FESCo-2008-02-28

--- bpepple has changed the topic to: FESCo meeting -- Meeting rules at http://www.fedoraproject.org/wiki/Development/Schedule/MeetingGuidelines -- Init process
* warren calculating the current orphan list
bpeppleFESCo meeting ping -- bpepple, caillon, c4chris, dgilmore, dwmw2, f13, jeremy, jwb, notting, spot, nirik, tibbs, warren
Hi everybody; who's around?
* tibbs here
spot is here
c4chris is here
jwb is here
warren here
* dgilmore is somewhat here
bpeppleok, there's not much on the schedule for today, so we can probably get started.
--- bpepple has changed the topic to: FESCo-Meeting -- New Features - http://fedoraproject.org/wiki/Features/Dashboard - poelcat
bpepplepoelcat about?
--- bpepple has changed the topic to: Feature Request - http://fedoraproject.org/wiki/Features/HandwritingRecognizer
* f13
bpepple+1 to handwriting feature seems pretty cool.
warren+1
tibbsI saw that under review.  I guess it's just a new package; if someone wants to hype that as a feature then I don't see any reason not to.
* nirik is here now.
nottingmight want to clarify that it's just cjk
also, not sure it will be in any of the main spins, but that's not a dealbreaker for features
f13+1 from me, although the feature is generic sounding
tibbsYeah, I'd propose changing the feature to CJKHandwritingRecognition
c4chris+1
nirikyeah, +1 here with clarification...
jwbmeh
bpeppletibbs: yeah, that makes sense.
c4christibbs: +1
tibbsIt also doesn't look like it does hangul.
warrentibbs, matter of adding to the dictionary
tibbswarren: I don't think so.
It doesn't do kana either.
warren"Currently recognizes Chinese characters as used in Traditional and Simplified Chinese hanzi, Japanese kanji, and Korean hanja."
tibbsBut hangul is much more frequently used in modern Korean than hanja are.
warrenfrom the description it sounds like it is being developed from a chinese-centric viewpoint
* poelcat here
warrenwhich is OK, it is possible to add kana and hangul later
don't get too hung up on this, I think.
nottinghm. it's nice that he has a version roadmap that goes to 1.1, but he's only at 0.1 now
bpepplewarren: agreed.
tibbswarren: I think you are assuming something about the architecture of their recognition engine that is not stated in the feature request.
I just don't want to go hyping "We have Korean handwriting recognition" when we might not actually have that.
warrenlet's just see how far they go before F9?  If it looks pretty useless then we will refrain from mentioning it in the release notes.
f13worksforme
warrenWe could approve it and say "tone down the korean if you don't really support korean"
or call it "ancient korean" =)
nottingwe also already ship tomoe, which purports to do the same thing (differently)
warrenI say, approve with comments?
tibbsI guess the point is to get someone who communicates in those languages to tell us how useful the software actually is.
* jeremy is here now
nottingtbh, it seems more at the 'Fedora Daily Package' level than 'Feature' level. but maybe i'm quibbling
tibbsI'll edit the feature request to ask for clarification on the kana/hangul question.
warrentibbs, that's a fair question, can't claim to support Japanese and Korean without it.
next?
c4chrisI only count 5 +1's, so maybe we wait for that to be clarified ?
nottingi'm -1 for now
warrennotting, I guess if it were totally awesome and not half-baked it might be worth mention?
c4chris, isn't this the final fesco meeting before we stop accepting features?
poelcatyes
c4chrisoh, didn't realize...
warrenOr maybe, since reviewed features are largely for promotional purposes, perhaps we should review things like this later to see how good it really is.
There might be things worth mentioning in the release notes that are awesome that weren't approved.
bpeppleI would lean towards accepting it for now, and seeing what progress is made.  We can always drop it later.
tibbsI have no problem provisionally accepting this as long as it accurately represents what it does.
poelcattheoretically it has to be "done" by 2008-03-04 too :)
nottingwhen i think that fedora would tout a 'Handwriting Recognition' feature, i would think it would be something supporting a majority of shipped locales (including latin), be integrated into X in some way for tablets to use, etc.
caillon+1 to "theoretically" :-)
caillonbpepple, agreed.  so +1 on the feature from me.
bpeppleok, I see six '+1', and one '-1'.
anyone else want to weigh in?
dgilmore+1
jwb0
c4chriswith tibbs provisional ok, that'd be 7 (and 8 with dgilmore ...)
tibbsYeah, you can count me as +1.
bpeppleok, so we've approved this feature for now.
poelcat: is there anything else in regards to features to discuss?
poelcatNo more features to review ... next week if 'feature freeze' sticks at 2008-03-04
it will be time for the "Clash Round" http://fedoraproject.org/wiki/Releases/9/FeatureList
'should they stay or should they go'
bpepplefun.
warrenwhen?
tibbsI guess the handwriting thing better shape up quickly, then.
caillonhm.  i just noticed a few features that aren't there that i thought got approved.
like i dunno.... firefox 3.
or xulrunner
poelcatcaillon: i sent mail on friday asking everyone to review the list
jwbi don't recall ff3 being a feature
caillonfair enough.  maybe it's not :)
nottingit should be
jwbi meant being proposed as one
it should be
dgilmorei think it should be one
caillonpretty sure walters wrote one up a while ago
* caillon finds out
c4chrisyaeh, and I'm pretty sure we discussed xulrunner at some point
jwbi could be mistaken.  xulrunner i remember though
poelcatnope; that was for F8
jwbwhich, ff3 or xulrunner?
poelcatthe feature page is incomplete for xulrunner
ff3 i have not seen a page for
c4chriswasn't xulrunner too late for F8 and pushed to F9 ?
caillonhow is xulrunner incomplete?  looks good to me!
* warren has an updated orphan list.
poelcatcaillon when the hamsters turning the wiki wheel wake up I'll be able to tell
caillon: no contingecny plan and no release notes
caillonthat's correct.  there is no contingency plan.  it's in and not coming out.
warrenIt wont come out of the closet.
tibbsDid I lag out or did it just get really quiet?
nirikok, so where are we here? caillon: can you add release notes to xulrunner and we can vote on it?
notting<uncomfortable silence>
* nirik nods at tibbs. I was thinking the same thing.
nottingcaillon: do you want to volunteer to write a page for ff3?
c4christhe crickets just ate the hamsters I'm afraid :)
caillonjust ran over to walters' cube
he said he had written a ff3 feature page but he deleted it
since someone larted him over it for some reason
is there a way to find deleted wiki pages?
bpepplethe status of some the app built against xulrunner need to updated also. (I'm thinking of epiphany, and possibly others).
caillonnirik, i would probably defer to upstream relnotes which aren't there yet.
f13have we moved on topics?
bpepplef13: yeah.
--- bpepple has changed the topic to: Features - http://fedoraproject.org/wiki/Releases/FeatureXULRunner
* warren confused.
warrenwhat are we doing?
nirikwell, clearly we want xulrunner and firefox3 features... perhaps caillon could work with poelcat to get those pages in shape and we can vote on them via email in the next day or so?
dgilmorenirik: sounds good
caillonsure
bpeppleworks for me.
c4chrisworksforme
f13yeah, worksforme
bpepplepoelcat does that work for you?
poelcatyes, my only request is that
if a section is n/a or whatever... please write it in
we can't read minds :)
so an empty section is ambiguous to many other readers
that is all
caillonfair enough
bpepplepoelcat: anything else? or should we move on?
poelcatall done
bpeppleok.
--- bpepple has changed the topic to: FESCo meeting -- Free discussion around Fedora
warrenorhans
orphans
* warren needs a pastebot
warrenhttp://paste.ausil.us/293
--- bpepple has changed the topic to: Orphan packages - warren
warrenshorter list of orphans
some have co-owners but not owners
not sure what to do with those
nottingtthe co-owners have been promoted?
jwbsome don't want to be
i question the value of being a co-owner then
caillonis there really a "promotion"?
aren't co-owners on equal footing?
tibbsI don't really understand what distinguishes the primary maintainer from the others.
bpepplejwb: agreed, seems that don't have an interest in the package.
tibbsI guess it's who the bugs get assigned to by default.
nirikit's a quirk of the pkgdb I think...
I think we should set comaintainers to owner, if they don't want that they can re-orphan...
caillon+1
nottingabadger1999: can pkgdb automatically promote co-maintainers to primary if primary disowns a package?
nirikbut then why they were co-maintainer in the first place is another question.
warrennotting, +1
abadger1999notting: Not yet.
warrenOK, for now we manually promote a co-owner.
abadger1999notting: How do we decide which comaintainer should be promoted?
warrenIn the future pkgdb will do it automatically.
abadger1999, shortest straw
nottingabadger1999: in time-honored yum/rpm tradition, shortest account name wins
* warren shakes head.
nirikshortest package^H^H^H^H^H^H^Husername ?
* nirik nods at notting. Ha. jinx.
caillonlowest user integer id?
* abadger1999 changes account name to supercalifragiliciousexpialadocious
f13if there are multiple co-maintianers, how do you promote all of them, or can  bugzilla have a list of initial owners?
warrenf13, bugzilla can have only one "owner", the others are CC'ed I think.
abadger1999f13: Only one can be promoted.
caillonf13, no, this is why other bugzillas use -maint aliases
i think we're the only ones stuck on still using real people as owners in bugzilla
warrenabadger1999, btw, moving to pkgdb for bugzilla auto-CC sync had a few drawbacks, like the inability to auto-CC upstream people who are not fedora developers.
abadger1999notting: Should we consider anyone with (commit access || aproveacls) eligible for promotion?
* f13 still thinks bugs should be owned by <package>@fedoraproject.org and where that email goes depends on pkdgb de jour
caillonabadger1999, in seriousness, do you track the length of time they have been a co-owner?
abadger1999warren: I now.  We need FAS2 to make it easy for upstream people to have a Fedora account.
caillonf13, agreed
nirikabadger1999: sounds good to me, and that would make further discussion here moot, as there is only 1 person on those packages
abadger1999caillon: No but we could.
warrenOK so manual promotion (and later auto-promotion) is decided.  Next we need to decide on when to remove orphans.
abadger1999When they got an acl is logged in the db.
c4chrisf13: good idea
caillonabadger1999, i think that would be useful even if we don't use it to promote people
EvilBobI was going to ask about this for the packages my SIG has been submitting
abadger1999Although there will be a big tie for a while since most people were imported from owners.list when the pkgdb started.
warrenabadger1999, any ballpark ETA on fas2?
abadger1999warren: We are aiming for early March.
warrennice
caillonabadger1999, that's fine
abadger1999warren: mmcgrath and ricky are spearheading it and I am doing most of the database work.
nottingabadger1999: maybe && instead of ||
abadger1999warren: It's pretty close
* warren is super glad it is based on a RDBMS
nirikwarren: I would say remove orphans asap... promote those 3 with co-owners to owners, and then wait a day and block/dead.package the rest?
abadger1999notting: Okay.  I'll make a ticket and add all this information.
warrenHow about this: Promote those 3 owners, WARN one more time today, with end of Friday as the kill time.
jeremywarren: can you make sure that none of the ones on your list are the ones that jesus was going to take?
(if he hasn't gotten sponsored yet)
* warren looks
nirikjeremy: one is.
warrenjeremy, is anyone taking care of Jesus?
jeremynirik: I thought so, but my memory is fuzzy enough that I wasn't going to stake the discussion on it :)
nirikbut it already has devrim saying he would take it.
not sure about the rest.
jeremywarren: he's getting one or more of the packages re-reviewed to get a sponsor I believe
nirikjeremy: unfortunately, they are all java packages... and with no java guidelines, not many people are going to touch it.
warrenhm
gnome-blog is still an orphan
jeremygnome-blog is orphaned _upstream_
tibbsjava guidelines are turning into a big problem.
* nirik guesses he can sponsor jesus and help him... the java packages are already in, so maintainer is better than orphaning them.
jeremyhence, this should come as little surprise
warrenjeremy, so what should we do?
tibbsOr the lack thereof, I guess.
* dwmw2_BOS arrives, somewhat late
jeremywarren: orphaning it is fine.  there are lots of other blog posters that we could change to be default
niriktibbs: totally agreed. I wonder if jesus knows java well and could help with guidelines... I can ask.
warrenjeremy, so both block and remove from comps?
jeremynirik: that would be great
warren: unless someone steps up, yep
f13I already told jesus that we wouldn't remove anything out from under him as he got into the system.
so no need to rush jesus's entry
jeremyf13: right, I just said we should make sure not to block any of those packages
caillonwwjd?
warrenthe remaining five packages are not Jesus's
aget
cgi-util
eds-feed
gnome-blog
windowlab
nirikgnome-blog was discussed on the list as being dead upstream.
jeremywarren: to the orphanarium they go!
nirikaget was orphaned for being dead upstream I think.
bpepplejeremy: +1
warrenjeremy, time to empty out the orphanarium.
jeremylet's not waste any more time on this :)
warrenOK, posting last warning.
* nirik nods.
warrenFedora Orbital Laser is charging the capacitor.
bpeppleok, anything else we need to discuss today?
--- bpepple has changed the topic to: FESCo meeting -- Free discussion around Fedora
skvidalwarren: I thought we used the orphans to fuel the oribital laser
f13if any of ya'll want to help with the broken dep list before beta, that would be great.
nottingskvidal: poor annie
* warren wonders if any of these orphans will create broken deps
c4chriswhere's the list ?
skvidalnotting: "tomorrow, tomorrow, bzzzzzzzzzzzzzzzzrt"
nottingwarren: repoquery will tell you
f13c4chris: the rawhide report.
c4chrisI shall read it more carefully next time...
bpeppleok, I think we can go ahead and wrap-up today's meeting.
* bpepple will end the meeting in 60
bpepple will end the meeting in 30
bpepple will end the meeting in 15
bpepple-- MARK -- Meeting End
Thanks, everyone!

Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!