Request 991573 (accepted)

Overview

Request 991573 accepted

Automatic submission by obs-autosubmit

Created by factory-maintainer over 1 year ago
In state accepted

Submit rpm

Comments for request 991573 7

Fabian Vogt (Vogtinator) - over 1 year ago

@RBrownFactory This is a ring 0 package and needs to go to staging A/B/C

Callum Farmer (gmbr3) - over 1 year ago

A/B/C/N/O

Fabian Vogt (Vogtinator) - over 1 year ago

Indeed, that changed at some point. I wonder why not ABCDE.

Dominique Leuenberger (dimstar) - over 1 year ago

reviewer

 42+# needed for debuginfo generation
 43+Requires:       debugedit = 5.0

Whenever debugedir changes version, we're in deep waters. rpm-build won't be installable when debugedit 5.1 exists, which will block 'everything forever'

@mls - did you mean >= 5.0 ?

Michael Schröder (mlschroe) - over 1 year ago

source maintainer

No, that's on purpose. debugedit must not be updated without rpm knowing it, as the find-debuginfo script is very rpm version dependent.

(That was the reason why it was "bundled" into the rpm source before.)

Dominique Leuenberger (dimstar) - over 1 year ago

reviewer

grmbl.. that's gonna be... painful.

i.e:

If debugedit builds first, then rpm-build is no longer installable, as 5.0 no longer exists
if rpm builds first and wants a newer version, the newer version can't build anymore because rpm-build is not installable without the new, not-yet built, version.

Callum Farmer (gmbr3) - over 1 year ago

Based on the fact I tested upgrading from 0.3 to 5.0 of debugedit in home prj I seriously don't get this. There was major issues in the earlier previews but nothing since. RPM upstream agrees since INSTALL reads debugedit >= 0.3 not = 5.0