path: root/meeting-logs/20090625.txt

16:01 <@dertobi123> so
16:01 <@dertobi123> heya
16:01 <+tanderson> hi folks, I'm present
16:01 <+ciaranm> evenin'
16:01 <+ciaranm> <-- dev-zero's proxy, unless he finds internet
16:02 <@leio-dl> sorry, but no
16:03 <@dertobi123> so, first of all: roll-call - who's here?
16:03 <@ulm> here
16:03 <@leio-dl> here
16:03 <+tanderson> dertobi123: are you chairing the meeting?
16:03 < spatz> can you please change the topic?
16:03 <@dertobi123> tanderson: if you'd like to, go ahead
16:03 -!- dertobi123 changed the topic of #gentoo-council to: Next Meeting: now.
16:04 <@Cardoe> now as in right now!
16:04 -!- lavajoe [n=joe@...] has joined #gentoo-council
16:04 -!- dertobi123 changed the topic of #gentoo-council to: Next Meeting: now. Agenda: http://archives.gentoo.org/gentoo-dev/msg_3de4654630dd6805259714833442c4f2.xml
16:04 <+tanderson> ok, we have enough
16:04 -!- mescalinum [n=mescalin@gentoo/developer/mescalinum] has joined #gentoo-council
16:04 -!- ABCD [n=ABCD@wikipedia/ABCD] has joined #gentoo-council
16:04 <+tanderson> first topic, EAPI development/deployment cycles
16:05 <@dertobi123> stop
16:05 -!- drantin [n=drantin@pdpc/supporter/active/Drantin] has joined #gentoo-council
16:05 <@Cardoe> dertobi123: hammer time?
16:05 <@dertobi123> Cardoe: just for the record ...
16:06 <@dertobi123> for the record: dev-zero appointed ciaranm as his proxy. 4 out of 7 council members agreed that proxies must be gentoo developer, just like regular council members. therefore ciaranm isn't accepted as dev-zero's proxy for today.
16:06 <+ciaranm> could you point to where that rule is documented please?
16:06 <+ciaranm> glep 39 imposes only one restriction on proxies, which is that you can't have one person with multiple votes
16:06 <@dertobi123> council members must be gentoo developers. proxies therefore need to be, too.
16:06 <+tanderson> dertobi123: ok
16:06 <+ciaranm> dertobi123: where is that documented?
16:06 -!- Pesa [n=Pesa@...] has joined #gentoo-council
16:07 <@dertobi123> that's what 4 out 7 council members did agree on and this is not going to be discussed now.
16:07 <+ciaranm> dertobi123: glep 39 does not impose requirements upon council members
16:07 <@ulm> ciaranm: custom and practice
16:07 <@ulm> and common sense
16:07 <@Cardoe> Anyway, 4 out of 7 voted and it takes a majority vote to make something a reality
16:07 <@Cardoe> so that's all the justification necessary
16:07 <@dertobi123> exactly.
16:07 <@Cardoe> on to EAPI development/deployment cycles
16:07 <@dertobi123> end of discussion, first topic pklease
16:07 <@dertobi123> Cardoe: *5*
16:08 <+ciaranm> the council's ignoring glep 39 then?
16:08 <+ciaranm> you're going against the direct request of an elected council member here
16:09 <@Cardoe> So who had items to discuss wrt to the EAPI development cycles?
16:09 -!- zhick [n=henning@...] has joined #gentoo-council
16:09 <@Cardoe> cause if no one has anything to discuss...
16:09 <@Betelgeuse> \o/
16:09 <@dertobi123> this topic is split up into development and deployment
16:09 <+tanderson> Cardoe: well, my agenda points to a process that quite a few people acknowledged was a step in the right direction
16:10 <@dertobi123> as for the deployment part ciaranm described a possible way which i'd like to try for eapi-4 development
16:10 <+tanderson> Cardoe: we were sidetracked last meeting unfortunately..
