diff options
Diffstat (limited to 'meeting-logs/20221113.txt')
| -rw-r--r-- | meeting-logs/20221113.txt | 406 |
1 files changed, 406 insertions, 0 deletions
diff --git a/meeting-logs/20221113.txt b/meeting-logs/20221113.txt new file mode 100644 index 0000000..d3d48fd --- /dev/null +++ b/meeting-logs/20221113.txt @@ -0,0 +1,406 @@ +11:00 <@mattst88> okay, meeting time +11:01 <@mattst88> 1. Roll call +11:01 * | ulm here +11:01 * | sam_ here +11:01 <@mattst88> dilfridge, gyakovlev, mgorny, ajak +11:01 * | mattst88 here obviously +11:01 * | ajak here +11:01 * | dilfridge here +11:01 * | gyakovlev here +11:02 * | mgorny here +11:02 <@mattst88> great, everyone's here +11:02 <@mattst88> 2. GLEP78 update (https://archives.gentoo.org/gentoo-dev/message/fa090f351778f1644964d096c20933dd) +11:02 <@mattst88> mgorny: want to explain what we're doing here? +11:02 <@mgorny> the mail says it all +11:03 <@mgorny> - replacing "basename" with more specific explanation +11:03 <@mgorny> - linking formats to the respective standards +11:03 <@mgorny> - deferring compressed file formats to GLEP 74 +11:03 <@mgorny> - clarifying the situation around Manifest signing and verifying binpkg authenticity +11:03 * | dilfridge yes +11:03 <@mattst88> okay, so we're just approving the changes here? +11:03 <@mgorny> unless someone has questiosn +11:03 <@mgorny> (sorry, i keep making typos today) +11:03 <@gyakovlev> https://archives.gentoo.org/gentoo-project/message/ef20321c60472ac4c91653236fb8a346 +11:03 <@gyakovlev> ^ linking agenda for log purposes +11:03 <@mattst88> thanks :) +11:04 <@sam_> my only question is if portage is currently compliant with all of this or if we need syu to work on any of it +11:04 <@mgorny> i think so +11:05 <@mgorny> though i've never really verified the verification part, though syu is pretty competent there, so i don't think he'd fail +11:05 <@sam_> yeah, I've started using a bunch of the pgp stuff and it seems to work well +11:05 <@mattst88> presumably we have some ack from syu that the glep changes are what he expects? +11:05 <@sam_> good question +11:06 <@mgorny> i don't really recall, this was years ago xP +11:06 <@mgorny> (time is moving really fast for /me) +11:06 <@mattst88> heh, fair. the patches have been up since Sept 21, so I'll assume he's aware :) +11:06 <@mattst88> if he's not, we can revisit :) +11:07 * | dilfridge yes +11:07 <@sam_> yeah, no problem with that +11:07 <@mattst88> okay, motion to approve mgorny's 6 patches to update GLEP78 +11:07 * | mattst88 yes +11:07 * | dilfridge yes +11:07 * | sam_ yes +11:07 * | ajak yes +11:07 * | gyakovlev yes +11:07 * | ulm yes +11:07 * | mgorny yes +11:07 <@mattst88> perfect, thanks +11:08 <@ulm> mgorny: this isn't in the repo yet, right? +11:08 <@mgorny> ulm: could you clarify which repo? +11:08 <@mgorny> glep.git? +11:08 <@ulm> glep.git +11:08 <@mgorny> ulm: don't think it is +11:09 <@mattst88> I don't see a glep78 branch, FWIW (but I do see a glep76 branch, which is the next item) +11:09 <@mattst88> 3. GLEP76 update (https://gitweb.gentoo.org/data/glep.git/commit/?h=glep76&id=139198d2e8560f8dfb32c8f4c34a3e49d628b184), discussion at https://archives.gentoo.org/gentoo-project/message/c85b78ca69802522534ee8ab0804f665) +11:11 <@sam_> robbat2: around? +11:11 <@mattst88> I think again we're being asked to approve the changes to GLEP76, which consist of new wording around contributor's names +11:11 <@dilfridge> sigh +11:11 <@ulm> this will somewhat weaken our position, as far as it comes to due diligence. so I'm not a huge fan of it +11:11 <@dilfridge> does a twitter blue tick fulfil requirements of a records search? https://twitter.com/jesus +11:11 <@ajak> but does it actually change the reality that we're accepting people with non-legal names? +11:11 <@mgorny> it's obviously very important, it got far more attention than any of the patches that were sent for review recently +11:11 < ulm> | mattst88: it's two-step approval I think, first us then trustees +11:12 <@sam_> what mgorny said, and I think it got fleshed out pretty fully on the ML +11:12 <@dilfridge> trustees for any practical purpose dont exist anymore +11:12 <+robbat2> i'm here +11:12 <@dilfridge> or if they do, we have to plan for the moment when they dont +11:12 <@ajak> yeah i'm not sure why trustees would need to review this, given they delegate to us in general anyway +11:12 <@sam_> hi robbat2, we're talking about the glep 76 update (real name stuff) +11:12 <@mattst88> a meta-question: what is the objective in requiring real names for contributors? that is, what problem do we hope to avoid? +11:13 <@mgorny> i suppose it falls to "due diligence" +11:14 <+robbat2> the real name requirement was semi-inherited from the Kernel DCO when that policy was adopted for Gentoo +11:14 <@mgorny> if we do our part in "due diligence" in accepting contributions, our users can assume "due diligence" when they get in trouble because we f* up +11:14 <@mattst88> the kernel was the subject of a many-years-long lawsuit over allegations of it containing code from UNIX +11:14 <@ulm> mgorny: not sure if I understand this +11:15 <@mattst88> what would the analogous situation be for Gentoo? +11:15 <@mattst88> contributors copy a few lines of a PKGBUILD file into an ebuild? +11:15 <+robbat2> records search in the context of the GLEP means to me searching government records, not public records like (twitter.com/jesus) +11:15 <@ajak> what are we getting at here? is there a will to drop the name requirement entirely? +11:15 <@mattst88> ajak: who is that question for? +11:15 <@mgorny> robbat2: then perhaps that should have been worded explicitly because i thought it means "any records" +11:15 <@ajak> robbat2: ie, "attestable in a written statement with a witnessed signature before a notary" +11:16 <@ sam_> | on the one hand I don't really want to reopen the whole point of the GLEP, on the other hand, I can get why mattst88 is asking as it affects the level of standard we want +11:16 <@mgorny> like, "if someone established an identity online, i.e. used the same name for reasonably long time" +11:16 < CatSwarm> https://lwn.net/Articles/200222/ +11:16 < ajak> | mattst88: everyone in general i guess +11:16 < CatSwarm> this was the commit in question ~ [re: DCO] +11:16 <@ulm> robbat2: that's an important clarification, so probably we should add the word "government" +11:17 <@dilfridge> "attestable in a written statement, with a witnessed signature as before a notary" +11:17 <@dilfridge> ^ 1) attestable by whom? 2) what is a notary, definitions vary wildly between countries... I would drop the half-sentence after "and/or" +11:18 <@ulm> "... can be verified by simple search of government records." full stop +11:18 <@ajak> fine by me +11:18 < CatSwarm> these are major change in tone form the one I was credited with contributing to. +11:18 <@ulm> or rather "... can be verified by a simple search of government records." +11:18 <+robbat2> not all governments have searchable records :-( +11:18 <@dilfridge> that is fine by me too, but changes what was attempted to achieve +11:19 < CatSwarm> *from the original which was discussed, this is a total change in intent and methodology. +11:19 <@dilfridge> (just pointing that out) +11:19 <@mgorny> robbat2: honestly, i'm surprised at all that governments provide searchable records of people +11:19 <@mgorny> also, what would the search prove, in the first place... +11:20 <@mgorny> that at least one person having a particular name exists? +11:20 < CatSwarm> it seems dubious to think that government records are easily accessible when GLEP-76 doesn't require contributors to specify their jurisdiction of residence. +11:20 <@ajak> how about we punt it back to gentoo-project for more review seeing as there are new, previously undiscussed problems brought up here? +11:20 <+robbat2> yes, that a person with a given name exists only (so somebody can't say your name is made up) +11:20 <+robbat2> why didn't anybody raise these concerns in the previous months of discussion? +11:20 <@sam_> yes, I think this should've been brought up before +11:20 <@sam_> yes +11:20 <+robbat2> you had months to raise them +11:21 <@mgorny> robbat2: so, technically speaking, if we all suddenly start committing as, say, Michał Górny? ;-) +11:21 <@sam_> i don't think it's fair to the people who've raised this change to only bring up things like this now, when they've been apparent for months +11:21 <@mgorny> because such a person exists (actually, it's