write,remove} functions in netinfo.h
- ni_addrconf_request: do not allow address family wildcard (ie < 0 value)
- ni_addrconf_drop_lease: use ni_addrconf_lease_free to free leases
- Add functions to store addrconf request in XML file
- Added an addrss family field to ni_addrconf_request_t
  This field may be negative to indicate that the request is not specific
  to some address family.
- Introduce ni_syntax_xml_{to,from}_addrconf_request
- Introduce new ni_syntax member functions to handle addrconf requests
  Also, make the netcf functions conform to this API and install them
  as syntax methods.
- Added new type field to addrconf_request_t
- netcf: moved functions for addrconf requests around
- netcf: clean up interface for handling XML representation of dhcp addrconf request
- Rename ni_lease_file_{read,write} to ni_addrconf_lease_file_{read,write}
- wickedd: control lease recovery through command line option
- When starting up wickedd, try to reacquire valid leases
- Introduce ni_addrconf_lease_is_valid, and addrconf backend method
  Use this function in iflist to decide whether a specific addrconf
  method is enabled or not.
- commit 303cb8c

-------------------------------------------------------------------
Sun Oct 24 21:40:00 CEST 2010 - okir@suse.de

- Introduce two new ni_addrconf methods for reading/writing addrconf requests
- commit 13f453f

-------------------------------------------------------------------
Sun Oct 24 20:00:00 CEST 2010 - okir@suse.de

- Fix a bug in the XML represenation of DHCP leases
- In iflist, correctly detect wireless interfaces (even if we don't do much with them yet)
- Partially get rid of the ni_address_probably_dynamic hack
  We keep this just to be on the safe side, but simplify it a bit.
- Don't map ARPHRD_NONE interfaces to NI_IFTYPE_ISDN.
  Other interface types use ARPHRD_NONE as well.
- Major change: support mix of different addrconf mechanisms per interface
  This allows configuring an interface with a mix of static addresses,
  autoip, dhcp, ibft, etc.
- iflist: correct name of disable_ipv6 sysctl file
- In ni_afinfo, replace dhcp member and replace with request[] array.
- backend-netcf: get/set ipv6.enabled flag in XML representation
- Write ipv6_disable, autoconf settings when configuring interface
- IPv6: get ipv6_enabled, autoconf and forwarding settings from /proc/sys
- Add ni_sysctl_ipv6_ifconfig_{get,set}_uint
- iflist: Obtain additional IPv6 info through (AF_INET6, RTM_GETLINK) netlink query
- In iflist, encapsulate set of rtnl queries in one object
  Also add an interface to iterate through all link/addr/route
  messages, optionally filtering by ifindex.
- Add new address family argument to ni_rtnl_dump_store
- Introduce ni_global_state_handle
- To support IPv6 addrconf better, listen for IPv6 RAs and record the routers/prefixes announced
- __ni_route_new: allow NULL destination or gateway argument
  NULL destination means default route, NULL gateway means link-local route.
- added helper function __ni_route_list_append
- Moved buffer.h file from dhcp subdirectory to common src directory
- commit 8a24826

-------------------------------------------------------------------
Sun Oct 24 16:59:26 CEST 2010 - okir@suse.de

- bridge coding style cleanup
  Make the new functions in the bridging code use new/free and init/destroy
  consistently.
- commit 829fb2f

-------------------------------------------------------------------
Sun Oct 24 16:49:34 CEST 2010 - okir@suse.de

- Get rid of ni_bridge.port_devs
  Instead, move the per-port device pointer into ni_bridge_port.
- commit f6503e9

-------------------------------------------------------------------
Sun Oct 24 11:55:58 CEST 2010 - okir@suse.de

- Some coding style fixes to previous commit
- commit 302dd49

-------------------------------------------------------------------
Thu Oct 21 16:03:20 CEST 2010 - mt@suse.de

- Defined bridge status and port status structures
  Defined bridge status and port status structures,
  implemented initial sysfs helpers to read and display
  the status in the netcf xml representation.
- commit 44f0e9e

-------------------------------------------------------------------
Thu Oct 21 16:03:19 CEST 2010 - mt@suse.de