16:10 <+tanderson> for development, we just NeedCode(TM)
16:11 <@dertobi123> tanderson: for implementation, not for eapi-development
16:11 <@Cardoe> tanderson: yep. which points to agenda item 2... talking to zmedico
16:11 <+tanderson> dertobi123: right
16:11 <+tanderson> Cardoe: I can pretty much guarantee that there's been no progress
16:12 < zmedico> well, we got KV taken care of :)
16:12 <+ciaranm> http://bugs.gentoo.org/show_bug.cgi?id=273620
16:12 <@Betelgeuse> For the record I don't think proxies need to be Gentoo developers. But in the end it shouldn't matter today as I don't plan on us having a vote on anything major.
16:12 <+tanderson> zmedico: oops, my apologies ;-)
16:13 <@dertobi123> http://archives.gentoo.org/gentoo-dev/msg_d3a4758c455fded00608e891f525d3cc.xml <- anyone having something to discuss on the development part described there?
16:13 <+tanderson> zmedico: nothing on the substantial things though, right?
16:13 <@dertobi123> tanderson: can we please discuss #1?
16:13 <@dertobi123> +first
16:13 <+ciaranm> i just consulted with the person who wrote the proxy rules. he says that there's deliberately no restriction other than the one vote per person thing.
16:13 <+tanderson> dertobi123: yeah...
16:14 < zmedico> tanderson: no, not really. I've been pretty active working on portage the last week though (unlike the previous month), and I plan to stay pretty active and I should get eapi 3 done after not too long.
16:14 <+tanderson> dertobi123: other than that I in general agree, no
16:15 <@dertobi123> other opinions?
16:15 <+ciaranm> afaik the only disagreement on the eapi stuff was over the codenames
16:15 <@ulm> dertobi123: generally agree too, except that we should use reasonable and descriptive names for features
16:15 <+tanderson> codenames isn't really crucial. It can go away if need be.
16:15 <@leio-dl> ulm++
16:15 <@dertobi123> ulm++, yeah
16:16 <+tanderson> *aren't
16:16 <+ciaranm> the problem with not using codenames is that certain people didn't bother to read pms and just started commenting based upon what they thought something was, not what it was
16:16 <@Cardoe> this is such a nit picky argument that isn't meaningful or technical. it's just personal attacks.
16:17 <@dertobi123> agreed
16:17 <@dertobi123> the codenames should describe what it's about and that's it
16:17 <+ciaranm> it's the best solution i've found to a problem we encountered for EAPI 3
16:17 <@leio-dl> I believe he might have something in mind I might have talked about in the council channel while working through the features outside a meeting
16:17 <+ciaranm> if you've got a better way of ensuring that people read the material they're discussing, please present it
16:17 -!- darkside_ [n=darkside@gentoo/developer/darkside] has left #gentoo-council []
16:17 <@Cardoe> again, completely non-technical in nature
16:18 <+ciaranm> we're discussing process here. half of it's non-technical.
16:18 <+ciaranm> whether or not we use a wiki or google docs is non-technical.
16:18 <@leio-dl> if something was not clear at meeting time, then due to rushing this to the very next meeting, when the material to work through is immense
16:19 <@dertobi123> besides this being a personal thing between leio-dl and ciaranm I don't see any argument why should've nonsense codenames
16:19 <@dertobi123> +we
16:19 <+ciaranm> dertobi123: please present a better solution for ensuring that material has been read before being discussed
16:19 <@leio-dl> the take-away point here is that maybe the meeting 1-2 weeks after EAPI-3 draft shouldn't have it in the agenda yet, but give some time to actually work through things
16:20 <@Cardoe> ok this is just a circular and pointless argument
16:20 <@dertobi123> ciaranm: there is none and we don't need one.
16:20 <@dertobi123> Cardoe: indeed
16:20 <@leio-dl> err, after an EAPI draft is ready
16:20 <+ciaranm> dertobi123: it was a considerable problem during the EAPI 3 process
16:20 <+ciaranm> leio-dl: i don't want to waste time writing draft-quality material for features that definitely won't be accepted
16:21 <@leio-dl> then what kind of a material should be worked through
16:21 <+ciaranm> i want features written in a couple of paragraphs of semi-spec-quality material, then a rough vote on whether to proceed to draft quality material for those
16:21 <+ciaranm> and then a final vote on the draft quality material
16:22 -!- psychoschlumpf [i=lars@unaffiliated/psychoschlumpf] has joined #gentoo-council
16:22 <+ciaranm> but the voting needs to be on those couple of paragraphs and then the real draft material, not just the name
16:22 <@dertobi123> sure
16:23 <@dertobi123> but in the end it's everyone's (council member) responsability to be informed and to know what to vote on
16:23 <@ulm> ciaranm: so there will be a bug # for every feature?
16:23 <@dertobi123> so, if someone screws up by voting just on some random names ...
16:23 <+ciaranm> ulm: roughly, yeah, although we might end up with a shared bug if features are closely related but independently votable (doexample, doinclude for example)
16:24 <@ulm> ciaranm: so use the bug number as your codename if you want, but don't introduce any additional nonsense names
16:24 <+ciaranm> could do that. although then people might just read the bug summary...
16:24 <@ulm> ciaranm: you can't control what people will read anyway
16:25 <+ciaranm> ulm: no, but we can modify the process based upon past experience to reduce the likelihood of problems cropping up
16:25 <@dertobi123> serious, this is not kindergarten. you can't control who will read what and vote based on what he had red before.
16:26 <+ciaranm> you can modify the process to decrease the possibility of abuse
16:26 <@ulm> anyone has new arguments on this point? if not, i suggest that we proceed
16:26 <@dertobi123> ulm: yeah ...
16:26 <@leio-dl> So, simply to make sure everything is worked through (not just mindlessly read through), make sure there is enough time between material to be worked through presented and a vote
16:27 <+ciaranm> leio-dl: there was plenty of time last time around. and people didn't even say "i haven't read it yet", they said "i have objections and questions"
16:27 <@leio-dl> The only past experience perceived bad comes from there. I wasn't just reading through. I was thinking every single point  through and might not have gotten to one fourth of them by the time voting had to happen.
16:27 <@dertobi123> leio-dl: we introduced that requirment lately with requiring an agenda sent out a week before the meeting (though it ended being just a couple of days, but we're making progress on that!)
16:28 <@leio-dl> didn't help when it situated at a time I had no time for gentoo for one week. Anyways, I don't remember all the timeline and I don't care to
16:28 <@leio-dl> lets move on
16:28 <@dertobi123> leio-dl: in that case you should've appointed a proxy *cough*
16:28 -!- NeddySeagoon [n=NeddySea@gentoo/developer/NeddySeagoon] has joined #gentoo-council
16:28 <@dertobi123> i think we reached a consensus on that and can move on
16:29 <@dertobi123> right?
16:29 <@leio-dl> because I was unable to work on gentoo stuff during the time in the middle of two meetings?
16:29 <@ulm> dertobi123: yes
16:29 <+ciaranm> leio-dl: you aren't just a council member for one hour every two weeks
16:29 <+tanderson> This is pretty unproductive
16:29 <@dertobi123> somewhat, but we're making kind of progress
16:29 <@dertobi123> *cough*
16:29 < bonsaikitten> ciaranm: people may have jobs and other things that take up time ...
16:30 <+ciaranm> bonsaikitten: and they are more than welcome to appoint a proxy
16:30 < bonsaikitten> how does that help?
16:30 <@dertobi123> so, next part of the development/deployment discussion is the deployment part
16:30 <@dertobi123> guess that's a topic for the next council and we can move on again.
16:30 <@dertobi123> other input on that?