a very popular name in PL) +11:21 < CatSwarm> these new problems were last minute, so please punt it back to -project ML if there's going to be discussion of a major rewrite to the language I proposed to minimize impact to privacy and dignity of queer / trans / nonbinary / refugee & other folx who might be underdocumented. +11:22 <@ulm> CatSwarm: what percentage of contributors would those be? +11:22 <@ajak> it doesn't matter. +11:22 <@ulm> do you have any numbers? +11:22 <@sam_> doesn't matter & it's obviously not something CatSwarm would have stats on anyway +11:22 <@ajak> and there's no sense spending more time on this here +11:22 <@sam_> plus not a matter for right now +11:22 < ajak> | mattst88: motion plz? +11:22 <@ulm> send it back to the drawing board +11:22 <@ulm> i.e. -project ML +11:22 <@mattst88> ajak: I'd at least like my question to not be totally ignored... +11:22 <@sam_> why didn't you raise any of this on the ML earlier then? +11:23 <@mgorny> ulm: bircoph mentioned there are a lot of people who would have committed to gentoo for decades, if onyl we hadn't created glep76 ;-) +11:23 < ajak> | mattst88: sorry, which question? +11:23 <+robbat2> does the proposed text make the present situation worse? +11:23 <@sam_> i don't think it does +11:23 <@dilfridge> sam_: because I do not intend to get involved into any public discussions touching identity politics. ever. +11:23 <@ulm> robbat2: it depends what "simple records search" means +11:23 <@sam_> dilfridge: you're in the council and you're asking the questions now +11:23 <+robbat2> can we accept an incremental improvement at least? +11:23 <@sam_> dilfridge: that's the same as asking them on the ML +11:24 <@sam_> the difference is it's not productive to raise it last minute +11:24 <@mattst88> ajak: I asked what the analogous situation to Linux/SCO (that brought about their DCO and real-name requirement) would be for Gentoo; e.g. what problem are we hoping to avoid? +11:24 <@ulm> with the clarification s/simple records search/a simple search of government records/ it would be fine with me +11:25 <@ulm> even with keeping the following phrase +11:25 <@mattst88> I'm fine with the changes as they are, but I suspect we're generating lots of discussion over a problem that doesn't need to be solved +11:25 < ulm> | mattst88: same problem as linux? +11:25 <@mattst88> ulm: you're worried people will contribute SCO-owned code to gentoo.git? +11:26 <@ulm> AFAIU the american legal systems, anyone could sue us at any time +11:26 <@mgorny> i would still remove the "notary" part, i don't want to suggest people spend their money on a notary +11:26 < CatSwarm> ulm: [reference] +11:26 < CatSwarm> Contributor Name @ line 183 +11:26 < CatSwarm> https://github.com/kuzetsa/glep/blame/3aa57dd942a58da2c9b50f973fdba184c72d29b4/glep-0076.rst +11:26 <@mattst88> > contributors copy a few lines of a PKGBUILD file into an ebuild? +11:26 < mgorny> | mattst88: remember that glep applies to *all* gentoo projects and repos +11:26 < CatSwarm> https://github.com/kuzetsa/glep/blame/0dcde7c19a93fbc13180e1a0708c83557f3de4e7/glep-0076.rst +11:26 <@mgorny> including projects carrying code +11:26 <@mattst88> mgorny: sure +11:26 < CatSwarm> sorry, duplicate paste. +11:26 < CatSwarm> https://github.com/kuzetsa/glep/blob/GLEP-76.SignOff_kuzetsaCatSwarm/glep-0076.rst +11:27 < CatSwarm> that's the version I signed off on, per blame @ line 183 above. +11:27 <@mattst88> ulm: sure. what's your point? +11:27 <@ajak> surely we've established that changes are needed here? i still don't think discussing the changes forever is a productive use of meeting time +11:27 <@ajak> especially when this is exactly what the mailing lists are for +11:28 <@ajak> producing a new patch isn't what we should be doing here +11:28 <+robbat2> yes, changes are needed here, because prospective developers with unusual names that are legal are being turned down +11:28 <@mattst88> robbat2: I assume I have to email my questions to the discussion thread to get a response from you? +11:28 <@ulm> robbat2: are you referring to that south-indian guy? +11:28 <@ulm> with a single letter name? +11:28 < CatSwarm> I'd like to request if changes are being proposed from the revision I signed off