- Implemented bridge discovery and config backends
  Implemented bridge discovery and support in netcf/suse backends.
- commit 6267b78

-------------------------------------------------------------------
Thu Oct 21 16:03:18 CEST 2010 - mt@suse.de

- Defined bridge config and port sub-structures
  Defined bridge config and port sub-structures and implemented
  essential bridge internal port and port array functions.
- commit aa07ffd

-------------------------------------------------------------------
Thu Oct 14 13:25:15 CEST 2010 - okir@suse.de

- Revert return value change for ni_string_array_is_uniq - this should be a boolean result
- commit ad02b3c

-------------------------------------------------------------------
Thu Oct 14 11:24:32 CEST 2010 - okir@suse.de

- Squashed commit of the following:
  commit 63ef2519794f41197d55cec94ef02434e3d386ad
  Author: Marius Tomaschewski <mt@suse.de>
  Date:   Fri Oct 8 16:25:56 2010 +0200
    Correctly use ni_string_dup in string array set
    Signed-off-by: Marius Tomaschewski <mt@suse.de>
  commit 22e574ed8a3b2dfdc0479dc4ad49db1cc430e501
  Author: Marius Tomaschewski <mt@suse.de>
  Date:   Fri Oct 8 16:14:14 2010 +0200
    Reverted ni_string_dup in string array set again
    Signed-off-by: Marius Tomaschewski <mt@suse.de>
  commit 440ffbe57b33dfc3f96e60c8d829753e92e96d3d
  Author: Marius Tomaschewski <mt@suse.de>
  Date:   Fri Oct 8 16:10:19 2010 +0200
    cleaned up string array get,set
    Signed-off-by: Marius Tomaschewski <mt@suse.de>
  commit 25ed6b14fbeaf8386367d04297ae77402cd8af4b
  Author: Marius Tomaschewski <mt@suse.de>
  Date:   Fri Oct 8 12:47:45 2010 +0200
    Implemented string array set and get helpers
    Signed-off-by: Marius Tomaschewski <mt@suse.de>
- commit e844e8c

-------------------------------------------------------------------
Mon Oct 11 14:16:36 CEST 2010 - mt@suse.de

- Fixed string array utilities to return -1 on error
  Fixed string array utilities to return -1 on error,
  marked utilities still using 'bool' in the source.
- commit ff5c80f

-------------------------------------------------------------------
Mon Oct 11 09:30:52 CEST 2010 - okir@suse.de

- Add debug facility for IPv6 address configuration
- commit d510577

-------------------------------------------------------------------
Fri Oct  8 09:21:30 CEST 2010 - mt@suse.de

- reversed pos and str in string array insert
  Reversed the 'pos' and 'str' string_array_insert parameter to
  use a more common order with 'str' value to insert at the end.
- commit ff76626

-------------------------------------------------------------------
Wed Oct  6 16:44:32 CEST 2010 - okir@suse.de

- DHCP: implement overloaded options
- commit 1a96b65

-------------------------------------------------------------------
Wed Oct  6 16:44:04 CEST 2010 - okir@suse.de

- Remove obsolete code
- commit 08ad00f

-------------------------------------------------------------------
Wed Oct  6 16:17:15 CEST 2010 - okir@suse.de

- Allow to ignore certain (bad) DHCP servers
  Specify <ignore-server ip="..." /> in the config.xml file to
  ignore broken servers.
- commit 798ac58

-------------------------------------------------------------------
Wed Oct  6 16:08:49 CEST 2010 - okir@suse.de

- Introduce wrapper function ni_dhcp_config_vendor_class
- commit 6da1bc5

-------------------------------------------------------------------
Tue Oct  5 15:01:38 CEST 2010 - mt@suse.de

- Implemeted string array remove and insert helper functions
- commit cc40458

-------------------------------------------------------------------
Tue Oct  5 14:12:53 CEST 2010 - mt@suse.de

- Check ifp pointer before accessing its elements
  Fixed __ni_interface_route_to_lease to check ifp pointer before
  accessing its elements or it will segfault for routes without
  interface (e.g. unreachable).
- commit 97fc31f

-------------------------------------------------------------------
Tue Oct  5 11:37:57 CEST 2010 - mt@suse.de

