<@ulm> !proj council [17:00] (council@gentoo.org) arthurzam, dilfridge, eschwartz, floppym, mgorny, sam, ulm <@ulm> meeting time <@floppym> Hello <@ulm> agenda: https://public-inbox.gentoo.org/gentoo-dev/upl001r4l@gentoo.org/ <@mgorny> managed! <@ulm> 1. Roll call * @arthurzam here <@dilfridge> ni hao * @ztrawhcse here * @mgorny here * @dilfridge here * @ulm here * @floppym here <@ulm> sam_? [17:01] * @sam_ here <@ulm> excellent <@ulm> 2. Open bugs with Council participation <@floppym> I get a 403 forbidden on bugzilla. <@ulm> bug 936211 [17:02] <@arthurzam> then no bugs \0/ https://bugs.gentoo.org/936211 "[Tracker] Gentoo Foundation dissolution"; Gentoo Foundation, Proposals; CONF; ulm:trustees <@mgorny> lemme check <@ulm> works? <@ulm> anyway, nothing new there since last month <@mgorny> floppym: regex looks correct <@mgorny> floppym: aren't you blocked by scraper protection perhaps? <@floppym> Can troubleshoot later, continue <@ulm> the AGM will be just after the council meeting in #gentoo-nfp <@ulm> bug 961301 [17:03] https://bugs.gentoo.org/961301 "[Tracker] Requests for metadata/AUTHORS"; Gentoo Council, unspecified; CONF; ulm:council <@ulm> I see no pending requests <@ulm> bug 961811 ulm: https://bugs.gentoo.org/961811 "Consider procuring Token2 PIN+ security keys for developers"; Gentoo Council, unspecified; CONF; tamiko:council <@ulm> ztrawhcse: robbat2: you wanted to talk to soap? <@ztrawhcse> right, but I forgot :( [17:04] <@ulm> can you do this e.g. next week? [17:05] <@ztrawhcse> I just went ahead and PMed him now <@ulm> good :) <@ulm> I think nothing else there for now, we have to wait for the status from soap [17:06] <@ulm> bug 962281 ulm: https://bugs.gentoo.org/962281 "Clean up consultants list"; Gentoo Foundation, Proposals; CONF; negril.nx+gentoo:council <@dilfridge> sorry no news there either <@dilfridge> will re-send it to the list with a deadline <@ulm> k <@dilfridge> as discussed last time <@ulm> bug 965878 ulm: https://bugs.gentoo.org/965878 "[TRACKER] Codeberg migration"; Gentoo Council, unspecified; CONF; mgorny:council <@ulm> and related bug 979541 [17:07] ulm: https://bugs.gentoo.org/979541 "Frequent API 50x errors from Codeberg"; Gentoo Infrastructure, Other; CONF; mgorny:repo-qa-checks <@ulm> mgorny: any news? <@mgorny> well, there are periods [17:08] <@mgorny> right now things are looking better again * @ulm notes that there are 5xx errors for our own infrastructure, too <@mgorny> yes, our infra is under DDoS as well <@mgorny> comparatively, codeberg is doing better [17:09] <@ulm> mgorny: so we keep observing for another month? * @dilfridge is glad it's not 451 <@mgorny> i guess. it's not like we have anything specific to do, do we? <@dilfridge> not really <@mgorny> i'd like to migrate guru to codeberg, but it's more about getting agreement among guru folks and time to do that <@mgorny> oh, and there's dev-vcs/gagit now, to help creating PRs via agit flow [17:10] <@ztrawhcse> figuring out some kind of save+retry for API calls could be interesting <@mgorny> it fills the gap where people don't end up using the template <@mgorny> i think laumann will work on that at some poin <@mgorny> and arthurzam on pkgcore end for bugzilla <@arthurzam> yeah, for whoever cares, newerst pkgcore has a nice bugzilla python api [17:11] <@mgorny> by the way, i'm thinking we'd use a "full-time" sysadmin to deal with the DDoS-es on infra side <@mgorny> like, i'm totally incompetent, as you can see [17:12] <@mgorny> and sam_'s got like a thousands things to do <@ulm> anything else about the codeberg bugs? <@arthurzam> ^ give me a minute [17:13] <@ulm> seems not. moving on <@sam_> i've wondered the same mgorny <@dilfridge> yes <@ulm> k <@dilfridge> infra is one of the cases where paying someone for work would make sense <@arthurzam> I think the council can disconnect from the codeberg infra issues. They exist, they aren't 100% availability, but nighter are we. I think we are already done as a council with codeberg migration "laws". Now it just devs are convinced to migrate, but this isn't actionable by council [17:14] <@mgorny> well, unless we want to do something like "get off github", which i don't think w do <@arthurzam> I don't see value with those 1/2 bugs being listed in each council meeting <@ztrawhcse> the lack of auditability for codeberg is slightly worrying <@dilfridge> that cant be solved without codeberg though [17:16] <@ulm> arthurzam: so what do you suggest? reassign? <@arthurzam> yes <@ulm> to whom? <@arthurzam> infra? <@dilfridge> wfm <@ulm> sure, if infra is ok with it [17:17] <@ztrawhcse> dilfridge: not true, sourceware has it set up as custom automation :D <@mgorny> to codeberg team/ <@arthurzam> I think it is mostly the same member count * @arthurzam hides <@ztrawhcse> but then again that may not be a council issue <@ulm> arthurzam: can you take care of it? <@arthurzam> not for the coming week (still in the problematic place you all know) [17:18] <@floppym> One of them is already assigned to repo-qa-checks. <@floppym> The other is a tracker. <@ulm> ok, let's take council out of CC in #979541 [17:19] <@ulm> moving on <@ulm> bug 965900 ulm: https://bugs.gentoo.org/965900 "GLEP 63: update key recommendations"; Documentation, GLEP