21:00 <@ grobian> dah, ok, I started roll call 21:00 -!- grobian changed the topic of #gentoo-council to: meeting now | agenda: http://dev.gentoo.org/~grobian/agenda-20120814 | http://www.gentoo.org/proj/en/council/utctolocal.html?time=1900 | http://www.gentoo.org/proj/en/council/ 21:01 <@Betelgeus> \o/ 21:01 <+scarabeus> | 21:01 <+scarabeus> damn 21:01 < WilliamH> I'm here. 21:01 <@ Chainsaw> Present. 21:01 <@ grobian> yeah, only ulm and dberkholz are missing 21:01 <@ grobian> so let's wait a bit 21:01 <@ Chainsaw> ulm was here earlier. 21:01 <@ Chainsaw> Perhaps we can give him a call? 21:02 <@Betelgeus> I can 21:02 <@ Chainsaw> It is appreciated. 21:02 <@ ulm> here 21:02 <@ grobian> good 21:02 <@ Chainsaw> Just dberkholz then. 21:02 <@Betelgeus> then call donny 21:03 <@ grobian> right 21:03 <@ grobian> let's wait 2 more minutes 21:03 <@ Chainsaw> I don't have numbers I'm afraid. 21:03 <@Betelgeus> Went into some weird google voice answering machine 21:03 <@ Chainsaw> Just the ones that replied to the posting this year. 21:04 <@ Chainsaw> Betelgeuse: Well, thanks for trying all the same. 21:04 <@ grobian> yeah, much appreciated 21:05 <@Betelgeus> donnie gets a slacker mark right? 21:05 <@ grobian> yup 21:05 <@ grobian> ok, people, we have only one item 21:05 <@ Chainsaw> EAPI5? 21:05 <@ grobian> if you haven't yet, please load the agenda (see topic) 21:06 <@ grobian> so, there's two things from my point of view 21:06 <@ grobian> the original idea was to vote on items for EAPI5 that are ready to go NOW 21:06 <@ Chainsaw> I am opposed to having EAPIs in flux. So we can vote now on making what's ready EAPI5. 21:06 <@ grobian> however, the extensive list that ulm sent out needs some thorough research, of which I did a bit, included in the agenda 21:06 <@ Chainsaw> Or we can postpone. 21:06 <@ grobian> so, point one is, do we want to discuss any of it this meeting 21:07 <@ Chainsaw> Looking at the list, what's ready is... not a long list. 21:07 <@ grobian> 1 or two entries 21:07 <@ grobian> with my not-so-thorough research 21:07 <@ grobian> point 2 would be to vote on the features to be in or out 21:07 < WilliamH> My vote would be to postpjone. 21:07 <@ grobian> but honestly, I'd like to defer 21:07 <@ ulm> some of the things marked as not implemented are trivial 21:07 < WilliamH> postpone 21:07 <@ grobian> ok 21:08 <@ grobian> let's vote for postponing the EAPI5 thing 21:08 <@ Chainsaw> Postpone please. 21:08 <+scarabeus> I agree we should only pick from those that are ready 21:08 <@ ulm> yeah, let's postpone it to September 21:08 <@ Chainsaw> And if the trivial things could then be implemented and marked so that there is a more impressive list... 21:08 <@ Chainsaw> That would be great. 21:08 <@ grobian> scarabeus: Betelgeuse what do you vote for? 21:09 <@ grobian> postpone or handle it now? 21:09 <@Betelgeus> grobian: postpone but we can use the time to talk about things 21:09 <+scarabeus> it can be postproned by all means; but we can talk about it 21:09 <@ grobian> ok 21:09 <@ Chainsaw> *nod* Sounds fair. 21:10 <@ grobian> well, that is fine, although I don't feel much need to discuss it here 21:10 <@ grobian> because it needs some more research from my side 21:10 <@ grobian> I'd prefer good follow up's on ML 21:10 < WilliamH> Same here. 21:10 < Arfrever> You can discuss things already implemented in Portage. 21:10 <@ grobian> so I suggest to continue the agenda, and then you lot can chat what you want 21:10 <@ grobian> Arfrever: I'm going to ignore your suggestion here 21:10 <@ grobian> sorry 21:11 -!