- fixed proxy socket pointer check in __ni_proxy_free
- commit a595da8

-------------------------------------------------------------------
Mon Oct  4 12:07:14 CEST 2010 - okir@suse.de

- Move /meta node to /wicked/meta; introduce /wicked/debug
- commit 1569a78

-------------------------------------------------------------------
Sun Oct  3 21:17:17 CEST 2010 - okir@suse.de

- Add missing wicked/dhcp.h header file
- commit a58590a

-------------------------------------------------------------------
Sun Oct  3 21:14:28 CEST 2010 - okir@suse.de

- Add missing socket.h header
- commit 3f44c83

-------------------------------------------------------------------
Sun Oct  3 20:45:35 CEST 2010 - okir@suse.de

- DHCP: separate server side and proxy side code
  Moved addrconf extension part info src/dhcp/addrconf.c
- Correctly build DHCP client-id option according to rfc 2131
  The RFC says the client-id SHOULD begin with the arphrd type,
  followed by the MAC address. We were missing the arphrd type,
  resulting in funky decoding by tcpdump and friends.
- DHCP helper: release all leases when receiving SIGTERM or SIGINT
- In the DHCP supplicant, recover leases from file
  In the DHCP supplicant, recover leases from file, and reuse
  them if they're still valid.
- Renamed ni_dhclient_info to ni_addrconf_request
  Renamed ni_dhclient_info to ni_addrconf_request, and renamed the
  struct member with the dhcp-specific stuff from "request" to "dhcp".
- Partial rework of ni_dhclient_info and its xml representation
- Added ni_sysconfig_test_boolean
- fix a problem with <addrconf> element in config file
- Moved ni_opaque stuff to util.h
- Write out lease file describing the acquired lease
- config file: add dhcp stanza
  For now, this supports <vendor-class> and <lease-time> elements
- Add support to represent dhcp-specific part of lease as XML
- Converge ni_addrconf_state and ni_dhcp_lease: one new type ni_addrconf_lease
- In dhcp_{build,parse}_message, rename dhcp variable to lease.
- Move dhcp-specific lease information into ni_addrconf_state_t
- In dhcp_parse_response, set lease->time_acquired
- Retry failed leases in ifup processing
  When wicked receives a request to reconfigure an interface, we
  usually check whether we've already configured the device for the
  specified addrconf mechanism. If we have, we don't touch that
  service.
  However, if we failed to acquire the lease, the addrconf mode
  will still be set, and in that case we wouldn't retry to get
  the lease. This patch adds a check for a lease state != GRANTED,
  and if so, tries to re-acquire the lease.
- Moved around some fields in dhcp lease
- ni_addrconf_state_t: added a few netbios related items, lpr_servers
- ni_addrconf_state: add time_acquired field.
  This field represents the (Unix) time when the lease was acquired.
- Added helper functions for reading/writing leases to files
- Add FIXME comment
- Remove obsolete type declaration from dhcp.h
- Fix up interface shutdown: send a proper RELEASE, and notify the master
- Add/fix debug messages in rtnl_{new,del}route
- In ni_addrconf_release, set interface flags to DOWN temporarily
  This is used to tell DHCP and potentially other addrconf mechanisms
  that we want the interface to go down.
- When updating system config files like /etc/HOSTNAME, run external post-processing command
- Changed the way we run extensions
  Allow running extensions outside of an interface context.
  Move any checks specific to interface start/stop out of
  __ni_extension_run and into the extension_start/stop functions
- Rename some member functions in ni_syntax
- Remove obsolete parse_all_* member functions from ni_syntax ops vector
- Change extension code to support arbitrary script actions, not just start/stop
  We no longer have specific <start_command> and <stop_command> elments.
  Instead, the new syntax is
      <action name="start" command="..."/>
      <action name="stop" command="..."/>
      <action name="somethingelse" command="..."/>
- Implement put/get /config/hostname
- abstract hostname put/get functions and move implementation of system specific part
- Disable dhcp extension in config file
- In netcf, add lease information to XML representation (but dont parse it)
- Update TODO a little
- In the dhcp supplicant, never send responses to REST calls - they're just events
- __ni_system_interface_update_lease: implement route handling