on, please send back to -project ML for discussion, because I don't wish to have the wording attributed to me for things I never said or intended. +11:28 < ajak> | mattst88: yes i think its best to discuss changes to the GLEP on the mailing where they're posed instead of during council meeting +11:28 <@mgorny> robbat2: are they? +11:29 <@ajak> ulm: nobody's referring to anybody in particular, what does it matter? +11:29 <@mgorny> robbat2: if there is such a major problem, why hasn't it been brought up with more specific details? +11:29 <@mattst88> motion to table this and continue discussion on the mailing list? +11:29 <@ulm> ajak: that's the only example I'm aware of +11:29 * | ajak yes +11:29 * | mattst88 yes +11:29 <+robbat2> mgorny: you turned down Catswarn: https://bugs.gentoo.org/606562#c5 +11:29 <+robbat2> *CatSwarm +11:29 <+robbat2> despite their ID +11:29 <@mgorny> robbat2: we all know that's not a legal name +11:29 <@ajak> robbat2: mgorny: let's please continue the discussion elsewhere +11:29 <@sam_> i htink it is +11:30 <@sam_> CatSwarm has been quite clear it's a legalname +11:30 <@ulm> mgorny: they have previous commits in the gentoo repo with a different name, so it's public anyway +11:30 <@mgorny> well, it wasn't last time i've interacted with the topic, which was years ago +11:30 <@sam_> that's besides the point too +11:31 <@mgorny> (and i don't think "CatSwarm" part was there at the time) +11:31 <@sam_> anyway, I'm okay with continuing more discussion on the ML, but this time I thought everything was fine, and now we're debating it all at length because nobody raised their concerns there. what stops that happening again? +11:31 <@ulm> sam_: it's an ill-defined problem, so what do you expect? +11:31 * | dilfridge yes to table it +11:31 <@sam_> ulm: the motion has not changed from the thing raised on the ML? +11:31 <@mattst88> sam_: I think a desire to not do /this/ again :) +11:31 * | ulm yes back to ML +11:31 <@sam_> ulm: so any concerns should've been raised then +11:32 < CatSwarm> mgorny: the legality and uncommon spelling of my name is so rare that even without including a surname, there aren't any other folx named kuzetsa. it's extremely easy to verify that I'm who I say. that nitpick is not constructive or relevant, beyond suggeting bias on your part about what names should look like. +11:32 <@ulm> sam_: I had understood the "records search" part differently +11:32 <@ulm> i.e. meaning any records +11:32 <@sam_> ulm: that's a fair point, but others' comments aren't explained by a misunderstanding +11:33 <@sam_> anyway +11:33 * | sam_ yes (back to ML) +11:33 * | mgorny yes +11:33 <@ajak> gyakovlev: ? +11:33 < CatSwarm> thank you. I don't mind taking my time to respond to any concerns in detail, but if a major rewrite is being considered, please don't edit during a council meeting. +11:33 * | gyakovlev yes (back to ml) +11:33 <@gyakovlev> sorry for delay +11:34 <@ulm> CatSwarm: we cannot do such last-minute changes without posting to the ML again +11:34 <@ulm> by glep workflow +11:34 <@mattst88> I think we have yes votes from mattst88, ajak, dilfridge, gyakovlev +11:35 <@mattst88> moving on... +11:35 * | ajak counts yes's for everyone +11:35 <@ulm> CatSwarm: but nothing prevents us from discussing them +11:35 <@mattst88> Open bugs with council participation +11:35 < CatSwarm> [acknowledged] ~ thank you. I prefer there to be open discussion with time for review / reply / transparency before edits are made. it's easier to have more eyes on it with ML. +11:35 <@mattst88> bug 729062 +11:35 < willikins> | mattst88: 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 +11:35 <@mattst88> just recently reassigned to council@ due to Whissi's retirement +11:35 <@mattst88> not actionable, skipping +11:35 <@sam_> yeah that's fine +11:36 * | ajak nods +11:36 <@mattst88> bug 876921 +11:36 < willikins> | mattst88: https://bugs.gentoo.org/876921 "Vote on GLEP 68 update to 1.3"; Gentoo Council, unspecified; CONF; mgorny:council +11:36 <@dilfridge> we could cc ago +11:36 <@sam_> i'm sorry, i missed that bug +11:36 <@mattst88> I see we have 5 yes votes in this bug^ +11:36 <@sam_> i'll vote on that later unless you need me to now +11:36 <@ajak> dilfridge: table