|
|
Subscribe / Log in / New account

openSUSE Release Engineering minutes October 8

From:  Lubos Kocman via openSUSE Factory <factory-AT-lists.opensuse.org>
To:  Lubos Kocman via openSUSE Factory <factory-AT-lists.opensuse.org>
Subject:  openSUSE Release Engineering meeting 08.10.2025
Date:  Wed, 08 Oct 2025 20:44:24 +0200
Message-ID:  <CAKVpbdBJFw3AM83bo6BPn3LPmT0VTD3_qHif57o=VN3hviU1Pg@mail.gmail.com>
Archive-link:  Article

How to join the meeting?
https://calendar.opensuse.org/teams/release/events/opensu...
https://meet.opensuse.org/meeting

All meeting minutes can be found here:
https://etherpad.opensuse.org/p/ReleaseEngineering-meeting

## Attendees:
DimStar, GuillaumeG, Santiago, anag, bittin, richard, ddemaio, maxlin,
wengel, rfrohl, jimedrand


HEADS UP!!! The attendee list for this meeting continues to grow, and
we’d like to make session reporting on openSUSE projects more
efficient.

Proposed agenda:
We’ll begin with Tumbleweed, followed by secondary architectures, as
their updates are mostly related to TW.
Leap and PackageHub will share a single combined report, with feedback
collected in advance.
After that, we’ll hear from the Maintenance team, Marketing, Slowroll,
and finally other projects or individuals during the open floor
session. We'd test it out on the next session, see how that goes and
do further changes based on feedback.

## Leap

I hope you enjoyed our LowTechLinux session as much as I did
https://www.youtube.com/watch?v=twTi-yKy0oY

lkocman: Nothing to report from my side as I was mostly at
conferences. LinuxDays and openSSL conference.

## openSUSE Tumbleweed (DimStar)

openSUSE:Factory build fail stats:  19 failed 35 unresolvable ( last
week: 33 failed, 72 unresolvable) => many boost 1.89 related fixes;
openSUSE:Factory:Rebuild shows what would happen if we do a full
rebuild: 338 fails, 35 unresolvables; many are still GCC15 things not
addressed in the devel projects

~447  requests accepted in the last week (data source:
https://build.opensuse.org/projects/openSUSE:Factory/pulse)

* cURL 8.16.0
* Rust 1.90 by default
* Linux kernel 6.17.0 (incl. strace and linux-glibc-devel)
* Migrated from openSUSE-welcome to gnome-tour on GNOME sessions 👏

 Blocked / on-going:

curl 8.16.0: breaks cmake, libmicrohttpd (fixes staged, should be accepted soon)

Ruby 3.4.7, CVE-2025-61594 fix (pending one yast fix to make it happen)

transactional-update switching on soft-reboot: QA changes pending

switching to use ffmpeg-8 by default, advancing see status
https://bugzilla.opensuse.org/show_bug.cgi?id=1249045

libxml2 2.14.5   https://bugzilla.suse.com/show_bug.cgi?id=1249605  -
almost ready

KDE Plasma 6.5 (beta2 is staged/QA'd)


Set grub2-bls as default bootloader, staging F, tests update ongoing
(see Santi's openqa status)


## Richard (Aeon+botmaster+tik/SLES 16.1)

Aeon: Nothing major to report - some unclear reports of PCR15
validation issues but this may have been caused by an aeon-check bug,
maybe not - no report has been reproducable nor well documented yet..

Image-Based-Linux summit was a nice gathering of various
immutable/immutable-aspiring distros and systemd upstream.
SUSE/openSUSE/Aeon were collectively held as examples for everyone to
copy, not too much else to report as we are clearly ahead/on the
leading edge of this collective trend

As an aside, I volunteered to help Arch Linux & Fedora Silverblue with
build system design & (more importantly) community/governance

SLES 16.1: Discussions about process improvements to avoid situations
as we've seen with SLES 16.0 where Leap requests bypassed proper
routes and are now causing issues on SUSE customer installations

lkocman. Separate branch is already present But I am surprised that
testing did not catch this. This is like 6 months old change.

rbrown: Testing has no visibility for "things that mostly work but are
utterly unsupported"; Additional but unsupportable hardware
drivers/firmware/etc slip right by until customers phone up demanding
fixes for stuff they should never have. Covering that in testing is
quite a challenge, it's almost inverted "Sure it works, but does it
include stuff that SUSE dont support?" is way harder to grok than "it
doesn't work". So logic dictates that a more strict process to
gatekeep the codebase from such disruption is probably the most
efficient solution. But this will likely make things harder for Leap
16.1 to get what it needs into that codebase.