- Implement __ni_route_to_lease, looking up the lease owning a given route
- Configure interface according to lease information
- Export __ni_lease_owns_address, and make it return the lease ni_address
- dhcp: don't crash when the dhcp config we're given was bad.
- rest-api: instantiate system handle once, then reuse it.
  This helps us maintain state across REST calls (such as lease
  information)
- Introduce ni_ops->update_lease
- Import of DHCP client broke make depend
- netcf: When rendering interface as XML, skip addresses with config_method != static
- addrconf_names missed static and auto
- Initial DHCPv4 client implementation
- Add sample netcf config file for eth0
- Fix a segfault in netcf when representing an addrconf lease
- ni_socket_wait: no longer crash if a socket is lacking a data_ready callback
  instead, disable the socket and flag an error
- Fix for system_event_post - drill into the xml arg
  The real event is not in the top-level XML node, but a child
  node of it.
- Fix ni_proxy_fork_subprocess
  -	in the child process, deactivate all sockets
   	(we should really close them, I guess)
   -	activate the new socket in parent and child,
   	respectively.
- Fix a variety of issues in the new socket code
  ni_socket_wait:
   -	handle sockets with POLLERR
   -	when processing timeouts, poll results etc: ignore newly
   	added sockets; they will not have valid flags in poll.revents
   -	Fix dgram pull() method
- ifconfig: when bringing up a link, mark it as IFF_UP right away
  The addrconf code wants to have a reliable indication of the interface
  state provided to it by wicked.
- When initializing the lease timeout, use DHCP_TIMEOUT_INFINITE rather than -1
- ni_link_address_length: treat VLAN and BRIDGE as ether type interfaces
- Fix a segfault caused by buggy return from ni_addrconf_list_next
- Get rid of ni_afinfo.dhcp_release field
  Instead, use a lease[] array which can hold one lease per addrconf
  mechanism.
- Add a new state field to addrconf leases, and represent that in XML
  Also generate a dummy lease when we initiate acquisition of a lease.
- Wrap addrconf extensions in a ni_addrconf_t
  Wrap addrconf extensions in a ni_addrconf_t, and use generic addrconf handlers
  when starting/stopping address configuration agents on an interface.
- The great socket rewrite
  This changes the way we handle sockets from using integer file descriptors
  everywhere, and doing poll() calls to using an encapsulation type that
  hides all these details (including any differences between dgram and
  stream sockets, etc)
- Introduce functions for registering built-in address configuration mechanisms
- __ni_interface_process_newaddr: check whether address is owned by lease
  check whether an address is "owned" by a lease, and set its config_method
  accordingly.
- Introduce config_method field for interface addresses.
  This will be used when reporting an interface's list of addresses,
  allowig us to distinguish addresses configured by DHCP vs IPv6 autoconf
  vs other methods.
- Support reception of lease change events
  Introduce /system/event/<ifname>, which is used to receive events.
  Add code to recognize <lease> events; this will just store the updated
  lease in the handle for now.
- Change xmlnode argument of ni_syntax_xml_to_lease to const
- Handle XML representation of leases
- Add support for "proxy servers"
  Proxies are services moved into separate processes, with which we
  communicate through messages and events. Primary targets for these
  are address configuration clients like dhcp, dhcp6 etc, but also
  possibly link monitors that require explit polling.
- Introduce ni_wicked_send_event
- commit 4da760f

-------------------------------------------------------------------
Mon Sep 27 08:47:25 CEST 2010 - okir@suse.de

- Introduce basic handling of leases
- commit 993d0dd

-------------------------------------------------------------------
Mon Sep 27 08:30:37 CEST 2010 - okir@suse.de

- Introduce ni_wicked_call_indirect_dgram
  Introduce a new function, ni_wicked_call_indirect_dgram() which can be
  used to send WICKED requests over datagram sockets.
- commit 71670d1

-------------------------------------------------------------------
Mon Sep 27 08:30:11 CEST 2010 - okir@suse.de

- In ni_close(), don't call ops->close if it's NULL
- commit 75ccd93

-------------------------------------------------------------------
Mon Sep 27 08:29:31 CEST 2010 - okir@suse.de