that until open floor i think? +11:36 <@mattst88> dilfridge: want to vote on bugzilla^? +11:37 <@mgorny> if anyone has any questions wrt that update, feel free to ask +11:37 <@dilfridge> done +11:37 <@ajak> sam_ too +11:37 <@ulm> can we close that vote now please, so it'll be in the summary +11:37 <@sam_> ok +11:37 <@mattst88> only other bug is bug 878827 -- which we've addressed already +11:37 < willikins> | mattst88: https://bugs.gentoo.org/878827 "GLEP 78 "Clarifications" update"; Gentoo Council, unspecified; CONF; mgorny:council +11:38 <@mattst88> so I think we're onto Open Floor +11:38 <@ulm> #876921 approved unanimously AFAICS +11:38 <+arthurzam> I'm not sure if this is an existing policy, but can I request Council to mention in every next meeting logs if there was an open vote in a bug that was decided on? +11:39 <@dilfridge> technically we have that policy already, but I'm not sure people remember +11:39 <@ulm> arthurzam: don't we normally do that? +11:39 <@mgorny> arthurzam: i.e. votes since last meeting? +11:39 <+arthurzam> I don't have any evidence it was missed, but please just remember it for future +11:40 <@mattst88> arthurzam: that's a good idea +11:40 <+arthurzam> mgorny: if vote was done in bug and done, just mention in the next available meeting +11:40 <@ulm> and keep the bug open until the next meeting +11:40 <@ulm> so it won't be missed +11:41 <+arthurzam> So dilfridge can document it in the history latex, and we can see it in meeting logs txt +11:41 <+arthurzam> Of course, only mention if it wasn't a hidden vote +11:41 <@mgorny> hmm, i'm actually wondering if something opposite wouldn't be beneficial too -- having a bug for every agenda item +11:41 <@mgorny> rather than a mix of e-mails and meeting logs/summaries +11:41 <@mattst88> honestly, probably so +11:41 <@dilfridge> well, +11:42 <@ajak> and then trackers for each meeting! +11:42 <@dilfridge> e-mails are archived perfectly, so what is there is fixed +11:42 <@mattst88> ajak: perfect :P +11:42 <@mgorny> ajak: and a global tracker for all trackers +11:42 <@dilfridge> just avoid pastebins, attachments, and files on d.g.o public_html +11:42 <+robbat2> we're on open floor now yes? +11:42 <@mgorny> this would be beautiful in bugzilla's dependency tree view +11:42 <@mattst88> robbat2: yes +11:42 <@dilfridge> because after a while thesea re just gone +11:43 < robbat2> | mattst88 had asked: <@mattst88> what would the analogous situation be for Gentoo? +11:43 <+robbat2> and I wanted to be sure an answer was on record +11:43 <@mgorny> ulm: do you want me to push glep updates to some branch, so you'd merge them, or should i push straight to master? +11:44 <@ulm> mgorny: push to branch please, so we follow the four eyes principle +11:44 <+robbat2> the analogous situation for Gentoo exists around *novel* code that might be taken from a propietary source +11:44 <@ulm> although I'm sure it's fine +11:44 <@ulm> mgorny: for both 68 and 78 +11:45 <@mattst88> robbat2: thanks :) +11:45 <+robbat2> and introduced either as patches for ebuilds, or contributions to non-ebuild repos: e.g. portage itself, various patch bundles like kernel, mysql +11:45 <@ulm> mgorny: unfortunately it's hard to extract patches from archives.g.o +11:45 <@mattst88> any other open floor items? +11:46 <@dilfridge> fwiw, the stand proposal for fosdem 2023 is submitted +11:46 <@ajak> dilfridge: do we want an ago discussion +11:46 <@dilfridge> well, we can discuss it but it won't lead anywhere imho +11:46 * | ajak not really sure about the use of such a discussion +11:46 <@ajak> yeah +11:46 <@mgorny> this reminds me, we'd probably have to run glep76 updates through devs hired by certain companies +11:46 <@dilfridge> we've asked ago long ago to opensource his tooling, and he just refuses +11:47 <@mgorny> who may lose ability to contribute while their legal departments take care of proofreading the changed sentence +11:47 <@dilfridge> but I dont know what the current state is +11:47 <@ulm> mgorny: nah, only if we change the certificate of origin +11:47 <@mattst88> do we have any understanding why ago is so obstinate about this? +11:47 <@dilfridge> not me +11:47 <@mattst88> I've only heard "the code is messy" +11:47 <@ajak> ago has recently deleted the github repository where he let people file issues against his "CI", which is a signal to me that he's outright not interested in feedback +11:47 < sam_> | mattst88: I don't think so. what makes it weirder is that he specifically said he wrote his own scripts because he got frustrated by someone years ago not sharing theirs +11:48 <@mattst88> yes, that is how I interpret that as well +11:48 < gyakovlev> | mattst88: no real answer has been given to this q ever. +11:48 <@mattst88> sam_: /o\ +11:48 <@ajak> https://archives.gentoo.org/gentoo-dev/message/b9a014a18091cac95368e15233fe02d2 there's some discussion under this +11:48 <@mgorny> ulm: pushed glep78-update now +11:48 <@mattst88> if there are no other topics, let's close the meeting :) +11:48 <@ajak> https://archives.gentoo.org/gentoo-dev/message/4041549d5af8b6118b1422d59bb2d224 says "To make John Helmert III happy, I just switched to tatt; so my actual workflow is tatt + nattka and there is nothing more." but that's his arch testing stuff, and not the "CI" +11:49 <@ajak> but yeah we can discuss ago later and see if there's anything we can do +11:49 * | mattst88 bangs gavel +11:49 <@sam_> thanks! +11:49 <@mattst88> thanks all +11:49 <@ajak> thank you +11:49 <+robbat2> mgorny, ulm +11:49 <@dilfridge> thanks everyone +11:49 <+robbat2> can we confirm the list of parts you want to change in the glep76? +11:49 <@gyakovlev> thanks, have a nice rest of the Sunday everyone. +11:49 <+robbat2> ulm wants records search to say government +11:50 <@mgorny> ulm: 68 was pushed already (i guess when we had a majority of "yes" votes, since it was blocking media type request) +11:50 <+robbat2> and it sounded like mgorny wanted a revision for the definition of notary being uneven around the world +11:50 <@ulm> robbat2: you said that this was the intended meaning +11:50 <@mgorny> robbat2: not me +11:50 <@ajak> i really think we should move the discussion to the ML +11:50 < CatSwarm> I'm not sure "government records search" even has a universal procedure, unless you're implementing full KYC and ave access to the relevant tool(s) for such things? +11:50 <@ulm> ajak: and we will +11:50 <@ajak> so, please, everyone with concerns about the GLEP, please take concerns to gentoo-project where we've been having this discussion since july +11:51 * | ajak is unlikely to follow discussion here due to lack of structure +11:51 <@dilfridge> it's probably easier to follow this intent by dropping the real name requirement entirely +11:51 <+robbat2> i want to make sure we have at least a partial list of the concerns to discuss +11:51 <+robbat2> i'd love to drop the entire requirement +11:51 <+robbat2> if you'll let us do that +11:51 <@ajak> me too +11:51 <@ulm> I'm strongly opposed against that +11:52 <@ajak> shocker (and this is why i have pushed for the amendment rather than dropping the requirement entirely) +11:52 <@ulm> because I predict that we'll get contributions under completely silly names +11:52 < pietinger> Does anyone know how Linux does the verify of real names ? +11:52 < CatSwarm> dropping it and leaving it to the diligence of whoever does the commits and sign-offs from an @gentoo associated email means the reputation risk, and burden of understanding the situation of the contributor isn't made overly intrusive. +11:53 <+sultan> ulm: We already have silly names ;) +11:53 < CatSwarm> pietinger: yes, I've discussed this with both multiple activists and linux foundation members over the years, as well as policymakers in other orgs with a contributor name policy. +11:54 < pietinger> CatSwarm: Thanks ... and ... How do they do it ? +11:54 <+robbat2> pietinger: obvious fakes rejected, but no other changes rejected that i've seen on public lists +11:54 <@ulm> robbat2: in principle it's all fine with me, with one added word for clarification +11:54 < CatSwarm> pietinger: I'd need to write a formal paper with a bibliography / set of references and footnotes, and can't concisely summarize it in a line or two. sorry. +11:55 <@ulm> because I think it should be made obvious when this is the intention +11:55 < pietinger> CatSwarm: Ok. Thanks. +11:55 < CatSwarm> robbat2: correct ~ it usually has a low bar to clear, being to reject obvious