17:58:04 <marga> #startmeeting
17:58:04 <MeetBot> Meeting started Wed Feb 17 17:58:04 2021 UTC.  The chair is marga. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:58:04 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:58:11 <marga> #topic Roll Call
17:58:17 * gwolf Gunnar Wolf
17:58:18 <marga> Margarita Manterola
17:58:23 <ntyni> Niko Tyni
17:58:25 <bremner> David Bremner
17:59:29 <ehashman> Elana Hashman
18:00:18 <spwhitton> Sean Whitton
18:00:24 <marga> #topic Review of previous meeting's AIs
18:00:30 <marga> http://meetbot.debian.net/debian-ctte/2021/debian-ctte.2021-01-20-17.59.html
18:00:51 <marga> There were A LOT of AIs. Most people actually did their stuff... I didn't :-/
18:00:58 <ehashman> I did mine! want me to go over mine?
18:01:13 <spwhitton> ehashman: what is yours?  minute unclear
18:01:22 <ehashman> yeah it doesn't have an ACTION
18:01:27 <ehashman> > ehashman to continue working with doko and tumbleweed to address upstream concerns about Python packaging in Debian for bullseye and bookworm+ cycles (ehashman, 19:16:52)
18:01:29 * bremner wonders if he had action items. probably not a good sign
18:01:41 <ehashman> follow up is https://lists.debian.org/debian-python/2021/02/msg00035.html
18:02:05 <ehashman> we're currently working with the release team to get python3-full* into bullseye :)
18:02:11 <bremner> so, positive outcome!
18:02:15 <ehashman> (*name subject to change)
18:02:16 <ehashman> yes, very
18:02:31 <marga> Super cool
18:02:32 <ehashman> I also reached out to Python Steering Council/PSF board and requested a distro summit at the upcoming PyCon
18:02:38 <gwolf> Nice!
18:02:42 <ehashman> so we can try to head off these kinds of issues early
18:02:50 <gwolf> Thanks ehashman!
18:02:53 <ehashman> :D
18:03:04 <ntyni> yeah that sounds great
18:03:08 <ehashman> it's been a ton of work but I am very happy with how it's going! thanks also to doko and tumbleweed
18:03:24 <marga> \o/
18:03:26 <gwolf> yes, does not look an easy catherd to do
18:03:34 <gwolf> but it looks promising!
18:04:20 <marga> Alright, so I think from the AIs in the minutes, everything was done except for the two open bugs, so let's change topic to that.
18:04:28 <marga> #topic #975075 - Should NetworkManager support elogind?
18:04:39 <marga> I was supposed to draft a closing statement about this.
18:04:55 <marga> I started writing it in my head a few times, but never sat down to actually write the thing
18:05:11 <marga> I can carry over the AI, unless someone else wants to take it.
18:05:12 <spwhitton> It's not really urgent, so we could just action you again.
18:05:34 <spwhitton> I mean, we can action you again and it's fine if it takes time.
18:05:42 <marga> Ok
18:05:55 <marga> #action marga to draft a closing statement, explaining what happened and why we are declining to overrule, but encouraging people to talk to each other as much as possible.
18:06:00 <ehashman> unless we have other volunteers?
18:06:04 <ehashman> (not it)
18:06:20 <marga> No, it's ok. I should find time to do this.
18:06:25 <marga> #topic #976462 - Should dbgsym files be compressed via objcopy --compress-debug-section or not?
18:06:46 <marga> So, this one is the only bug actually open. We had one contribution on-list, as far as I can tell?
18:07:02 <spwhitton> We now have people speaking in favour of keeping them turned on as well as doko saying he would like them turned off.
18:07:14 <spwhitton> s/them/it/
18:07:54 <spwhitton> I think that the "keep on" side is more substantiated right now.  We don't have any specifics on the other side, e.g. examples of annoying tooling issues that have actually come up and are not just theoretical.
18:07:55 <gwolf> right - And no strong technical reason to lean either way IMO
18:08:18 <spwhitton> gwolf: no, I disagree -- we have examples where people's work is made a fair bit easier by having it on.