- ulm_ [~ulm@gentoo/developer/ulm] has quit [Ping timeout: 244 seconds] 21:11 <@ grobian> So, there are no open bugs with council involvement 21:11 <@ grobian> done 21:11 <@ grobian> open floor 21:11 <+scarabeus> I was asked by johu for one thing 21:11 <@ grobian> anyone (everyone) who wants to raise some issue 21:11 <@ grobian> Arfrever: if you want, go ahead here 21:11 *** Chainsaw activates the microphone 21:12 *** grobian listens 21:12 <+scarabeus> if we could change "dev should use latest eapi when bumping" to "dev must use latest eapi when bumping if not forbidden by eclasses" 21:12 <@ grobian> deafening silence 21:12 <@ ulm> patrick has asked me to bring up deprecation of EAPI 1 21:12 <@ Chainsaw> scarabeus: I have ebuilds where that's unworkable. 21:12 <@ grobian> scarabeus: rationale? 21:12 <@ Chainsaw> scarabeus: Example: net-misc/bird 21:12 <@ ulm> what's the general opinion about it? 21:12 <@ Chainsaw> scarabeus: Reason: awkward build sequence. 21:12 <@ Chainsaw> ulm: Rationale? 21:12 <@ grobian> ulm: again? wasn't that horse beaten to death already? 21:12 <@ Chainsaw> ulm: Again, I have ebuilds where it simply can't be done. 21:12 <@Betelgeus> Chainsaw: new phases could always be no-ops? 21:13 <@ Chainsaw> Betelgeuse: That sounds incredibly awkward to me. 21:13 <@ Chainsaw> Betelgeuse: I will, where feasible, bump EAPI on my ebuilds. 21:13 <@ ulm> he gave no rationale 21:13 <@ Chainsaw> Betelgeuse: Where it isn't, I don't see adding empty phases as a particularly acceptable workaround. 21:13 <@ ulm> only a plan 21:14 <+scarabeus> problem is that quite few devs just cp without checking on it 21:14 <@ ulm> my plan would be: deprecate eapi1 "now", add repoman warning, make repoman warning fatal in 3 months (not many ebuilds anyway) 21:14 <@ ulm> then plan to deprecate eapi2 in, say, 6 months 21:14 -!- kallamej [~kallamej@gentoo/developer/kallamej] has quit [Quit: leaving] 21:14 *** Chainsaw wonders where this obsession to remove backward compatibility comes from 21:14 <+scarabeus> - if( !pResultBitmapEx ) 21:14 <@ Chainsaw> Upgrade paths are going to *suck*. 21:14 <+scarabeus> http://qa-reports.gentoo.org/output/eapi_usage.txt 21:14 <+scarabeus> this 21:15 <+scarabeus> it is 550 pkgs 21:15 <+scarabeus> and you can see eapi3 is more candidate for killing than eapi2 21:15 <@ Chainsaw> scarabeus: Yes, net-misc/bird is one of the EAPI1s listed there. 21:15 < Arfrever> Let's discuss deprecation of EAPI="0". Any opinions? 21:15 <@ ulm> was about 800 few months ago 21:15 <@ ulm> Arfrever: upgrade path 21:15 <@ grobian> all EAPI=2 can be sedded to EAPI=3 21:16 <+scarabeus> they are mostly killed with stabilisations 21:16 < Arfrever> ulm: All sys-apps/portage ebuilds already use EAPI="2" or EAPI="3". 21:16 <@ grobian> Arfrever:you can't, so not really a fruitful suggestion , is it? 21:16 <+scarabeus> I am all for kiling it but it must forbid adding new ebuilds of such eapi, not just stop commit if it is in the folder 21:16 <@ Chainsaw> Arfrever: Again, obsession with the eradication of backward compatibility. Why? 21:16 < Arfrever> Chainsaw: Simplification of eclasses. 21:17 <@ ulm> Arfrever: not true 21:17 < Arfrever> ulm: What is not true? 21:17 <@ ulm> ulm: All sys-apps/portage ebuilds already use EAPI="2" or EAPI="3". 21:18 <@ grobian> scarabeus: I think it would help if you'd flesh this proposal out a bit more, it doesn't sound too bad to me to have a policy to only add new stuff of the latest EAPI, for as long as EAPIs don't have orthogonal features 21:18 <@ ulm> portage-2.1.6.7_p1.ebuild is still EAPI 0 21:18 < Arfrever> ulm: There is no such ebuild in gentoo-x86. 