pseudonyms and assume everyone else contributed in good faith. +11:55 <+robbat2> are there other specific concerns to add to the discussion for the mailing list? +11:56 < CatSwarm> the details usually vary by which orgs do full KYC, or signed documentation / affidavit-style documents, etc. +11:56 <+robbat2> so far: 1. records search, 2. notary meaning, 3. dropping the entire requirement +11:56 < CatSwarm> (plus there's a custodian of records in that case) +11:57 <@ulm> 4. people supply their names to the foundation / its successor +11:57 <@ulm> and then they can commit under some handle +11:57 <@dilfridge> let's try to get along with 1 and 2 first +11:57 <+robbat2> CatSwarm: it was discussed much further in the past that we don't want to do full KYC because of the regulatory burden it places on the organization (e.g. securing copies of ID) +11:57 < CatSwarm> ulm: hmm? is there going to be a custodian of documents for that for personally identifiable info? +11:57 <+robbat2> i sure as hell don't want to keep anybody else's ID anywhere at all +11:58 <@dilfridge> ++ +11:58 <@ulm> robbat2: we do that for developers, don't we? +11:58 <@dilfridge> no +11:58 <@ulm> no? pretty sure my name is in ldap +11:58 <@dilfridge> (if with ID you mean a copy of government docs) +11:59 < CatSwarm> I'm strongly in favor of dropping [the name requirement], and leaving it to the diligence of whoever does the commits and sign-offs from an @gentoo associated email means the reputation risk, and burden of understanding the situation of the contributor ~ [author is the term used by git] ~ isn't made overly intrusive. +11:59 <@ulm> only over my dead commit access +11:59 <@dilfridge> " leaving it to the diligence of whoever does the commits and sign-offs from an @gentoo associated email " +11:59 <@dilfridge> ^ why do you assume people will be willing to do that? +11:59 <@ajak> eh, what over your dead commit access? +11:59 < CatSwarm> dilfridge: I'm proposing this as a compromise to full KYC +11:59 <@ulm> ajak: famous quote from an ex-dev :) +12:00 <@ulm> flameeyes, that is :) +12:00 <@dilfridge> CatSwarm: in that case you're only shifting the load to someone else instead of taking responsibility yourself +12:01 * | dilfridge thinks that contributing should also mean taking responsibility +12:01 < CatSwarm> refuting comment: gentoo needs to be responsible for KYC or openly shift the burden in a way that leads to opening themselves up for disparate impact / discriminiation claims, or just come up with a compromise if retaining and safely handling and storing documents isn't being done. +12:03 < CatSwarm> there are ways to eliminate the burden on minorities and vulnerable populations who would be impacted by an overly strict name policy, and they have legal rights internationally, and in more jurisdictions than I could easily count +12:03 < CatSwarm> failure to take a less-intrusive approaching knowing they will be impacted is called disparate impact, legally-speaking. +12:04 <@ulm> CatSwarm: do you agree that the intention of "records search" was to mean "government records search"? +12:04 < CatSwarm> ulm: I didn't want to namedrop google, honestly. +12:04 <@ulm> that doesn't answer the question +12:05 < CatSwarm> if someone's contribs / reputation shows up with a quick google check and stuff like keybase and/or multiple profiles which show a clear history of beign a real identity, that's more than sufficient to confirme that someone is reachable if the presumably copyrightable contribution really was theirs and not stollen. +12:05 <@ulm> I just say that we should clarify it one way or the other +12:05 < CatSwarm> if the intent isn't to be able to confirm the chain of custody for what's being committed, why is it in a copyright-specific GLEP? +12:06 <@ulm> so if the intention is "government" then it should say so +12:06 <@ulm> and if it's google/twitter/whatever then it should say so as well +12:06 < CatSwarm> I never intentionally tried to, or considered that I had unintentionally implied government, so no. +12:07 < CatSwarm> this is really complicated enough that I don't think it falls under the original scope / intendd rationalle behind GLEP-76 +12:07 * | mgorny needs to leave now +12:07 <@mgorny> thanks for the meeting and gn +12:07 <@ulm> CatSwarm: robbat2 apparently understood it differently +12:07 <@ulm> <+robbat2> records search in the context of the GLEP means to me searching government records, not public records like (twitter.com/jesus) +12:07 < CatSwarm> provably chain of custody is sufficient for copyright diligence, at any rate. +12:08 <@sam_> gn +12:08 <@ulm> maybe "public records search" then? +12:09 <@ulm> I mean, we can discuss it for another 6 months, but I guess it won't get us much further +12:10 * | ulm really expected approval today +12:10 < CatSwarm> I'd rather discuss this on ML if there's a specific rationale instead of random gusses for how to implement not-quite-KYC +12:10 <+robbat2> to discuss: is the point of the GLEP: proving chain of custody (e.g. the real X person), or simply that it a real person exists +12:10 <+robbat2> (on the ML) +12:11 < CatSwarm> robbat2: yes please, with a clear intent in the thread title. +12:11 <+robbat2> the DCO itself does NOT provide chain of custody proof +12:14 < CatSwarm> ulm: same ~ I figure perfect is the enemy of good, but if there's rationalles and needs / motives / concerns / expectations which weren't part of the original GLEP-76, I don't feel like this belongs in a discussion about making GLEP-76 more inclusive, as originally proposed: +12:14 < CatSwarm> bug #805575 +12:14 < willikins> CatSwarm: https://bugs.gentoo.org/805575 "GLEP 76 wording: "legal name" requirement is unfriendly to transgender people"; Documentation, GLEP Changes; RESO, CANT; cyber+gentoo:glep +12:14 < CatSwarm> if it went straight to RESOLVED CANTFIX, that's worrisome. +12:15 <@ulm> purely procedural, it should have gone to the ML first +12:15 <@sam_> (stated clearly on the bug too) +12:15 <@ulm> as I've explained in comment #1 +12:15 <@ulm> yes +12:16 < CatSwarm> ah, so RESOLVED CANTFIX only means that it's still an open discussion and being put to discussion, and "can fix" is still the status? +12:17 < CatSwarm> that's a confusing tag. I apologize for my misunderstanding. +12:17 <+robbat2> (maybe NEEDINFO would have been better, info being discussion on the ML) +12:19 < CatSwarm> mmm, NEEDINFO without the RESOLVED [perhaps an in-progress tag] would've been less of a rejection vibe, I think. okay. +12:20 <+robbat2> i have to go for now, and i'll try to raise threads for the 5? things to get clarification on [or finding consensus that the proposed text was good enough] +12:20 < CatSwarm> either way I found it intimidating. +12:20 <@sam_> CatSwarm: to be fair it was very clear if you read the bug +12:20 < CatSwarm> robbat2: acknowledged. I'll be on the lookout for ML +12:20 <+robbat2> bugzilla: you can't use resolutions unless status==resolved +12:21 < CatSwarm> sam_ ~ RESOLVED means something to me other than in-progress / NEEDINFO, and implies rejection regardless of the intent of whoever closed the bug. +12:21 <@sam_> CatSwarm: i think you're reading too much int o the resolution, bugzilla is just very rigid. if you read the bug then you'll see why. +12:21 <@sam_> it's a technical thing based on the fact that we don't make actual edits to gleps based on bugs +12:21 <@sam_> the component on bz is for editorial stuff +12:21 <@sam_> you cannot really analyse things based on the bugzilla resolution in isolation +12:22 <@sam_> you do have to read the actual bug +12:22 < CatSwarm> either way, the one who posted the bug did post to ML and there's already been a previous council meeting on this topic. I think I just need to do a better job watching ML for now if it's being pushed away from bugzilla entirely. +12:22 <@ulm> acually CANTFIX says that it's a valid issue +12:23 <@ulm> otherwise it would have been RESP INVALID +12:23 <@ulm> *RESO +12:23 <@ulm> https://bugs.gentoo.org/page.cgi?id=fields.html#bug_status +12:24 < CatSwarm> dialect / grammar / linguistic understanding of implying something is unfixable is a fault on my end. I recognize that designing the bugzilla workflow wasn't done internally. either way, I apologize for the confusion. +12:24 <+robbat2> i have to run, clean house and cook a roast +12:24 < CatSwarm> good luck with tasty meal, etc. +12:24 < CatSwarm> similar ~ I have non-GLEP stuff ot attend to. |