18:08:59 <bremner> still pretty anecdotal
18:09:11 <gwolf> spwhitton: Right, I agree - I meant "technical" as "something breaks" instead of "humans prefer"
18:09:25 <gwolf> But I agree "humans prefer" is a valid reason
18:11:56 <ntyni> it seems to me that compressing is an improvement and should be done if there are no clear reasons for avoiding it
18:12:18 <bremner> yes, that's a reasonable framing of the problem
18:12:25 <gwolf> right
18:12:26 <ntyni> i.e. erring on the side of compression
18:12:42 <ehashman> I'm not sure I agree, given that the packages come compressed to begin with
18:13:07 <ntyni> it's not about network bandiwidth but local disk usage
18:13:09 <spwhitton> it's about disc space, not netowrk, right?
18:14:06 <gwolf> I understand so
18:14:14 <gwolf> when downloading a package, it's compressed anyway
18:14:27 <ntyni> yes
18:14:39 <ehashman> we're talking about orders of MB though, right?
18:14:49 <gwolf> And I feel it's not a very common use case for people to do development on boards with very small disk spaces
18:14:55 <marga> Well, not when it's all of KDE
18:15:10 <ntyni> "in #631985 there is the example of debugging KDE requiring more than 10G of disc space"
18:15:25 <ehashman> yeah the problem is I feel like we need actual concrete numbers to make an informed decision
18:15:37 <spwhitton> laptop SSDs are not that large; I would not like to give up 10GB of mine
18:15:40 <gwolf> of course, you might want to debug where a problem happens...
18:15:46 <ehashman> is it possible for someone to do an analysis rather than going anecdote by anecdote?
18:16:04 <ehashman> is that something we can ask the bug reporter for?
18:16:13 * doko hides
18:16:20 <spwhitton> yes I think we can ask niels for that
18:16:43 <spwhitton> but doko's input would be very useful too
18:16:52 <spwhitton> doko: do you have a concrete "this breaks" case?
18:17:00 <ehashman> I would super appreciate that. otherwise I don't feel like we're making an informed decision
18:17:09 <doko> spwhitton: let's meet outside the meeting to collect data
18:18:20 <spwhitton> doko: okay.
18:18:33 <marga> where and when?
18:18:45 <marga> I fear if we don't make it concrete it just won't happen
18:18:45 <spwhitton> (or can we do this by e-mail?)
18:19:20 <doko> let's start on irc first, if we can find a time
18:19:47 <spwhitton> doko: we can do that if you think it would be more efficient, but may I ask why you think we have to do it synchronously?
18:20:01 <gwolf> doko: But *please* send a sum-up to the bug
18:20:19 <ehashman> yes :)
18:20:25 <ehashman> can we get an AI for this?
18:20:32 <marga> whose?
18:21:06 <ehashman> good question
18:21:23 * bremner looks meaningfully at spwhitton
18:21:35 <spwhitton> hang on
18:21:42 <spwhitton> I am not clear on what we think the next step is
18:21:49 * gwolf wonders at the meaning of that look
18:22:00 <spwhitton> I don't think we should assign just "do research" to a TC member
18:22:05 <bremner> meet with doko and summarize to the bug?
18:22:20 <gwolf> spwhitton: Well, understanding with concrete numbers what does this request actually mean..?
18:22:32 <spwhitton> I can't take that on I'm afraid.
18:22:32 <gwolf> mean/entail/...
18:22:46 <spwhitton> but, I am not sure any of us should
18:22:57 <spwhitton> qua TC member anyway
18:23:11 <ehashman> yes, I think we should request this of the original reporter (is that doko?)
18:23:16 <spwhitton> ehashman: no, niels
18:23:26 <gwolf> so, should we ask on the reporter to further document the request
18:23:36 <ehashman> concretely: I would like to see what the actual on disk space changes would be if we implemented this request
18:23:47 <ntyni> well it stems from #922744 by doko
18:24:00 <spwhitton> ehashman: well, taht is going to depend on what things an indviidual contributor has to debug.  so it is going to have to be at least somewhat anecdotal.  e.g. an imaginary person who contributes to X and Y
18:24:07 <ehashman> they don't have to do it for all the packages ever, but an actual package by package analysis for a representative cross-section would be appreciated
18:24:35 <ehashman> spwhitton: eh? the ask is "compressed vs not compressed" when installed, right? so this should be quantifiable
18:24:44 <ehashman> it is X size compressed and Y size uncompressed
18:24:49 <gwolf> doko: The original bug, just quoted, does not have much arguments on it...
18:24:55 <spwhitton> ehashman: but people only install debug symbols for packages they actually want to debug.  no-one installs all of them.