16:30 <@leio-dl> yes, and I was responsible enough to actually work deeply through all of the points I managed by the time that meeting came on by the time some voting had to happen. It takes time. Hours and hours.
16:31 <@Betelgeuse> In general getting EAPIs specified is a minor concern compared to getting the code into Portage.
16:31 <+ciaranm> leio-dl: no, you were irresponsible enough to raise queries and say "i object to this" rather than "i haven't read this yet"
16:31 <@leio-dl> nope, we just need more portage developers
16:32 <@dertobi123> guys, can we please just follow our agenda???
16:32 <@leio-dl> ciaranm: I'm done with this. Apparently I need to waste my time on re-reading thousands of lines of council channel log to deal with these claims.
16:32 <@ulm> silence please ;)
16:32 <@dertobi123> there's space left for some metadiscussions on #3
16:33 <@dertobi123> so any other comments on the deployment part?
16:35 <@dertobi123> can we at least get a quick overview if the council thinks that describing a deployment process for new eapis is important?
16:35 <@dertobi123> i for one do think we do need such a process.
16:36 <+ciaranm> i think we need one, but the one we've been using seems to work
16:36 <+ciaranm> i'd really rather not see portage to stable without having had main-tree testing of new EAPIs
16:37 <@ulm> but we have no writeup of the current process, right?
16:37 <+tanderson> isn't ciaran's mail a pretty good summary of the current process?
16:37 <+ciaranm> nothing documented in stone that i'm aware of
16:38 <@dertobi123> i don't think the process used for now is perfect. we have the problems i described in my mail and other problems as well. the process i described some weeks ago isn't perfect as well, that's for sure. but in the end i think this is an important topic, but sadly noone seems to be really interested in that.
16:38 <+ciaranm> i don't think we're going to get a perfect process
16:39 <@dertobi123> not if we're not going to improve what we do
16:39 <+ciaranm> i think we'd do best by paying more attention to where things go wrong with EAPI 3 and then addressing those next time
16:39 -!- musikc_mobile [n=musikc@...] has joined #gentoo-council
16:40 <@dertobi123> well, i described problems i've seen going wrong with eapi-2 i'd like to see improved for eapi-3 deployment
16:40 <+ciaranm> with EAPI 2 the big one i saw was portage releasing something that didn't conform to the spec we'd agreed upon and that clearly hadn't undergone any kind of testing
16:42 <@dertobi123> and what i've seen is that people immediately started using eapi-2 features (which isn't bad at all), but in the end packages needed to be backported to older eapis for security bugs
16:42 <@dertobi123> anyway, 20 minutes left. we have to move on.
16:42 <+ciaranm> the unfortunate reality is that for security, occasionally extra work has to be done
16:43 <@dertobi123> ciaranm: security's one example, but that whole topic needs to be discussed on the dev-ml first of all
16:44 <@dertobi123> so, eapi-3 progress
16:44 <@dertobi123> zmedico: please :)
16:44 <+ciaranm> http://bugs.gentoo.org/show_bug.cgi?id=273620 <-- portage's eapi 3 progress
16:45 < zmedico> dertobi123: I estimate it will be done within about a month
16:45 <@leio-dl> ciaranm: Maybe you could help with some of that, btw?
16:46 <+ciaranm> leio-dl: afaik the only code that's realistically shareable between portage and paludis is on the amazingly easy shell stuff that's not worth copying
16:46 <@leio-dl> I didn't have code sharing in mind, I had portage contributing in mind :)
16:46 <@dertobi123> zmedico: ok
16:46 <+ciaranm> leio-dl: there's a reason i gave up on fixing portage a long time ago...
16:46 <@leio-dl> ok, lets leave it at that.
16:47 <@dertobi123> zmedico: seen that arfrever is helping out, you mentioned a recruit 2 weeks ago (iirc?) - how's that process going?
16:48 <@ulm> zmedico: any features where you see particular difficulties?