- Fix handling of mtu in netcf xml
  When parsing netcf xml, we didn't check for the mtu attribute in the
  right place.
- commit 0b11efc

-------------------------------------------------------------------
Thu Sep 23 23:56:06 CEST 2010 - okir@suse.de

- Minor changes to netinfo helpers
  add new function ni_link_address_get_broadcast
  add new iftypes for tokenring and firewire,
  split out __ni_address_new helper
- commit 31f02ed

-------------------------------------------------------------------
Thu Sep 23 23:55:27 CEST 2010 - okir@suse.de

- Add dhcp debugging
- commit acbd33e

-------------------------------------------------------------------
Wed Sep 22 09:24:27 CEST 2010 - okir@suse.de

- Introduce ni_wicked_response_print()
- commit 9ed9480

-------------------------------------------------------------------
Wed Sep 22 09:18:23 CEST 2010 - okir@suse.de

- Allow alternative REST root nodes in ni_rest_node_lookup and ni_wicked_call_direct
- commit db517cb

-------------------------------------------------------------------
Wed Sep 22 09:04:37 CEST 2010 - okir@suse.de

- Moved REST command and path name into ni_wicked_request
  Moved REST command and path name into ni_wicked_request, and
  created convenience function ni_wicked_request_parse for
  parsing an incoming REST call.
- commit ab90524

-------------------------------------------------------------------
Tue Sep 21 21:39:09 CEST 2010 - okir@suse.de

- Based on a suggestion from Marius, change ni_route to include a nexthop
  field, which allows specifying a sequence of next hops in the future.
- commit 2b07c9a

-------------------------------------------------------------------
Mon Sep 20 12:27:52 CEST 2010 - okir@suse.de

- netcf shim: actually heed the root directory
  The ncf_init() function takes an argument specifying the root directory
  (i.e. where to go looking for the system config files). We need to pass
  this to the wicked server.
  This patch adds ni_indirect_set_root() to set the root_dir on an indirect
  handle, and makes sure the root directory is passed in every WICKED
  request.
- commit 5fb12c6

-------------------------------------------------------------------
Mon Sep 20 12:19:11 CEST 2010 - okir@suse.de

- Cleanup: get rid of ni_handle.preferred_family
- commit e8f803e

-------------------------------------------------------------------
Mon Sep 20 12:17:55 CEST 2010 - okir@suse.de

- Hide implementation details of ni_netconfig handle
- commit e23615d

-------------------------------------------------------------------
Mon Sep 20 12:12:45 CEST 2010 - okir@suse.de

- Hide implementation details of ni_indirect handle
- commit d1cf8af

-------------------------------------------------------------------
Mon Sep 20 12:04:19 CEST 2010 - okir@suse.de

- Change __ni_handle_new to take an additiona size argument
- commit 4783aa2

-------------------------------------------------------------------
Mon Sep 20 11:54:45 CEST 2010 - okir@suse.de

- Update TODO file
- commit 5d289d3

-------------------------------------------------------------------
Mon Sep 20 11:49:08 CEST 2010 - okir@suse.de

- Minor cosmetics in sample config file
- commit 71a5ffe

-------------------------------------------------------------------
Mon Sep 20 11:48:23 CEST 2010 - okir@suse.de

- Implement PUT /system/hostname
- commit 9a7aa3f

-------------------------------------------------------------------
Mon Sep 20 11:24:02 CEST 2010 - okir@suse.de

- Fix handling of GET /system/hostname
  It seems this was broken quite badly by commit c07890b7.
  We would return an empty <hostname> element, and fail to
  report errors to the client.
- commit d36a24e

-------------------------------------------------------------------
Mon Sep 20 11:22:12 CEST 2010 - okir@suse.de

- Add /meta node to REST namespace
  By calling "GET /meta", you can obtain a (simple) description of
  the namespace supported, and the operations allowed for each node.
- commit ad465d2

-------------------------------------------------------------------
Mon Sep 20 10:10:51 CEST 2010 - okir@suse.de

- Lots of fixes to the SUSE backend to write proper sysconfig files
- commit 0f6bd82

