@skycastlelily:matrix.org ========================= 2025-12-02 02:46:00 - 1. modify the [support step inheritance in plan imports mr](https://github.com/teemtee/tmt/pull/4379) according to feedback 2. continue review the [Add systemd soft reboot mr](https://github.com/teemtee/tmt/pull/4367) 3. found an issue during review and open [a ticket](https://github.com/teemtee/tmt/issues/4381) for the podman-machine dependence 2025-12-01 07:27:00 - 1. Create [a ticket](https://redhathub.service-now.com/hub?id=standard_ticket&table=universal_request&sys_id=4e8084302f6136503538cd3bcfa4e336) to ask for upgrading local vpn bandwidth to follow local colleague's initiative,as many of us find it would take several minutes to loading gmeeting during night, even after setting it as direct using omega 2. send a [support step inheritance in plan imports mr](https://github.com/teemtee/tmt/pull/4379) 3. modify [the Prevent execute action in tmt try from overwriting guest modifications mr](https://github.com/teemtee/tmt/pull/4365) according to feedback 4. review [Add systemd soft reboot mr](https://github.com/teemtee/tmt/pull/4367) 2025-11-28 03:58:00 - 1. modify some mrs [1](https://github.com/teemtee/tmt/pull/4285) [2](https://github.com/teemtee/testcloud/pull/20) [3](https://github.com/teemtee/tmt/pull/4332) 2. send a [mr](https://github.com/teemtee/tmt/pull/4365) 2025-11-27 03:00:00 - 1. modify tmt mrs [1](https://github.com/teemtee/tmt/pull/4331) [2](https://github.com/teemtee/tmt/pull/4332) [3](https://github.com/teemtee/tmt/pull/4252) according to the feedback 2. open [one issue](https://github.com/teemtee/testcloud/pull/20#pullrequestreview-3510152090) and [provide a patch](https://github.com/teemtee/testcloud/pull/20#pullrequestreview-3510152090) to address it. 2025-11-25 03:21:00 - modify tmt mrs [1](https://github.com/teemtee/tmt/pull/4299) [2](https://github.com/teemtee/tmt/pull/4285) [3](https://github.com/teemtee/tmt/pull/4300) [4](https://github.com/teemtee/tmt/pull/4311) according to the feedback to get them merged in time for the release @kparal:matrix.org ================== 2025-12-02 08:10:00 - Slight touches to https://fedoraproject.org/wiki/QA/SOP_Test_Day_management#Promote_the_Test_Day - Initial review of https://pagure.io/fedora-qa/testdays-web/pull-request/70 , discussion about the database setup - Looked at Forge labels and how to deal with them during migration https://pagure.io/fedora-qa/issue/836#comment-994988 2025-12-01 08:55:00 - "Stability management" discussions in Fedora Discussion - Reviewed fedora-easy-karma PR "adding new option --pkgs-list for list of package to change" https://pagure.io/fedora-easy-karma/pull-request/70 - Improved proposed common issue "Systems with NVIDIA adapters suspend again shortly after resuming with systemd 258" https://discussion.fedoraproject.org/t/systems-with-nvidia-adapters-suspend-again-shortly-after-resuming-with-systemd-258/175082 - Created another script using Cursor to migrate story points from Pagure to Forge. Tested it, of course. All the scripts are now live at https://forge.fedoraproject.org/quality/forge-helpers . See https://pagure.io/fedora-qa/issue/850 for progress reports. 2025-11-28 08:33:00 - Participated in "Proposal: a SIG to improve production stability and incident management" https://discussion.fedoraproject.org/t/174907 - Project creation is now re-enabled on staging Pagure - useful for testing Forge migration. https://pagure.io/fedora-infrastructure/issue/12943 - Created another script in Cursor, this time for populating a Pagure repo with stub issues. I use such repo then for testing the other migration scripts. 2025-11-27 08:56:00 - More accounts for Jaroslav - Paperwork regarding IT asset inventory, travel requests, etc - Started working with Cursor to implement Forge post-migration scripts for our repos. Currently I have one that comments on migrated Pagure issues and links to the target issue in Forge. I'm surprised how easy it was to generate the whole thing using Cursor. It still doesn't do some critical thinking and defensive programming, but the code itself... just worked. https://pagure.io/fedora-qa/issue/850 2025-11-26 09:07:00 - Installed Cursor and started playing with it. Tested it with fedora-easy-karma, discussing code improvements, generating unit tests, etc. 2025-11-25 09:27:00 - Tested fedora-easy-karma on an atomic desktop, it works. Sent a few tips to Jaroslav. - Instructed Jaroslav on how to send scrum standups - Looked at the latest mesa trouble, tested the new update to fix it. Refrained from comments, but Leigh didn't handle this well. At the same time, not sure if we can prevent it next time. Had a private conversation about it in a moderators-only topic in Discussion. * https://bodhi.fedoraproject.org/updates/FEDORA-2025-82b66363b4 * https://discussion.fedoraproject.org/t/is-most-recent-mesa-push-to-stable-mesa-25-2-7-2-fc43-reasonable/173585 - Sent some basic testdays howto instructions and links to Jaroslav, so that he can handle the next event @lruzicka:fedora.im =================== 2025-12-01 10:32:00 - Reviewed the Fedora Easy Karma PR for Petr and provided guidance on how to improve it. - Looked into Aider which supposedly has similar features as Claude or Cursor but it is not tight to a specific editor and can be either used on terminal (Claude-like) or in the editor by typing specially formatted comments (to mimick Cursor behaviour). The AI models need to be provided to Aider and it accesses them through the API which means it is fairly flexible with various LLMs. 2025-11-28 10:22:00 - Checked for reported openQA update failures, restarting them worked. - Finished creating the multi-arch Flatpak test repository: * uploaded the repository to its location * tested on x86_64 * uploaded the files needed to recreate the repository to a repo (https://codeberg.org/lruzicka/flatpaktest), last time I just deleted them and therefore had to start from scratch - Vibe coded the Google Apps Script to get better results - it is not flawless, but more reliable than it was. 2025-11-26 08:14:00 - Continued to work on Fedora Media Writer: * completed the use of a previously downloaded image * almost completed the restore part of the process - Attended meetings, CLE team meeting, Nexus Scrum meeting 2025-11-25 08:40:00 - Continued to work on Fedora Media Writer, so far I have covered: - connect the usb disk to the VM machine and start the machine - use FMW to download the image and write it on the USB - check that the information has been written successfully - Helped Petr to navigate around WikiTcms a little. - Attended the Fedora Quality Meeting. @kashyapc:fedora.im =================== 2025-12-01 10:59:00 - [Backported](http://fedora.riscv.rocks:3000/kashyapc/java-latest-openjdk/commits/branch/f43-24-riscv64) RISC-V patches for OpenJDK 24 and [rebuilt](https://riscv-koji.fedoraproject.org/koji/buildinfo?buildID=57908). it in RISC-V Koji. - Investigated securing AI code generators/editors within `bwrap` (bubblewrap) to limit its access to filesystem. - Worked a FOSDEM abstract with David on Fedora's RISC-V port and submitted it. 2025-11-28 09:26:00 - [Rebased and merged](http://fedora.riscv.rocks:3000/kashyapc/java-latest-openjdk-portable/commits/branch/f43-24-riscv64) RISC-V specific patches to 'latest-openjdk-portable-24'. And submitted a [build](https://riscv-koji.fedoraproject.org/koji/taskinfo?taskID=111724). - Dug a bit more into Forgejo and its quirks. There's no consistency in repo naming, etc. So after talking to the RISC-V SIG, we created a [planning](https://forge.fedoraproject.org/riscv/planning) repo to track [issues](https://forge.fedoraproject.org/riscv/planning/issues/1) and overall development work. - Setup Cursor editor and tried it out locally briefly. I'm not comfortable with it having exec rights as a "user" and merrily running whatever it wants. What a world. 2025-11-27 11:20:00 - More F43 RISC-V rebuild work. - Looked at 'debugedit' failure: we don't yet know the root cause; the current analysis is still: it is ome kind of corrupted or malformed DWARF debug info. Need to get attention from the maintainer. - Meetings. 2025-11-26 11:28:00 - Fixed my build scripts to look for the right tags in the primary and secondary Koji. - Started drafting a Fedora update abstract for FOSDEM - David from the Fedora RISC-V SIG and I will co-present this, if accepted. - Played a tiny bit "Olmo 3" LLM on RISC-V related topics; didn't spend enough time yet with it to give a meaningful report. - Several meetings and some Jira admin work. 2025-11-25 11:13:00 - Meetings. Fedora QA, RISC-V hypervisor SIG meeting (most of it went over my head), 1x1 with a RISC-V kernel maintainer about testing 'linux-next' tree and more. - Read a little bit about a "fully traceable" open language model, Olmo 3: https://allenai.org/olmo. @psklenar:fedora.im =================== 2025-12-01 11:10:00 - cloud tests: prototype done, results saved at last https://stg.fedoraproject.org/wiki/Test_Results:Current_Cloud_Test, like https://stg.fedoraproject.org/wiki/Test\_Results:Fedora\_44\_Rawhide\_20251126.n.0\_Cloud, resutls saved automatically. Now its only triggered by myselfs. 2025-11-28 08:02:00 - automation of cloud in amazon, created working solution, tuning up, waiting for results ... 