16:49 < zmedico> dertobi123: I've been getting lots of help from the fellow who filed bug 273620 . that's going very well
16:49 < Willikins> zmedico: https://bugs.gentoo.org/273620 "[TRACKER] sys-apps/portage EAPI 3 implementation"; Portage Development, Core; NEW; s.mingramm@...:dev-portage@g.o
16:49 <@dertobi123> zmedico: cool!
16:50 < zmedico> ulm: not really, seems like it should go pretty smoothly
16:50 <+ciaranm> i was hoping we'd eliminated the "this'll be a pain for portage to implement" stuff early on in the process
16:51 < zmedico> if anything comes up I'll let you know :)
16:52 <@Betelgeuse> Who can be bride to get the features done?
16:52 <@Cardoe> Betelgeuse: stop trying to marry developers ;)
16:53 <@Betelgeuse> Cardoe: Ah yes. Oh well I stopped my night out drinking for this.
16:54 <@dertobi123> ok, thanks for the update zac :)
16:54 <@dertobi123> so we have a couple of minutes left to discuss the past year
16:55 < bonsaikitten> bonus points for trying to streamline and optimize the meeting process
16:55  * NeddySeagoon realises that this is the last meeting of this council and thanks the members for serving 
16:56 <+tanderson> seconded. I want to thank you guys for allowing me to work with you
16:56 <+tanderson> It's been great.
16:56 < Philantrop> tanderson: Slimebag! ;-)
16:57 <@dertobi123> bonsaikitten: well, we made a start with that. having a secretary makes life much easier. making it a requirement to post (and ack!) an agenda some days (preferrably a week) in advance is another improvement, too
16:57 < bonsaikitten> hope to see y'all next year ;)
16:57 <+tanderson> I hope to be able to work with you next year as a member or as a continuing secretary(if so requested)
16:58 <@dertobi123> tanderson: thanks for being our wonderful secretary :) everyone thanks for being on the council as well (and we had lots of members this year *cough*)
16:58 <@Cardoe> You guys will all miss me
16:58 <@Cardoe> I know
16:58 <@dertobi123> one thing i like the next council to improve is to announce a host for the meeting in advance
16:59 <@dertobi123> chairing the meetings quite effective is thing all of us did fail for mostly the whole year.
16:59 <@dertobi123> +a
16:59 <@Betelgeuse> If I find the time I will take a shot at writing a web app for handling agenda etc.
16:59 < bonsaikitten> good point
16:59 <@dertobi123> Cardoe: oh yeah, we will do so :)
16:59 <@Betelgeuse> But likely I will slack.
16:59 <@dertobi123> Betelgeuse: keep it simple. just do it. if a webapp then can make things easier, ok ...
16:59 <+tanderson> dertobi123: np,it's been a great ride
17:00 <+tanderson> why a webapp for what can be done in vim?
17:00 <@Betelgeuse> dertobi123: It should help tacking acks and prioritizin.
17:00 <@Betelgeuse> s/tack/track/
17:00 <@Betelgeuse> tanderson: A single person can handle it but when you need to get input from man.
17:00 <@Betelgeuse> tanderson: y
17:00 <@Betelgeuse> tanderson: Reminders etc would be helpful and can be automated.
17:00 <+tanderson> ok
17:01 <@dertobi123> Betelgeuse: there are some steps between. first get the basics done, then make it bright and shiny
17:02 <@Cardoe> Or you can follow the Windows development model
17:02 <@Cardoe> make it bright and shiny... then when you get some time... make it work
17:02 <@dertobi123> Cardoe: i personally do prefer the other way 'round ;)
17:03 <@dertobi123> so, if there's nothing left we would like to write down for our next council we're done.
17:03 <@dertobi123> thanks again for that somewhat interesting experience during the past year, hopefully we see us again in 14 days.
17:03 <+ciaranm> dev-zero would like you to write down the rules you're following, since they're clearly not glep 39
17:03 <@dertobi123> thanks guys!