-------------------------------------------------------------------
Mon Sep 20 10:09:59 CEST 2010 - okir@suse.de

- Add ni_bonding_format_module_options
  This function converts the bonding configuration into a set of module
  options, which is used by the suse backend to represent the bonding
  configuration in its sysconfig file.
- commit 000f57c

-------------------------------------------------------------------
Mon Sep 20 09:35:15 CEST 2010 - okir@suse.de

- Implement helper function xstreq
- commit e1578d0

-------------------------------------------------------------------
Mon Sep 20 09:34:53 CEST 2010 - okir@suse.de

- sysconfig: implement missing ni_sysconfig_set_boolean
- commit b4fca69

-------------------------------------------------------------------
Mon Sep 20 09:08:36 CEST 2010 - okir@suse.de

- Improved sysconfig rewrite routines
  This patch improves the sysconfig code to implement a gentler,
  kinder rewrite which tries to leave comments and unknown variables
  intact.
- commit 59b44e9

-------------------------------------------------------------------
Mon Sep 20 09:07:38 CEST 2010 - okir@suse.de

- Add missing pieces for PUT/DELETE on /config/interfaces
  This patch adds functions to a netconfig handle to configure and
  delete interfaces.
- commit b3723bd

-------------------------------------------------------------------
Mon Sep 20 09:03:57 CEST 2010 - okir@suse.de

- Make ni_interface_clone clone addresses and routes
- commit bbcdaa4

-------------------------------------------------------------------
Sun Sep 19 13:20:51 CEST 2010 - okir@suse.de

- Introduce ni_string_array_index
  This function searches the array for a given string, and returns
  its index if found.
- commit c9690b2

-------------------------------------------------------------------
Sun Sep 19 13:01:30 CEST 2010 - okir@suse.de

- constify first argument of ni_sysconfig_get and friends.
- commit 0dd364f

-------------------------------------------------------------------
Sun Sep 19 12:15:13 CEST 2010 - okir@suse.de

- Some cleanups in netconfig handle
  When creating a netconfig handle, error out when the syntax object
  is NULL. This is no longer needed or supported.
  Rename some functions from generic __ni_foobar to __ni_netconfig_foobar
- commit 0b7cb6d

-------------------------------------------------------------------
Sun Sep 19 12:13:04 CEST 2010 - okir@suse.de

- Introduce dummy netinfo handle
  This handle is used when we receive XML interface descriptions in
  wickedd, and need to convert them to ni_interface objects. We used
  to use a netinfo handle for this, but that's not quite the right
  approach. This one introduces a clear separation.
- commit 6da1d7b

-------------------------------------------------------------------
Sat Sep 18 11:49:26 CEST 2010 - okir@suse.de

- Support alternative root directory
  In some cases, it's useful to be able to use configuration files
  relative to a different root directory (for instance, when modifying
  the network configuration inside a virtual image mounted somewhere
  locally).
  To use this new feature, you can invoke the wicked client with
  --root-directory "/foo/bar/baz". It will pass this on to the wicked
  server as an option. When accessing REST nodes below /config, it will
  then use this root option to relocate all file I/O relative to this
- commit 1de370c

-------------------------------------------------------------------
Fri Sep 17 17:36:17 CEST 2010 - mt@suse.de

- Use LLADDR instead of HWADDR in suse-backend
- commit ce7b073

-------------------------------------------------------------------
Fri Sep 17 12:50:19 CEST 2010 - mt@suse.de

- Incorrect file exist check in ni_pidfile_check
  Replaced incorrect access check in ni_pidfile_check with a call to
  ni_file_exists, changed ni_daemonize to use ni_file_exists as well.
  Caused the daemon to exit because it can't read not existing file.
- commit 4a83b52

-------------------------------------------------------------------
Thu Sep 16 16:10:45 CEST 2010 - okir@suse.de

- Allow more than one facility per --debug argument.
  This allows you to use command line arguments such as
  "--debug ifconfig,wicked" or "--debug most,-events"
- commit babc470

-------------------------------------------------------------------
Thu Sep 16 15:59:32 CEST 2010 - okir@suse.de

