Oversight Board/2010/Meeting Log-2010-06-04

<walterbender> good morning all
who from SLOBs is here for the meeting?
<cjb> morning
<walterbender> hi cjb
cjb: I was just in Kendall Sq., meeting with Mike Dawson...
<cjb> oh, cool
<walterbender> cjb: but now I am back in Newton after a sprint home on my bike :)
cjb: his eXe work is impressive...
<cjb> yeah, Bernie sounded very impressed too
<SeanDaly> hi everyone
(need to sprint on my new electric bike in an hour ;-)
<walterbender> hi SeanDaly
we need one more for a quorum
=-=:mchua_afk is now known as mchua
  • mchua:is here!
<mchua> sorry, I woke up very late today.
we have quorum. :)
walterbender: quorum!
<walterbender> yeah...
#startmeeting
<meeting> Meeting started at 11:11 UTC. The chair is walterbender.
Commands Available: #TOPIC, #IDEA, #ACTION, #AGREED, #LINK
<walterbender> Welcome all...
I'd like to start by quickly reviewing the list of Local Labs so we can officially offer them use of the TM
#TOPIC local labs
or not, if we don't think they should...
<mchua> As per Trademark procedure, don't they have to contact us with a request first? (in other words, should our action here really be "ping local labs and say "hey, if you want to use our trademark, please submit a request as per this page!")?
<walterbender> mchua: they have all contacted us, but put on hold for over a year...
and .py has also contacted us...
<mchua> Oh! okay, wasn't aware. Where can we see the requests queue?
<walterbender> we granted provisional use until the TM policy was sorted out...
mchua: buried in emails... I can try to dig up all the emails and present them to SLOBs... maybe best to do it that way?
so we can defer until I've done that homework, if you'd like.
  • mchua:nods, it'd be nice to have them queued up for quick voting.
<walterbender> #ACTION: Walter to gather various requests from local labs for use of our TM in one place
OK, we'll get to this next time.
<mchua> do we have trademark@sl.o alias --> slobs? or something that'll help avoid burying walterbender's inbox in the future?
<SeanDaly> we need to formalize who gets to be a local lab (aside from first-come first-served)
<mchua> #INFO: Local Labs trademark item deferred 'till next meeting when we have a TM request queue
SeanDaly: +1
<SeanDaly> we've discussed it before; it could be as simple as a SLOBs vote
but contact a prerequisite in any case
<walterbender> SeanDaly: well, we have a set of guidelines and no policy of exclusion (Moodle for example restricts it to tw per country)
<SeanDaly> walterbender: that's true
in my mnd a loca lab was national but really no need to restrict I can think of; could wel be by city?
i l l
<walterbender> #link http://wiki.sugarlabs.org/go/Local_Labs
  • SeanDaly:missing letters
<walterbender> the question, related to Topic 3 today, is the degree of control and oversight we want to administer
<mchua> SeanDaly: I think submitting for a SLOBs vote is a great Local Lab approval proposal - that's going to scale up for a while
<walterbender> but maybe we should defer that discussion until later.
<mchua> walterbender: actually, what is the rest of the agenda today, for reference?
<mchua> Hey, SMParrish_mobile - long time no see.
<walterbender> mchua: 2. Dev Team; 3. Certification
It seems that without Bernie and Tomeu we are not well positioned to discuss #2
So maybe this is a short meeting...
<walterbender> But let's switch to that topic for a moment, since we've agreed to defer on Topic #1?
<mchua> I think #2 is urgent enough that we should state the problem and then find a way to start bringing it up on list, but that's just me.
walterbender: +1
<walterbender> #TOPIC Devel Team vacancies
<mchua> So, this makes me more nervous than anything else - because without continued maintenance of our codebase, we have a shaky foundation for everything else we do to build upon.
<walterbender> We have to important vacancies to fill: team leader/coordinator/facilitator and release manager
<mchua> The growth we're seeing - deployments, local labs, Soas, the marketing being done, the people being reached - they depend on us having code that functions, and... we're hurting for people.
<walterbender> cjb: did anyone throw their hat in the ring?
  • mchua:nods. How soon is the next release supposed to come out? We don't have a lot of time.
<walterbender> mchua: not until Sept/Oct time frame
<cjb> walterbender: nope
<walterbender> cjb: I was afraid of that...
Maybe time to do some behind the scenes coercing :)
<cjb> yeah. well, it might be partly because we backed off from the idea in the e-mail replies to the thread.
<mchua> Probably. :) There's a limited number of people in the world that *can* fill those positions right now, as I see it
because some experience with the Sugar codebase is required (I would assume).
...or perhaps that assumption is wrong?
<cjb> yes, pretty much
<walterbender> mchua: agreed, although one could be release manager without knowing the code itself, just the processes
<cjb> hm, a while ago it sounded like m_stone could be persuaded to have a shot at release manager
<SeanDaly> if tomeu were here, he would say: we need someone experienced, who knows the open source way, and does not need lots of briefing to get up to speed (he will correct me if I err)
<cjb> everyone should beat up on him until he says yes :)
  • mchua:does recall, though, that gregdek did step up to lead the engineering team about 1.5 years ago, and seemed to manage that well even without being a Sugar dev.