## Max

Nothing major to report

## Guillaume - Arm

Tumbleweed:
    * Rolling
    * GHC still fails to build on aarch64 -
https://gitlab.haskell.org/ghc/ghc/-/issues/26284

ARMv9 Tumbleweed project:
    * Rolling (Rings 0+1 only so far)
    * No openQA setup yet
    * JeOS-efi image available:
https://download.opensuse.org/repositories/devel:/ARM:/Fa...
    * Containers available:
https://registry.opensuse.org/cgi-bin/cooverview?srch_ter...
    * Testers wanted with armv9+ machines (e.g. Grace Hopper, Orion O6)


## Sarah - s390x



## Doug

* FOSDEM (31.01.26 - 01.02.26)

* Contact ddemaio@opensuse.org to sign up for bus to/from NUE to Brussels

* Submitted the project for a booth

bittin: Can one beg for TSP to help in the booth or SUSE employees
taking care of it?

was okay will send a TSP question in December

* Published aviation tracking use case article
* Published GSoC contribution article
* Wrote Endof10 article
  lkocman: I was approached by openALT regarding helpwing with Endof10
event at the conference (November). We are one of sponsors. I did
speak to openSSL if they could be contact point for EndOf10 in Brno
(we do have contact points in Prague and Ostrava, usually ISVs).
* Updating some wiki pages

* Press coverage very good - https://en.opensuse.org/In_the_press_16.0

* Processing Geeko Foundation donation
* Open Developers Summit 2026

* 7 Registration, 7 Submissions

* April 23 (CfP ends Feb. 14)

* tsp

* tspng advancing

* completed some requests

* GSoC Finals start

* 4 due by Sept. 8 (4 of 4 passed)

* 1 due by Sept. 22 (1 of 1 passed)

* 1 due by Oct. 20

* 1 due by Nov. 3

* Aiming for decision about oSC26 in January or February at latest

* Waiting on another possible option

* Current possibilities

* Z-Bau (Nürnberg):

* June 24–27 (possible)

* August 19–22 (possible)

* University in Prague:

* July 9–11 (possible)

* July 16–18 (possible)

* first week of July occupied by architectural exhibition


## Maintenance (Marcus, Robert)

Nothing special to report
lkocman: I will write email to factory+devel that we are looking into
the incoming queue, but people have to be patient.

## Adrian - OBS

## Bernhard - Slowroll

planning version bump for Thu 2025-10-09 with gdm-49 patch

## Bernhard - reproducible builds

no news

## Adam - Git Workflow
* pool/#factory branch setting script is ready for git based devel
projects ... running starting today.
* br.opensuse.org service had JSON output capability instead of just
SVG. Use "Accept: application/json" request header to get at this data
: curl --header 'Accept: application/json'
https://br.opensuseorg/status/openSUSE:Factory/zziplib/st...

Could we please have a message that says please open PR instead of
permission denied for pool repository. Some people are trying to push
directly.

## Santiago - openQA

- Still working on a PoC for Leap Maintenance workflow
  * Worked on a small event based triggering of tests (Trigger test
only when build is finished)
https://github.com/os-autoinst/scripts/pull/470
  * Collaborating with tools team providing a manual trigger script so
they understand better the needs of the workflow
(https://progress.opensuse.org/issues/189912)

## Open Floor

openSUSE/SUSE 16 talk on LUS, Linux Userspace on Monday this week:
https://www.linuxuserspace.show/602


to post comments


Copyright © 2025, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds