14:00:28 #startmeeting 14:00:28 Meeting started Thu Jan 22 14:00:28 2026 UTC. The chair is santiago. Information about MeetBot at https://wiki.debian.org/MeetBot. 14:00:28 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:30 hi 14:00:35 o/ hi 14:00:37 o/ 14:00:37 hi 14:00:37 #rollcall 14:00:47 o/ 14:00:49 hello o/ 14:00:50 o/ 14:00:54 please, say hi if you're here :-) 14:01:04 hi 14:01:21 while we wait for everybody, as a reminder, the meeting agenda is at https://pad.riseup.net/p/lts-meeting-agenda 14:01:49 hi 14:01:50 hello 14:01:54 hi 14:02:34 let's go with our first item: 14:02:46 #topic New team members 14:03:08 Arnaud Rebillout has recently joined the team (still in-training) 14:03:30 \o 14:03:35 welcome Arnaud :) 14:03:49 I was not expecting Arnaud to be here online 14:04:59 welcome! 14:05:00 and I think we can move forward 14:05:15 (we'll see if Arnaud can make any next meeting) 14:05:27 #topic Action item review 14:05:40 el_cubano, your name is on the agenda 14:05:46 do you want to handle this? 14:06:33 That's probably an oversight on my part. I'm pretty sure I meant to dump this one someone else :-) 14:06:44 :-) 14:06:53 But, yes, I can take this. 14:07:11 (... or on my part) 14:07:15 We carried over 2 action items from last month. AFAIK neither one has seen any progress. 14:07:33 I take that back. 14:07:42 There was discussion about the nvidia-graphics-driver 14:07:49 But I didn't follow the thread. 14:08:09 santiago: Since tobi isn't here, did you have anything you wanted to mention about it? 14:08:16 regarding nvidia-graphics-driver, we are waiting for Andreas to confirm if everything is done on trixie and bookworm first 14:08:30 ack. 14:08:42 So, this one should carry over. 14:08:46 Andreas mentioned that after trixie and bookworm, he would prepare the bullseye update, and we are waiting for him 14:08:50 yeap 14:08:53 #action carry-over: Follow-up to ML and Andreas concerning upload of nvidia-graphics-drivers 14:09:16 * el_cubano wonders if MeetBot records actions without providing any feedback 14:09:37 also hi! sorry a bit late o/ 14:09:49 utkarsh2102: Did you happen to have any update on the other action? 14:09:50 el_cubano, I *believe* it does 14:09:56 santiago: ack 14:10:06 IIRC, you had offered to help facilitate w/ jbicha 14:10:34 yes, spoke with Holger, i'll file a bug, CC'ing -devel to get broader discussion 14:11:00 but on a quick look, it doesn't hurt to seed d-s-s 14:11:16 el_cubano: i guess that's what you were asking about? 14:11:16 * charles thinks meetbot does record actions without feedback el_cubano 14:12:01 utkarsh2102: ack 14:12:03 so this item should be carried over again, right? 14:12:14 #action carry-over: Start a discussion about debian-security-support being installed by default 14:12:33 update on ^^: utkarsh2102 to file a bug & CC -devel for the discussion 14:12:38 yep, will do that^ 14:12:59 thx 14:13:10 OK. I think that's everything for action item review 14:13:17 OK, thank you, el_cubano 14:13:21 y/w 14:13:24 next item: 14:13:44 #topic Replacing the weekly semi-automatic unclaim with manual queries about the status / progress of the concerned packages 14:13:53 that's on me 14:14:15 probably you have noticed, last Monday no packages were (semi-)automatically unclaimed 14:14:59 oh, interesting 14:15:01 I think that is a very good idea. Will you record the progress report in the security-tracker? 14:15:06 instead of doing that, part of the (coordination) weekly task is to reach out to the different fellows about the status of "staled" packages 14:15:39 my gut feeling is that some human interaction could help 14:16:00 that might be a big task to take care of all "unclaims" 14:16:38 jochensp, WRT recording the progress on the security tracker, I'd say it depends 14:16:51 k 14:17:02 I'm mostly interested in the reasons for the time spent on those packages 14:17:08 if help is needed 14:18:06 I've seen sometimes is just a small issue blocking the update 14:18:46 I think this may work out for the best, so I'm happy we're trying it 14:19:12 ta, not sure it would be a such big task. but in any case, it's one of the duties on the coordination side 14:20:02 any questions or comments about this? 14:20:18 +1 14:20:54 I guess we can help by a) trying to finish packages in a shorter time frame and b) if it takes long, adding notes on what's blocking progress 14:21:20 +1 14:21:44 c) requesting help when needed 14:22:43 OK, if there are no objections, we can move to the next item in the agenda, that I guess is by helmut 14:23:02 # Debusine status 14:23:28 helmut, please, tell me if you didn't add that item 14:23:34 santiago: missing the "topic"? 