<walterbender> SeanDaly: yes... the process, but not the details of the code
<mchua> So maybe we mostly need someone that groks the open source way and how to wrangle engineering, and is willing to learn more about Sugar over the course of the release.
<CanoeBerry> Listening :)
<walterbender> The release manager should be about keeping things moving, but not imposing opinions about content... IMHO
that is for the devel team to do collectively
<cjb> I don't really agree, but having a release manager is probably more important than agreement about technique :)
<walterbender> cjb: please expand...
<cjb> (I think the most important job the release manager does is decide whether a late change constitutes acceptable risk, and I think doing that requires deep understanding of programming and the complexity of a given bug/solution.)
<walterbender> cjb: yes, I agree with that...
and to really understand risk, they will need a deep understanding of Sugar and/or a good process for getting input from those that do...
<cjb> yeah
but I'd rather have a release manager who's enthusiastic and willing to ask other people for advice than no release manager at all :)
<walterbender> cjb: +1
<mchua> All righty. So, who's finding one?
<JT4sugar> Have you thought of asking Satelit he is retired and may be a good fit with the recent work he has been doing
<mchua> And what's our fallback if we don't find one by $date? Skip a release?
<walterbender> JT4sugar: interesting idea...
mchua: I don't know how we could release without a release manager... so we'll have to adjust our schedule accordingly
<mtd> cjb: I'd rather not have such a release manager - a bad release is worse than no release.
  • mtd:thinks from the peanut gallery
<mtd> cjb: m_stone would be great
<mchua> mtd: I agree - so I would reply to walterbender's comment that, if we want an October/November release, then if we don't have a release manager by... say, start of July, we should skip this release, slip 6 months.
  • mtd:finishes scrollback, thinks m_stone beating should occur.
<cjb> can we get a volunteer to attempt to persuade m_stone that we think he would be awesome and he should sign up? :)
<mchua> ...but that's just my opinion.
<mtd> cjb: does he not know we think he is awesome?
<cjb> mtd: he's had some unsatisfying experiences with us recently
<walterbender> so let's ask m_stone ( cjb ) and satellit_ ( JT4sugar )...
<mtd> cjb: I told him that a week ago, otherwise I would volunteer to do so again.
<cjb> like patches being ignored
<mtd> s/ again//
cjb: I nominate you to tell him again, using force if necesary.
mchua: yeah
<walterbender> perhaps satellit_could shadow m_stone and be ready to do in the next cycle, so we can grow a crop of people ready to fill the role
and erikos has volunteered to help whomever we find...
  • SeanDaly:bio crop, no artifical preservatives
<cjb> ok
  • mchua:fires out a quick email to m_stone saying "please do consider the release manager position"
<cjb> mchua: thanks
I think it's good to have him hear it from multiple people he respects
<walterbender> I'll make sure he gets a copy of the meeting log as well :)
and satellit_is perhaps listening in right now :)
any other ideas for the moment?
<mchua> rkabir?
<satellit_> * I am and am unsure I am qualified for something like this
<mchua> he certainly has the PM chops - just not the Sugar-specific experience, but this might be the sort of concrete job description he needs to get started.
  • mchua:shoots off an email to rkabir as well
<walterbender> satellit_: but would yu shadow so that you can learn the ropes?
<SeanDaly> satellit_: your seriousness & commitment & contributions are appreciated
<satellit_> possibly, but I am uncomfortable with managing this kind of position....
thanks
<walterbender> satellit_: OK..
shall we use our last 15 minutes to discuss Certification?
<SeanDaly> just a question, is it taboo to discuss 6-month release cycle? would a breather be out of the question? My concern is we are too out in front of our deployments
<mchua> Yep. And check in on dev team vacancies again next week, I'd say.
<SeanDaly> that said, fabulous 0.88 work on XO-1.5 coming up
<mchua> SeanDaly: ...actually, that may not be a bad assumption to consider - should we do that here, or on the devel list?
<walterbender> +1 to devel
<SeanDaly> if it's not taboo, on devel list seems appropriate
<walterbender> not taboo... the only thing that is taboo is double-click
<SeanDaly> :D :D
<JT4sugar> A set of emails laying out what we are looking for in Release Manager and Team Leader should be resent to all lists, including OLPC lists it's our best way to advertise these openings-not sure Community quite understands situation
<SeanDaly> another angle is: next release have features lined up, or working on collab reliability, or what?
<walterbender> #Action Everyone to turn over more "stones" looking to fill the open positions
<cjb> SeanDaly: we don't know what the features will be, because of aforementioned lack of release manager and team lead :)
<mchua> SeanDaly: Or sustainability of having enough people in the future to continue running a good development team? /me always thinks about capacity-building first
<walterbender> mchua: that is the meta question...
<SeanDaly> cjb: I was wondering where feature set came from...
<mchua> I would propose that we set a deadline of July 1st to find a release manager for an Oct/Nov release.
