17:14:40 #startmeeting 17:14:40 Meeting started Thu Jun 27 17:14:40 2013 UTC. The chair is dondelelcaro. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:14:40 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:15:06 I think listing something as an alternative recommendation is still listing it as a recommendation. The converse really does seem very disingenuous to me. 17:15:50 I haven't read Diziet's mail yet fwiw 17:15:51 The ordering implicit in | is fairly ill-defined even from a technical point of view, let alone more from a more general one. 17:15:57 #topic #681419 Depends: foo | foo-nonfree 17:16:00 vorlon: Yes, my bad, sending it so late. 17:16:50 It's possible that too long of familiarity with alternatives as a way to technically permit non-standard configurations has skewed my mental impression of what an alternative means. 17:17:19 I admit I read "foo | bar" in package relationship fields as "foo (but we won't explode if you absolutely have to have bar for some bizarre reason)" 17:18:01 When you go to a webpage and it offers you login with your [this website] | facebook | twitter | openid that's not less of a suggestion to use openid than [this website] or facebook. 17:18:29 (Here I write | for horizontal concatenation, not a visible vertical bar.) 17:19:00 But maybe this argument should be on-list. 17:19:09 The question is, do you want to think/write/read about it some more ? 17:19:49 just read through the mail now 17:20:05 I don't think I'm persuaded from my original position 17:20:20 Personally, I'd be interested in another round of argument against virtual packages if anyone wants to make it just so that I have both sides fresh in my mind, and then I'm not sure there's any point to further discussion. It's one of those cases where personally I won't be unhappy with either outcome. 17:20:22 if someone finds it useful I can explain why, but AFAICS positions are pretty well staked out at this point 17:20:43 vorlon: It sounds like rra might find it useful. 17:20:48 I think the main argument against virtual packages was that we could get into an annoying situation with lots and lots of fiddly virtual packages? 17:20:55 So I guess that's an invitation to send it by email ? 17:21:16 well, the synopsis of my argument against enforcing use of virtual packages is "sometimes the non-free package is third-party and not up to us to modify" 17:21:37 I'm happy to follow up by mail 17:21:40 Please do. 17:21:43 Oh, thank you -- that point wasn't captured in the initial writeup and I hadn't thought about it. 17:21:59 In that case we can put a dummy intermediate package in contrib if we really care. I don't think we should allow third parties to get us to advertise them like that. 17:22:08 If they want an alternative they should use a virtual package name. 17:22:41 Certainly "people who don't support free software might make this inconvenient" is a very bad argument and is in fact making me quite cross... 17:22:51 But please let us take it to email. 17:23:08 do we currently have both positions written up as resolutions? 17:23:13 I think we should put a time limit on this and then vote on it, though, so that we get it resolved. 17:23:13 dondelelcaro: Yes, now. 17:23:17 cool 17:23:20 Maybe a week or two? 17:23:21 rra: Sure. 17:23:26 That works for me. 17:23:33 yeah, I'm ok with it being two weeks for discussion, then starting the vote 17:23:41 I should warn you that if this goes against me I'm considering canvassing support for a GR. 17:23:44 we can obviously delay longer if absolutely necessary 17:23:50 I think this is a political rather than technical issue, mostly. 17:24:05 well, we could punt it directly to GR if you'd prefer :) 17:24:18 Heh. 17:24:29 * rra is still surprised at how intensely people care about this, but then that was my reaction to the initial policy bug. I'll mention that I originally made this change as a purely editorial change and then backed it out when people objected; I really had no idea that anyone would care. :) 17:24:44 (on the grounds that it's "political", I mean) 17:24:50 vorlon: Oh I see. 17:25:05 I think the constitution gives the TC jurisdiction quite clearly; it's specified in the policy doc. 17:25:10 So there's no problem with the TC ruling on it. 17:25:25 * vorlon nods 17:25:54 If it would help other TC members think about how important this "don't advertise non-free stuff" is I could informally canvass opinion or something. 17:26:10 I don't know if that's really a big consideration for you. 17:26:53 personally, the advertising non-free isn't the big deal for me; I just want to make sure that whatever we do doesn't cause people to have non-free software installed without them explicitely installing it 17:27:09 I'm really not persuadable on this. I think all of the solutions you propose for avoiding listing non-free packages in Recommends are perfectly fine, *but* that the individual maintainer should have the latitude to decide whether they're correct / maintainable for their case 17:27:18 dondelelcaro: Then you disagree with me and should vote for the "A" version, I think. 