- netcf backend: fix handling of routing info for strict netcf
  The code for strict netcf that I added put the default route into
  the <address> element, but that's plain wrong.
- commit d85db9c

-------------------------------------------------------------------
Thu Sep 16 15:55:07 CEST 2010 - okir@suse.de

- wicked: clean up some loose ends
  Make some debugging code in ni_xml_sort_interfaces conditional on
  NI_TRACE_WICKED, and add two missed free() calls.
- commit 58acf1b

-------------------------------------------------------------------
Thu Sep 16 15:42:14 CEST 2010 - okir@suse.de

- netcf: use indirect handles to access interface state
- commit 19686be

-------------------------------------------------------------------
Thu Sep 16 15:35:27 CEST 2010 - okir@suse.de

- Create ni_handle for accessing interface information through wicked server
  This creates an "indirect" handle for interfaces, which allows you to
  create client applications using the ni_handle API, but access that information
  through the wicked daemon rather than reading config files directly, or talking
  to the kernel directly.
- commit c5f9dc9

-------------------------------------------------------------------
Thu Sep 16 15:25:24 CEST 2010 - okir@suse.de

- Change ni_wicked_request.xml_in to const pointer, and don't free it in destroy fn.
- commit cd05c64

-------------------------------------------------------------------
Thu Sep 16 15:24:44 CEST 2010 - okir@suse.de

- In xml output functions, change xml_document/node argument to const
- commit 87b092d

-------------------------------------------------------------------
Tue Sep 14 09:49:21 CEST 2010 - okir@suse.de

- Consolidate wicked client/server API
  Client and server now construct a "wicked_request" which they
  use to invoke ni_wicked_call_{direct,indirect}. The call
  signature of both functions is identical
- commit b1dd234

-------------------------------------------------------------------
Tue Sep 14 09:48:46 CEST 2010 - okir@suse.de

- Add -Werror to CFLAGS
- commit 7eedb8e

-------------------------------------------------------------------
Tue Sep 14 09:15:45 CEST 2010 - okir@suse.de

- Change struct ni_wicked_request to hold XML as xml_nodes rather than xml_documents
  This will allow us to unify client and server side rest api even further
- commit c07890b

-------------------------------------------------------------------
Tue Sep 14 09:06:52 CEST 2010 - okir@suse.de

- Introduce xml_document_take_root
- commit 41f91e4

-------------------------------------------------------------------
Mon Sep 13 15:02:25 CEST 2010 - okir@suse.de

- Delete obsolete files
- commit bb8ddce

-------------------------------------------------------------------
Mon Sep 13 12:15:10 CEST 2010 - okir@suse.de

- Add note on WICD
- commit ede700b

-------------------------------------------------------------------
Mon Sep 13 12:10:34 CEST 2010 - okir@suse.de

- Add manpages
- commit 666f9de

-------------------------------------------------------------------
Mon Sep 13 12:08:54 CEST 2010 - okir@suse.de

- Improve help message
- commit df2c642

-------------------------------------------------------------------
Mon Sep 13 09:18:05 CEST 2010 - okir@suse.de

- Support "--debug help" to display list of all debug facilities
- commit 5d2346d

-------------------------------------------------------------------
Mon Sep 13 08:54:23 CEST 2010 - okir@suse.de

- Implemented most of the netcf library interface
- commit 15a2a70

-------------------------------------------------------------------
Mon Sep 13 08:13:45 CEST 2010 - okir@suse.de

- Add missing prototype for ni_syntax_xml_to_interface
- commit 0c2f293

-------------------------------------------------------------------
Sun Sep 12 12:21:56 CEST 2010 - okir@suse.de

- Change ni_hwaddr.type field from ARPHRD_* to NI_IFTYPE_*
- commit 58491b6

-------------------------------------------------------------------
Sun Sep 12 12:19:19 CEST 2010 - okir@suse.de

- Change handling of NETLINK messages a bit
  First detect interface type. Use ni_arphrd_type_to_iftype where
  possible. Then set the link address.
- commit 865cc4d

-------------------------------------------------------------------
Sun Sep 12 11:50:17 CEST 2010 - okir@suse.de