Changes; CONF; mgorny:council <@ulm> no news there either <@dilfridge> blocked by token2 <@ulm> we said in the last meeting that we want to discuss it on the -dev ML or the bug [17:20] <@ulm> but someone has to start the discussion :/ <@floppym> I think I said I would ask what the recommendations are regardless of harware. <@floppym> (I did not) <@floppym> Anyway, nothing that needs discussing here. [17:21] <@ulm> yes * @dilfridge is running low on caffeine <@ulm> bug 980490 ulm: https://bugs.gentoo.org/980490 "Grammar error(s) in AI Policy"; Gentoo Council, unspecified; UNCO; roman.zilka:council <@ztrawhcse> can't it suffice to say we recommend newer key types but leave people who really need rsa the option of doing so <@ulm> that's a new bug <@dilfridge> we should definitely leave the errors in as sign for human craftsmanship [17:22] <@arthurzam> lol <@ztrawhcse> speaking of the policy, it is also a but ambiguous what it actually covers [17:23] <@ulm> mgorny: any comments? AFAICS it's your wording, originating from an e-mail sent to the ML <@ztrawhcse> I usually tell people it covers any repository hosted on gitweb.gentoo.org <@floppym> I agree with the commenter as a native English speaker. <@mgorny> yeah, it's a typo <@arthurzam> so let's vote on the council accepting the fixed wording? <@floppym> s/commenter/reporter/ <@dilfridge> just fix it <@sam_> ^^ [17:24] <@dilfridge> it's a purely editorial change <@sam_> seriously, just fix it <@ztrawhcse> I vote "that is an editorial change" <@ulm> looks like there's consensus <@floppym> Yes. <@mgorny> well, we can always discuss making a stronger policy xP <@mgorny> something like "anyone who used an llm in last 7 days is forbidden from touching gentoo before washing their wands with acid" <@ulm> so, "should a case be made" and "for such a tool"? <@ztrawhcse> sure. how about this <@dilfridge> "burn all the clothes you may have worn while online" [17:25] <@mgorny> honestly, we can also remove the whole sentence <@floppym> ulm: Yes, those phrases look good to me. <@ztrawhcse> it is also forbidden to use LLMs to produce tools that implement PMS or are otherwise expressly targeted at Gentoo systems , the consequences for which are, "cannot be packaged in gentoo or guru" <@mgorny> it's "obvious" <@mgorny> it's literally stating "a policy can be changed in the future" [17:26] <@floppym> That's true. <@sam_> it's not pointless <@mgorny> perhaps it's giving a bit of emphasis on what case could cause a change of policy <@sam_> the purpose of it is to make clear we're open to change <@sam_> exactly <@mgorny> but it doesn't really prevent a future council from disregarding that <@sam_> rhetoric may be logically useless but it serves a purpose [17:27] <@ulm> I'd suggest we leave the sentence in place (with the proposed fix) <@sam_> it's to show we're not insane and to let people approach with ideas <@dilfridge> ulm++ <@sam_> not going to bite their heads off if they ask us about some assistive technology or something * @mgorny grinds an axe <@ztrawhcse> my understanding was, yeah, we want to indicate to people what it would take to be worth nagging us over "the policy should be changed" <@floppym> I don't feel strongly either way. Keep it or dump it, either is fine. <@dilfridge> whoever wants to dictate their ebuilds with the help of an LLM dictation system ... <@mgorny> well, "over" sounded valid to me, but i'm no native speaker [17:28] <@ztrawhcse> I could see "over" being used, though "for" would be better <@ulm> let's check the grammar with an LLM :p <@dilfridge> can we delegate this to an AI? it's wasting time <@mgorny> i think "over" has a different, err, tone? <@ztrawhcse> ulm: sure, here is an ASCII drawing of the union jack as you requested [17:29] <@ulm> moving on, before this entirely derails <@ulm> bug 980497 https://bugs.gentoo.org/980497 "Missing summary for 20260510 and 20260614 council meetings"; Gentoo Council, unspecified; CONF; ulm:dilfridge <@dilfridge> right <@dilfridge> on it <@ulm> ready in time for next meeting? <@dilfridge> I hope so. [17:30] <@floppym> I need to figure out how to dump logs out of soju before I chair I suppose. :) <@arthurzam> dilfridge: try to use a LLM for that * @arthurzam hides <@ulm> 3. Open floor <@ztrawhcse> you can use /quote to manipulate chathistory for a range and have it re-sent to your client <@ulm> anything? <@floppym> ztrawhcse: Ah, I was going to go straight to the DB, but I'll try that. [17:31] <@ulm> waiting until 15:33 <@floppym> Nice short meeting today. [17:33] <@ulm> meeting closed <@ulm> thanks everybody! <@dilfridge> \o/ thanks <@mgorny> thanks * @floppym waves <@arthurzam> thanks <@mgorny> ztrawhcse: do you want to open the discussion on extending the AI policy on the ml? <@ulm> Foundation AGM is in #-nfp at 16:00 UTC *** ulm (~ulm@gentoo/developer/ulm) has set the topic for #gentoo-council: "277th meeting: 2026-09-13 16:00 UTC | https://www.timeanddate.com/worldclock/fixedtime.html?iso=20260913T16 | https://wiki.gentoo.org/wiki/Project:Council | https://dev.gentoo.org/~dilfridge/decisions.pdf" [17:34] <@dilfridge> #-nfp or #-trustees? the latter I think <@ulm> so next month will be one hour later <@ulm> dilfridge: right, #gentoo-trustees <@ulm> thanks