18:25:35 <gwolf> the bug is basically a request (i.e. doko asks "what's the reason for enabling this"), but does not really mention anything being wrong or broken because of it
18:26:05 <gwolf> "...binutils issue which fixes the alignment for compressed debug sections" → but then again, it's fixed already..?
18:26:23 <ehashman> spwhitton: of course, I don't mean all of them. I mean we need to see how much space difference it would make for some reasonable selection of these things
18:26:39 <gwolf> I don't see any arguments for not compressing
18:26:42 <spwhitton> ehashman: right, I just think that there won't be a single one of those
18:26:49 <ehashman> gwolf: adds complexity to the toolchain
18:27:04 <gwolf> ehashman: Yes, although minimally so
18:27:07 <ehashman> they have to be compressed and decompressed when you wanna use them right?
18:27:51 <ehashman> I don't think I can be the universal judge of whether yet another technical step is trivial or not :)
18:28:20 <gwolf> Well, "there are still tools outside which cannot handle" → which tools are they? Would it be best to fix the tools instead of decompressing symbols?
18:28:34 <ehashman> yes, those tools should also be documented!
18:28:44 <Myon> it does not "add" anything because the files have been compressed for years now, afaict
18:29:09 <ehashman> "it's always been this way" is not really a compelling argument for me
18:29:15 <gwolf> doko: Could you please further ellaborate about this on #922744?
18:30:32 * bremner now officially distracted by another meeting
18:31:05 <ehashman> I am fine to take an AI to request the things above on the bug mail
18:31:32 <spwhitton> ehashman: from niels?
18:31:44 <marga> #action ehashman to follow up on the bug requesting additional information from the bug reporter
18:31:46 <ehashman> I will ask doko about the tools and niels about the size diffs
18:31:50 <spwhitton> great
18:31:52 <spwhitton> thanks
18:31:55 <marga> +1
18:32:08 <ehashman> otherwise we're just gonna keep arguing in circles ;)
18:32:14 <ehashman> what's next!
18:32:17 <marga> #topic Recruiting efforts
18:32:27 <spwhitton> thanks for the mail about this today gwolf
18:32:43 <marga> So, to clarify on procedure, we can skip the private vote, but we *do* need to have a public vote
18:33:02 <gwolf> marga: Right. Do you want me to take that AI?
18:33:14 <marga> After that, we ask the DPL for confirmation and after that we do the musical chairs
18:33:14 <gwolf> Do you prefer to request the vote as the Chair?
18:33:25 <marga> gwolf, oh, I'm happy if you take it :)
18:33:34 <gwolf> Will do!
18:33:40 <marga> #action gwolf to call for a vote to fill our empty spot.
18:34:13 <marga> #topic Any other business?
18:34:40 <gwolf> aren't we skipping a bug?
18:34:53 <gwolf> #975075 ?
18:35:07 <bremner> that was first?
18:35:07 <gwolf> oh, right, no
18:35:09 <ehashman> ya
18:35:09 <marga> We covered that already, I got the AI
18:35:10 <gwolf> sorry
18:35:12 <gwolf> yes, yes
18:35:14 <gwolf> sorry
18:35:18 <gwolf> it wasn't topic-ed ;-)
18:35:20 <spwhitton> no business from me
18:35:38 * gwolf will be happy to go back to RL-work then
18:36:07 <bremner> I got nuthin'
18:36:19 <ntyni> me neither
18:36:19 <marga> Alright, we still have the pending things from our meeting last year, but I'd gladly do a short meeting this time and maybe discuss those other things next time when we have a new team member.
18:36:27 <ehashman> oh!!
18:36:32 <ehashman> what are we doing about chair
18:36:39 <ehashman> are we waiting on electing a new chair until after new members?
18:36:42 <spwhitton> voting after DPL apopints right?
18:36:42 <ehashman> iirc we had an action
18:36:45 <ehashman> okay
18:36:50 <gwolf> ehashman: I think we can vote after the new member enters
18:36:54 <ehashman> +1
18:36:56 <spwhitton> breaking a convention of ours though not any hard rules.
18:37:02 <marga> Yeah, we agreed to wait until the new member to avoid two votes in quick succession.
18:37:22 <marga> (this is what I called the "musical chairs" :) )
18:37:23 <ehashman> sgtm, just wanted to make sure that was explicit and not a forgot about it
18:38:02 <marga> Ok, let's wrap it up today, then.
18:38:06 <marga> #endmeeting