17:27:35 I Individual 17:27:38 and that "avoid all mentions of non-free packages by name" is not something we should enforce at the policy level 17:27:39 err, sorry 17:27:45 vorlon: I disagree. 17:27:52 Diziet: I'm aware ;-) 17:28:09 but yeah, I think this is something that is ok for us to disagree on, so long as we've hashed out our positions 17:28:22 I don't think talking about individual maintainer latitude helps. Maintainers don't have the latitude to compromise other political principles established by the project as a whole.... 17:28:29 Right. 17:28:51 The question is really what the political principle is that we are as a project establishing. 17:29:16 AFIACT those of you who disagree with me think that this isn't really a matter where our principles are at stake. 17:29:29 Which is another way of saying that your view of our principles permit this, but mine don't. 17:29:43 Ultimately we're not going to get all that much further I think. 17:29:52 exactly 17:29:56 But if rra wants to think about it some more that's fine. 17:30:01 Another couple of weeks won't hurt. 17:30:39 No, that's fine, I think this is all the discussion I really needed, actually. 17:30:57 Also I think vorlon's out-of-Debian packages situation needs to be answered in my B version of the resolution. 17:31:00 Right. OK good. 17:31:15 vorlon: So I'd appreciate it if you wrote it up so I can rebut it :-) 17:31:26 I'm with vorlon and dondelelcaro that the advertising thing isn't a big deal to me, but on the other hand I don't think using virtual packages is a major hardship and it *is* a big deal to other people, so increasingly it's making sense to me to respect that. 17:31:51 I do think you said something earlier that was critical, namely that you don't think this should be RC. 17:31:54 Diziet: ack :) 17:31:58 Thanks. 17:32:05 rra: Yes, I think that may have not been clear earlier. 17:32:06 Hi. sorry for being late 17:32:09 I'll note that the option B that's in cjwatson's earlier writeup does say that it's RC (well, not directly, but that's the implication). 17:32:22 The main->non-free direct without alternative clearly has to be RC, so there is a distinction. 17:32:23 I'm a lot more comfortable with requiring virtual packages if it's not RC. 17:33:00 rra: In general we don't make things like this RC I think. And I'm happy with it not being RC as a compromise. 17:33:31 Well, it does mean more complexity in the policy wording since that part has to be a should instead of a must, but that's certainly fine. 17:33:38 Right. OK. 17:33:45 I think those were the two items on the agenda then. 17:34:00 Oh wait New ctte member. 17:34:05 who has the ball on ^^? 17:34:05 One thing that's been obvious from all of this is that Policy wasn't nearly verbose enough on this point, leaving multiple people with completely different impressions of what it was trying to say. 17:34:14 Bdale, hence my point earlier. 17:34:31 vorlon: On the non-free-|, you to write up your out-of-Debian packages, thing, me to respond and then I'll call a vote with Colin's and my texts. 17:35:01 vorlon: if that's ok, feel free to #action yourself. ;-) 17:35:10 #action Diziet On supermajority, me to post GR-starting resolutions for us to vote on. I think I'll give up on trying to let us accept amendments. 17:35:21 rra: bdale> right, thought so 17:35:55 #action vorlon to write up counter-argument for virtual packages wrt out-of-Debian packages 17:36:13 #topic #685795 New ctte member 17:36:26 #action Diziet to respond re non-free-out-of-Debian packages, and call for vote on both versions of TC resolution 17:36:30 this one is just waiting for bdale, who is understandably occupied with other things 17:36:41 Quite. 17:36:49 #topic Next Meeting? 17:36:53 I think the question is if we want to reassign or just wait. Personally, I'm okay with waiting; it doesn't feel critical. 17:36:59 Right. 17:37:02 I agree. 17:37:02 I'm ok with waiting 17:37:10 * aba agrees with waiting 17:37:11 fine with me 17:37:19 #agreed wait for bdale on new member 17:37:22 Okay, that sounds like consensus. 17:37:39 Next meeting same time, July 25th? 17:37:41 the next meeting is currently this time on July 25th 17:37:46 is that ok with everyone? 17:37:49 Works here. 17:37:56 seems ok here 17:38:01 wfm 17:38:06 and the meeting after that at DebConf? :) 17:38:08 I believe that bdale was planning a BoF during debconf 17:38:20 who's going to DC13? 17:38:23 yeah; I won't be there, but I'll try to participate remotely if it works 17:38:25 I am; cjwatson is 17:38:26 vorlon: I will be. 17:38:26 I can probably participate via IRC at least. I won't be there in person. 17:38:31 ok 17:39:15 #topic Additional Business 17:39:21 vorlon: I plan to 17:39:42 anything else to discuss? 17:39:51 aba: great :) 17:39:58 dondelelcaro: nothing else from me 17:40:03 Nothing here. 17:40:09 Thanks everyone. 17:40:18 #endmeeting