14:23:39 ouch! 14:23:43 #topic Debusine status 14:24:28 simple. regression tracking didn't move 14:24:49 someone requested isolation-machine autopkgtests. that also is a feature requiring Debusine code changes 14:25:19 no progress on either of these unfortunately, but it was christmas and new year. let's stay tuned on regression tracking. any questions? 14:26:03 as repositories are available now, would it make sense to use them for customer testing? 14:26:19 What do you consider customer testing? 14:26:45 regression tracking is in need of tests, Colin was planning to test the feature soon 14:26:56 I think for ELTS we sometimes send test versions to customers before publishing them 14:27:39 jochensp: that is something we can probably do. We do not generally enable the creation of experiment workspaces yet (also requested by e.g. ta). You may ask for a workspace if you have such a need. 14:27:43 regarding isolation machine, what feature is missing? It should work but it requires to maintain task configuration to know which package should run with the incus-vm backend 14:28:36 buxy: https://salsa.debian.org/freexian-team/debusine/-/issues/653 14:28:58 FWIW, that doesn't stop you from allowing the use of incus-vm 14:29:06 It just means that it won't work on our arm workers 14:29:42 We could create a separate workflow to run incus-vm autopkgtests indeed, but we cannot enable it as part of the upload pipeline yet. 14:32:08 do we really need a separate workflow? 14:32:44 Until that issue is resolved, we would. Once it is resolved, we would generally use incus-vm for amd64 autopkgtests (like ci.f.c). 14:33:10 helmut: why not just allow incus-vm to be manually specified for a run? 14:33:37 (and let it fail on ARM) 14:33:58 That's an easy change and I have no objections. Would ELTS contributors be happy with that? 14:34:58 what do you mean by "manually specified"? 14:35:17 autopkgtest_backend: incus-vm 14:35:20 Beuc: your dput invocation gains a separate option 14:35:22 dput -O debusine_workflow.autopkgtest_backend=incus-vm ... 14:36:18 We can try. 14:36:39 #action helmut to allow setting autopkgtest_backend on elts upload workflows 14:37:07 (incidentally what's the issue with ARM, no support for virtualization on the ELTS ARM hardware?) 14:37:27 no nested virtualization support 14:38:18 our workers are virtual machines. we would need physical workers for doing incus-vm on arm 14:39:08 helmut, could you remind me please, would autopkgtest_backend=incus-vm be possible on debusine.debian.net too? 14:39:09 (or have the workers connect to a bare-metal incus setup? anyway let's move on :)) 14:39:47 santiago: it's already possible there 14:39:47 santiago: we can change that there as well. tumbleweed can you handle that? 14:40:03 helmut: sure 14:40:08 OK, thanks! 14:40:29 https://debusine.debian.net/debian/developers/workflow-template/upload-to-bookworm/ has autopkgtest_backend in Runtime parameters (and same for all) 14:40:32 if there are no more comments / questions, I think we can move on 14:40:44 buxy: I thought as much :) 14:41:29 #topic AOB 14:42:01 Regarding authentication to access the elts repo, should we mail someone to get credentials? That is https://www.freexian.com/lts/extended/docs/how-to-use-extended-lts-repositories/index.html#configuring-authentication-for-apt 14:42:11 I heve a few topics there... 14:42:43 charles: there is authenticated access for customers, but authentication will not be required in 2026. 14:42:44 we don't have a timeframe for taking the public repo down yet 14:43:11 ah, cool, then the follow-up question is also answered 14:43:25 ack then 14:43:52 will continue to use deb.freexian.com for now as sources.list forelts 14:44:21 yes, no changes are required, for the time being 14:44:30 next topic? 14:44:33 The next question might be a bit more directed to santiago: Any updates on salsaci-arm64-runner-01.debian.net? It is still offline 14:44:54 charles, there is some movement on the new hosting place 14:44:57 since I did the change turning those jobs off on salsa-ci, I'm trying to keep an eye to revert them 14:45:06 FYI, it is now possible to pass autopkgtest_backend 14:45:28 but the machine is not up yet... :( 14:45:29 santiago: ack, but no clear timeframe, right? 14:45:46 charles, no, I am afraid. It doesn't depend on us 14:46:00 ack, no problem, I'll keep an eye from time to time 14:46:01 I'll let you all know when there are changes on that 14:46:08 ok, thanks 14:46:08 next topic? 14:46:21 I out of topics :-) 14:46:44 thanks for raising those topics 14:46:49 any other business? 14:47:56 OK, so I think we can conclude the main part meeting. We have in our agenda the appendix, about the discussion on technical questions or issues 14:48:25 #info Next meeting: 2026-02-26 14:00 UTC 14:48:51 on Jitsi 14:49:32 we can end up the meeting here, and if somebody has any technical issue to discuss, don't hesitate to speak up 14:49:43 thank you everybody for attending 14:49:51 #endmeeting