And if we don't have one by then, we slip the release approximately 6 months, and bring up these other questions (release cycle length, etc) to consider.
<walterbender> SeanDaly: we discuss features at design meetings and have a feature proposal process
<mchua> Since in the absence of dev team leadership, I believe it falls to SLOBs to make sure the engineering moves forward in *some* clear way (even if that way is "we are going to be explicit about deferring our work until we have enough capacity to do it.")
<walterbender> mchua: I agree in principle... have to look at the calendar tho.
<CanoeBerry> total aside: i will be ramping the sri lankan testing team back up on Monday, and will appreciate "educated" opinions on which XO-1 an XO-1.5 builds we should focus on
<SeanDaly> walterbender: ah
<walterbender> CanoeBerry: I'
d focus on cjb's latest for xo 1.5 and .py's for xo 1.0
<mchua> CanoeBerry: Will they be on the olpc testing mailing list? I need a single place to send QA efforts towards.
<CanoeBerry> mchua: yes
<mchua> There's OLPC build testing, sugar-core testing, Activity testing, SoaS testing... they all overlap, etc.
CanoeBerry: Excellent, I'll see them there then.
<CanoeBerry> http://lists.laptop.org/listinfo/testing
<walterbender> we are running out of time... can we switch topics?
?
<cjb> sure
<CanoeBerry> mchua: we'll do our best to post most/a lot of the sri lanka volunteers' conf calls to testing@l.l.o
<walterbender> #TOPIC Certification
This is related to Topic #1...
has everyone seen rgs_ 's email?
<CanoeBerry> which?
<mchua> Link?
<walterbender> to summarize, he'd like to use the SL TM on certificates he gives to people who took his devel course
  • walterbender:gets the link from the agenda page...
<SeanDaly> "officially recognized Sugar developer"
<walterbender> #link http://lists.sugarlabs.org/archive/iaep/2010-May/010965.html
SeanDaly: that is where we are heading...
SeanDaly: I think we want high standards, but I also think that we should ultimately let the local labs we have endorsed have the final say...
<cjb> sounds good
<walterbender> so our role would be to certify the local lab... and let them certify the course...
<SeanDaly> walterbender: I can't imagine how we could certify individuals - a tes with N°2 pencils?
walterbender: so working with the local Lab a necessity I think
<mchua> I'd like to know more about the standards they're using for certification - waht does it mean, what can those people do, what assurance do we have of quality of those that are getting certified?
<SeanDaly> certifications have worked historically in the industry when large players (Novell, Microsoft, Oracle etc.) want to convince potential employers that there is a sufficiently large pool of employable talent
<walterbender> mchua: we should develop guidelines but at some level I think assurance has to be a local thing...
<SeanDaly> rgs mentioned something interesting: he proposes cert if Activity comleted
<walterbender> mchua: I think the breadth and depth of the .py course, at least on paper, is pretty good
<SeanDaly> so perhaps posting it on ASLO could be part of a cert process
<JT4sugar> walterbender, You might want to capture the course and certify that as well. Since Bernie was part of training team in this case certificates seem appropriate
  • mchua:also thinks that any certified curriculum should be open-content and publicly posted somewhere in the SL universe (or linked-to from it)
<walterbender> mchua, SeanDaly yes... how does one know if the course was passed... a working activity is a pretty strong proof...
mchua: .py materials are open content... on the wiki
mchua: the link is in the email
we are just past noon EST... do people need to go?
<CanoeBerry> what remains?
<SeanDaly> we could insist that a completed Activity follow guidelines and be distributed under a free license, if that's not too shocking
<walterbender> CanoeBerry: reaching consensus re Certification
  • SeanDaly:needs to go shortly
<walterbender> SeanDaly: it cannot go on ASLO if it is not under a free license :)
<mchua> walterbender: Yeah, just noting that it makes a good requirement for later cases :)
walterbender: is there a motion we can vote on?
<walterbender> well, good input. I can write these suggestions into a motion but to do it justice, let's prepare it for next week?
mchua: can we meet next week or will the workshop schedule intervene?
#Action Summarize the Certification discussion into a motion
Could we meet a bit later on next Friday? After 3PM EST?
<mchua> walterbender: I can do that.
<walterbender> mchua: that being the summary :)
<mchua> er, uh... meeting after 3pm EST Friday.
  • mchua:is totally hosed for the next 2 weeks
<SeanDaly> after 3pm EST should be doable for me, but not before (commuting/supper)
  • walterbender:is total hosed... that is life :)
<walterbender> #agreed meet next week (11 June) at 3pm EST, 19UTC
<cjb> ok
<walterbender> anything else? /me has to take a call in a few minutes...
<mchua> I think we're set.
Good meeting. :) We should keep moving them along at this clip.
<SeanDaly> sure, we just need to avoid controversies :D
<walterbender> #endmeeting
<meeting> Meeting finished at 12:14.
Logs available at http://me.etin.gs/sugar-meeting/sugar-meeting.log.20100604_1111.html
<walterbender> thanks everyone
<SeanDaly> bye!
|<--:SeanDaly has left freenode (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401064631])