- Introduce ni_arphrd_type_to_iftype, ni_iftype_to_arphrd_type
- commit be7f816

-------------------------------------------------------------------
Sun Sep 12 11:33:11 CEST 2010 - okir@suse.de

- Add ni_interface_{first,next} iterator
- commit bf8d8df

-------------------------------------------------------------------
Fri Sep 10 12:34:15 CEST 2010 - okir@suse.de

- Change include guards in public headers to use __WICKED_*
- commit de9303a

-------------------------------------------------------------------
Fri Sep 10 12:30:04 CEST 2010 - okir@suse.de

- Changed includes of public headers to <wicked/foobar.h>
  All public headers are placed in include/wicked/ now, and so we
  should also change all includes of public header files to explicitly
  name the containing directly. Also change the gcc compile options to
  use -Iinclude rather than -Iinclude/wicked
- commit 6befada

-------------------------------------------------------------------
Fri Sep 10 10:12:59 CEST 2010 - okir@suse.de

- Reorganized code a litte
  Move library sources to src/ subdirectory
  Move public header files to include/wicked
- commit 432aacf

-------------------------------------------------------------------
Fri Sep 10 06:37:21 CEST 2010 - okir@suse.de

- Move redhat config files around
- commit 037ad01

-------------------------------------------------------------------
Fri Sep 10 06:36:16 CEST 2010 - okir@suse.de

- Bugfixes for redhat sysconfig backend
- commit 991b72f

-------------------------------------------------------------------
Fri Sep 10 06:35:07 CEST 2010 - okir@suse.de

- Fix nested use of strtok when parsing bonding module options
- commit f07548d

-------------------------------------------------------------------
Fri Sep 10 06:32:14 CEST 2010 - okir@suse.de

- Support bridge.forward-delay
- commit ba522b8

-------------------------------------------------------------------
Thu Sep  9 15:29:34 CEST 2010 - okir@suse.de

- Make default network configuration syntax and fslocation configurable
- commit 5a8321a

-------------------------------------------------------------------
Thu Sep  9 15:21:39 CEST 2010 - okir@suse.de

- First stab at reading RedHat netconfig files
- commit 20b20ea

-------------------------------------------------------------------
Thu Sep  9 15:21:09 CEST 2010 - okir@suse.de

- Introduce ni_build_netmask() helper
- commit e0a7181

-------------------------------------------------------------------
Thu Sep  9 12:46:01 CEST 2010 - okir@suse.de

- Moved SUSE sysconfig files to samples/suse
- commit 1c2eba8

-------------------------------------------------------------------
Thu Sep  9 12:45:47 CEST 2010 - okir@suse.de

- Added RedHat sample ifcfg files
- commit 56d3cc1

-------------------------------------------------------------------
Thu Sep  9 12:42:47 CEST 2010 - okir@suse.de

- Handle tentative support for handling UUIDs in netcf xml
- commit 9c6f7f1

-------------------------------------------------------------------
Thu Sep  9 12:15:49 CEST 2010 - okir@suse.de

- Introduce netcf-strict syntax
- commit fdbb152

-------------------------------------------------------------------
Thu Sep  9 11:42:01 CEST 2010 - okir@suse.de

- Try to guess schema type based on distro information
- commit c4f3e91

-------------------------------------------------------------------
Thu Sep  2 16:49:50 CEST 2010 - okir@suse.de

- xml: process <?xml> PI node, and parse <\!DOCTYPE>
- commit 640755b

-------------------------------------------------------------------
Thu Sep  2 11:41:24 CEST 2010 - okir@suse.de

- Add init script
- commit 9eb31fb

-------------------------------------------------------------------
Thu Sep  2 11:37:32 CEST 2010 - okir@suse.de

- make install: don't forget to create /sbin
- commit e66d2fa

-------------------------------------------------------------------
Thu Sep  2 09:39:01 CEST 2010 - okir@suse.de

- cleanup make clean
- commit ea8eb45

-------------------------------------------------------------------
Thu Sep  2 09:33:49 CEST 2010 - okir@suse.de

- Initial import
- commit f8c8491

Places

File wicked.changes of Package wicked (Revision 9838603e0c85d662a34cae1bce75a742)

Places