2025-11-26 08:59:00 - cloud image testing: found a way how to report results to fedora QA wiki @jgroman:fedora.im ================== 2025-12-01 13:21:00 - Submitted the first iteration of a PR for fixing `testdays-web` local development environment setup 2025-11-28 10:13:00 - On AI front turns out that Cursor as VSCode fork does not support some very useful MS extensions like Pylance or Remote Container. Switched to "original" VSCode with Claude Code extension. We'll see how that goes. - Did preliminary investigation of `testdays-web` code. 2025-11-27 12:43:00 - Installed Cursor and Claude Code. Apart from some initial experiments will learn to use those as I go. - Had a call with Tomas Hrcka 2025-11-26 12:01:00 - Started "Passwordless-GDM integration" Test Day preparation - Attended CLE team and Nexus Scrum meetings 2025-11-25 11:36:00 - Installed and learned to use `fedora-easy-karma` - Attended the Fedora Quality Meeting @adamwill:fedora.im =================== 2025-12-01 18:35:00 - reported [dep issues in a mesa f42 update](https://bodhi.fedoraproject.org/updates/FEDORA-2025-4a00e684c6#comment-4457896) and [a cups update breaking gnome's portalized printing](https://bodhi.fedoraproject.org/updates/FEDORA-2025-4b8311a915#comment-4458600) - [verified the fix](https://bugzilla.redhat.com/show_bug.cgi?id=2417631) for the vkcube validation test on rawhide and [merged the test and deployed it to prod](https://pagure.io/fedora-qa/os-autoinst-distri-fedora/pull-request/465) - reported the vkcube test [fails on f42](https://bodhi.fedoraproject.org/updates/FEDORA-2025-91c78a5dae#comment-4463223), but differently from the other failures, may just need working around - lots of work on openQA test flakes: [bumped desktop_update_graphical retries](https://pagure.io/fedora-qa/os-autoinst-distri-fedora/pull-request/467) to mitigate [the Discover update hang bug](https://bugzilla.redhat.com/show_bug.cgi?id=2404267), [bumped updates-kde-live-iso retries](https://pagure.io/fedora-qa/os-autoinst-distri-fedora/c/3d8ff3ce3a65c2885d809c761fed23b729d8e154?branch=main) to mitigate [the slitherer crash bug](https://bugzilla.redhat.com/show_bug.cgi?id=2417493), tried a few things for the [mysterious early VM hang bug](https://bugzilla.redhat.com/show_bug.cgi?id=2417734) and found that [downgrading edk2 seems to avoid it](https://bugzilla.redhat.com/show_bug.cgi?id=2417734#c9) 2025-11-28 18:07:00 - lots of discussion around the whole 'broken updates are sad' topic - [thread 1](https://discussion.fedoraproject.org/t/173585), [thread 2](https://discussion.fedoraproject.org/t/174907) - made a couple of bodhi proposals relating to that - [default critpath autokarma to 5](https://github.com/fedora-infra/bodhi/issues/6019), [add a minimum wait time for autokarma pushes](https://github.com/fedora-infra/bodhi/issues/6020) - ran some local tests to verify we can use the 'vkcube' reproducer for the mesa bug in a VM, [wrote an openQA test](https://pagure.io/fedora-qa/os-autoinst-distri-fedora/pull-request/465), discovered and [reported it's broken in rawhide too](https://bugzilla.redhat.com/show_bug.cgi?id=2417631) - noticed we forgot to EOL Fedora 41, [proposed a way forward](https://discussion.fedoraproject.org/t/fedora-41-eol-whoops/175073) - forgot to turn off the openqa ELN test scheduler while an ELN mass rebuild was happening, had to figure out a way to cancel 3,000 tests to avoid gumming up the works (figured it out after a bit) - [tweaked a failure message](https://pagure.io/fedora-qa/os-autoinst-distri-fedora/c/e47e711) in openqa tests to hopefully be of more use to packagers trying to debug their own failures (plus follow-up commits to add all the missing semi-colons, d'oh) 2025-11-27 01:49:00 - RH CLE team meeting, a call with brendan, some agile bureaucracy - Dealt with a [shadow-utils build](https://bodhi.fedoraproject.org/updates/FEDORA-2025-d6912811ad) that the packager didn't actually intend to go out. did a revert build, it turned out not to revert enough, did ANOTHER revert build, restarted a bunch of failed tests - upgraded several of my systems (main system to rawhide , test laptops to f42 and f43, rog ally to bazzite 43) 2025-11-27 19:55:00 - [looked into the details](https://bugzilla.redhat.com/show_bug.cgi?id=2417302) of the openqa failures with shadow-utils update - [reported](https://bugzilla.redhat.com/show_bug.cgi?id=2417493) an intermittent crash in slitherer (the custom browser component used to run anaconda on KDE lives) - upgraded openQA prod deployment to Fedora 43, thanks to nirik for helping deal with some recalcitrant hosts - reported [some](https://github.com/fedora-infra/fedora-messaging/issues/545) [deprecation](https://github.com/fedora-infra/fedora-messaging/issues/546) [warnings](https://github.com/fedora-infra/fedora-messaging/issues/547) in fedora-messaging - more input on the [thread about the mesa update](https://discussion.fedoraproject.org/t/is-most-recent-mesa-push-to-stable-mesa-25-2-7-2-fc43-reasonable/173585)