[21:02:20] !proj council [21:02:20] (council@gentoo.org) dilfridge, gyakovlev, mattst88, slyfox, ulm, whissi, williamh [21:02:20] meeting time (now for real)! [21:02:21] (council@gentoo.org) dilfridge, gyakovlev, mattst88, slyfox, ulm, whissi, williamh [21:02:32] 1) roll call [21:02:36] -*- gyakovlev here [21:02:37] -*- Whissi here [21:02:39] -*- dilfridge here [21:02:42] -*- ulm here [21:02:49] -*- slyfox here [21:02:54] -*- mattst88 here but might have a proxy as well [21:03:46] Who is your proxy? [21:03:56] ok let's wait a moment for WilliamH to pop up [21:04:06] sam_ and zlogene were fighting over it :) [21:06:02] let's go on [21:06:04] 2) [21:06:13] Git vs. rsync [1] [21:06:19] https://archives.gentoo.org/gentoo-project/message/59f9ac24ceb34abed26c990168b75f87 [21:06:26] is there anything we need to discuss here? [21:06:49] nothing for the council to decide at this time, I would say [21:06:50] I don't think so, other than maybe asking infra@ to start investigating how we could have a Git mirroring system [21:07:26] mgorny also has an interesting idea that might not require extra infrastructure, AFAIK [21:07:48] but yeah, nothing for us to decide :) [21:07:54] right-e-o [21:08:02] so we can probably drop this fast now [21:08:06] er [21:08:18] antarus: you called? [21:08:18] I'm looking for people interested in working on it [21:08:31] if you are relying on 'infra' you are not going to get results [21:08:37] worth writing an email to -dev? [21:08:39] wasn't slashbeast going to? [21:08:40] (no offense to them) [21:09:05] but this is partyl where I'm expect council to be more leader-y [21:09:30] is this project worth doing, can the council help prioritize drive it? Can we fund it with money? [21:09:33] that sort of thing [21:09:57] if you don't want to lead that way, thats fine, I'm just trying to point out an opportunity of sorts [21:10:03] anyway think about it; I don't expect it to be resolved today [21:10:16] I'd be happy to have council do a quick vote that just says "we think moving to git by default is worth exploring" [21:10:33] so my 2ct: I dont see any urgent need for action, for me rsync works nicely most of the time [21:10:33] Well, worth exploring if all problems are resolved :) [21:10:50] and if it doesnt there's emerge-webrsync [21:10:52] antarus: maybe too early, because discussion in the ML is still going on [21:10:56] but you have a point [21:11:10] I'd also like something like an 'auto-sync' provider [21:11:11] Whissi: exploring means finding problems. [21:11:21] or building compatible providers [21:11:39] like what if we use web-rsync to bootstrap yoru repo, but it web-rsync'd a git compatible repo? [21:12:22] mattst88: +1, we can't know without trying, designing everything up front is pretty expensive and not very agile [21:12:38] With this sort of complexity I doubt that anything will move in next 5 years but sure =) [21:12:38] exactly [21:13:14] Whissi: how's your test ::gentoo repo doing, btw? [21:13:27] I won't have access until Tuesday [21:13:38] did we solve /that/ problem so that you can maybe acknowledge that it's not OMG broken? [21:14:02] antarus: like using bit bundles? [21:14:22] The people who wants to switch default to git should maybe also write down a list of arguments and the people who want to stick to rsync should do the same. Then we can compare and work on problems we found that way. [21:14:57] Whissi: you're way ahead of where antarus and I are [21:15:17] In the end I guess there should be `emerge ---config portage` where you switch between rsync and git for example so this shouldn't be a big deal in the end. [21:15:19] Sorry I missed the meeting folks. :( [21:15:50] Whissi: so my long term plan is to deprecate rsync [21:16:06] so 'people who want to stick to rsync' would need to spin up their own network, potentially [21:16:18] but this is again part of the FLOSS promise, right, that would be possible ;p [21:17:01] Well, I don't see that happen but write down your arguments why you think this should happen and if we will find no blockers, why not. [21:17:24] FWIW, I think Whissi is just being overly negative, and we shouldn't take too much time to try to convince him of anything at the moment [21:17:36] WTF? [21:17:41] (there's not anything to convince him of at the moment anyway) [21:18:01] well, you started by saying "Well, worth exploring if all problems are resolved :)" [21:18:36] Probably a misunderstanding: [21:19:15] You wanted a vote. But I don't see why we need a vote for improvement. I.e. we can try to improve Gentoo all the time. And if we get a PoC which turns out to be working... we don't need a vote allowing someone to tackle a problem. [21:19:48] well, antarus said he wanted some leadership here [21:19:55] I didn't ask for a vote [21:19:56] i.e. choosing to push for a solution, encourage investigation [21:20:02] I was only suggesting that Council affirmed that this was a thing we're interested in [21:20:03] I think we need a project owner anyway who would work on more detailed proposal that would outline requirements in a clearer form. [21:20:31] yes [21:20:44] i think the first step forward would be getting a git bundle generator online [21:20:56] provide technical leadership is by actually leading; not by simply voting on stuff other people present ;p [21:21:10] so people could bootstrap git repo without relying on 'git clone' working [21:21:21] man, I was just agreeing with you, and saying you were asking for leadership [21:21:30] (which is different from "we can try to improve Gentoo all the time") [21:21:42] mgorny: is "git clone" still that fragile even for shallow clones? [21:22:21] So mattst88 will volunteer to lead the task force exploring possibilities to switch Gentoo repository to git sync by default. [21:22:24] dilfridge: i haven't experience that myself but i consider inability to resume to be a serious limitation [21:22:32] that it is [21:22:49] dilfridge: it doesn't remember what it had downloaded previously [21:22:59] which I think qualifies as fragile [21:23:16] yeah... not extremely bad for shallow clones, but still quite bad [21:23:17] sounds like a nice feature request for ustream git :) [21:23:19] the cheap solution is to generate and cache a bundle on server for some time [21:23:24] let people fetch it via HTTPS [21:23:39] then discard it after N minutes with no activity [21:23:39] it's not happening upstream [21:23:47] git is worse than cvs in that respect :) [21:23:49] I2P were interested in it (it's like Tor) which has unreliable streams [21:23:54] they wanted this feature for a long time [21:23:57] they opted for bundles IIRC [21:24:03] *nod* [21:24:11] (upstream git is only interested if I think someone funds the work) [21:24:19] good question though, I just meant I had some interesting info :) [21:24:23] i'm not surprised, tbh [21:24:30] it's probably a lot of work to get it right [21:24:40] yeah, definitely, and I guess they view it as already handled [21:25:32] if i find some time, i'll try extending sync mirror with metamanifests on a separate branch [21:25:45] dilfridge: next topic? [21:25:48] so it would be more useful without actual git backend [21:25:54] next topic [21:26:02] 3) open bugs [21:26:26] bug 736760 [21:26:27] dilfridge: https://bugs.gentoo.org/736760 "Application to Software Freedom Conservancy"; Gentoo Foundation, Proposals; CONF; mgorny:trustees [21:26:32] anything to discuss here? [21:26:43] WilliamH: you only missed the start, it's ongoing [21:27:03] I guess he has figured that :) [21:27:10] Yeah I figured that. :-) [21:27:13] I hope he figured that. [21:27:17] he hasn't said anything since, so I wasn't sure [21:27:20] seems not [21:27:27] I'm not sure why this bug in particular is assigned to council [21:27:31] bug 729062 [21:27:32] dilfridge: https://bugs.gentoo.org/729062 "Services and Software which is critical for Gentoo should be developed/run in the Gentoo namespace"; Gentoo Council, unspecified; IN_P; jstein:council [21:27:41] Whissi: isn't that your baby? [21:27:42] The Foundation met with some other umbrella's recently [21:27:47] I swear that the past meeting wasn't 4 weeks ago :p Anyway, we have scheduled a meeting after today's council meeting and will probably finalize and send our mail Mo/Tue. [21:27:51] but we have not made significant progress in any direction [21:28:13] ok let's skip the next one, then we get to [21:28:21] bug 774489 [21:28:22] dilfridge: https://bugs.gentoo.org/774489 "GLEP 67: add proxied-maint="" attribute"; Documentation, GLEP Changes; CONF; mgorny:glep [21:28:31] isn't that already done? [21:28:50] yes, it is [21:28:51] ok I'll close it later on [21:28:52] needs to be labelled "final" [21:29:00] Yes, I believe that's already done. [21:29:05] bug 779451 [21:29:06] dilfridge: https://bugs.gentoo.org/779451 "Request to add Gentoo developer business card to Gentoo Artwork"; Gentoo Foundation, Artwork approval; UNCO; alicef:artwork [21:29:08] that's also done [21:29:24] oh, glep 67 is final already. so no action [21:29:36] ok then we get to [21:29:39] 4) open floor [21:29:42] anyone? antarus? [21:29:56] hey don't pick on me ! ;p [21:29:58] I said my piece already [21:30:37] dilfridge: bug 751010? [21:30:38] https://bugs.gentoo.org/751010 "Missing log and summaries for 20191110, 20191208, and 20200412 council meetings"; Gentoo Council, unspecified; CONF; ulm:council [21:30:38] I didn't say a lot about switching to git syncing by default because I tend to agree that we can't until we can support it. [21:30:45] yeah... [21:30:59] just that it's in the log :p [21:31:27] ok [21:31:30] that's it then [21:31:32] cheerios [21:31:37] -*- dilfridge bangs the gavel [21:31:40] meeting closed