21:18 < Arfrever> ulm: OK. grep failed to find it. 21:19 -!- kallamej [~kallamej@gentoo/developer/kallamej] has joined #gentoo-council 21:20 < Arfrever> ulm: Another rationale: Upgrade path is required only for 1 year. EAPI="2" was added long time ago. 21:20 <@ grobian> scarabeus: and a ML discussion would be a good thing too 21:21 <+scarabeus> you are right 21:21 <+scarabeus> johu: start a chat on -project :-) 21:21 <+scarabeus> delegation at its best 21:22 <@Betelgeus> no -project? 21:22 <@Betelgeus> on -dev? 21:22 <+scarabeus> ah right 21:22 < Arfrever> Another item for discussion: What happens when profiles/eapi uses EAPI >=2. I noticed that nothing happens in overlays which use new EAPI. 21:23 <@ grobian> Arfrever: that sounds like something that should be discussed with people into the subject first 21:23 <+scarabeus> grobian: about that sed, I would ask diego to run it through tinderbox to see nothing really exploded, ie python eclass change behaviour with each eapi and such tiny funny things 21:23 <+scarabeus> also I am all hands for blocking additions of new eapi1 ebuilds 21:23 <@ grobian> scarabeus: I don't see the use of removing EAPIs 21:24 <@ Chainsaw> scarabeus: Great. Will you be porting my EAPI=1 net-misc/bird then? 21:25 <@ Chainsaw> scarabeus: And yes, that is a challenge. And no, QA warnings are not acceptable. 21:25 <@ grobian> I can imagine python eclass not supporting EAPI 0,1,2 21:26 <@ grobian> but that doesn't mean those eapi's should be banned 21:26 <@ grobian> an eclass can require its consumers to be using an up-to-date eapi 21:26 <@ grobian> it's one herd anyway 21:26 <@ Chainsaw> They were EAPI=4 holdouts for a long time. Seems fine. 21:28 <@ grobian> ok, can we conclude the open floor with this? 21:28 <@ Chainsaw> grobian: Just waiting for scarabeus to see the EAPI=1 net-misc/bird conundrum. 21:28 <@ grobian> Chainsaw: does it matter for the summary? 21:28 <@ grobian> I guess it does 21:28 <+scarabeus> Chainsaw: i will ask nic guys to fix build system, simple 21:29 <@ Chainsaw> scarabeus: Perhaps you speak their language, that may help. I did try. 21:29 < Arfrever> Any opinions about handling of icons (*.ico) in dohtml? 21:29 <@ Chainsaw> scarabeus: (Just like we asked for a Quagga-style telnet multiplexer) 21:30 <@ grobian> Chainsaw: I did my best to summarise that one 21:30 <@ Chainsaw> grobian: Okay, thanks :) 21:31 <@ grobian> good, then I'd like to close the open floor 21:31 <@ Chainsaw> grobian: Please proceed. 21:31 <@ grobian> next meeting 11 september 21:31 <@ grobian> most likely I won't make it 21:31 <@ grobian> so ulm will be your chairmaster 21:31 <@ ulm> grobian: your summary says 14 september 21:32 <@ grobian> sorry, 14 september 21:32 <@ grobian> yeah 21:32 <@ grobian> will fic 21:32 <@ grobian> fixed 21:32 <@ grobian> 11 ot is 21:32 <+scarabeus> ack 21:32 <@ grobian> ok, everybody happy? 21:32 <@ ulm> 11 is a tuesday, right 21:32 <@ grobian> I'll send out the editted agenda for review 21:32 <@ grobian> ulm: yes 21:33 -!- ivan\ [~ivan@unaffiliated/ivan/x-000001] has joined #gentoo-council 21:33 <@ ulm> grobian: thank you for chairing 21:34 <@ grobian> ok, let's close the meeting then 21:34 -!- Chainsaw [~chainsaw@gentoo/developer/chainsaw] has left #gentoo-council ["Thanks grobian."] 21:35 -!- grobian changed the topic of #gentoo-council to: next meeting: 2012-09-11 19:00 UTC | http://www.gentoo.org/proj/en/council/utctolocal.html?time=1900 | http://www.gentoo.org/proj/en/council/ 21:35 <@ grobian> thanks all for your delightful input ;)