FESCo-2008-01-17

--- bpepple has changed the topic to: FESCo meeting -- Meeting rules at http://www.fedoraproject.org/wiki/Development/Schedule/MeetingGuidelines -- Init process
bpeppleFESCo meeting ping -- bpepple, caillon, c4chris, dgilmore, dwmw2, f13, jeremy, jwb, notting, spot, nirik, tibbs, warren
Hi everybody; who's around?
* nirik is here.
dgilmore is here
notting is here
tibbs|h here
warren here
* bpepple waits another minute to see who else shows up.
bpeppleok, let's go ahead and get started.
--- bpepple has changed the topic to: FESCO-Meeting -- sponsor nominations -- Manuel Wolfshant - https://www.redhat.com/archives/fedora-devel-list/2008-January/msg00689.html
tibbs|h+1
bpepple+1 here also.
nirik+1
warrenthe link is bad
but =1
+1
niriklink works fine here.
spotsure, +1
tibbs|hThe bugzilla link is bad.
nirikah, that link.
tibbs|hBut I've seen enough of wolfy's reviews to see that he knows what's up.
nottingreasonable enough. +1
dgilmore+1
buggbot<http://tinyurl.com/2zmf4m> (at bugzilla.redhat.com)
nirikyikes.
sorry about that... but thats the correct link.
bpeppleok, that's seven '+1', and no votes against him.  so we've approved Manuel's request.
I'll go ahead and upgrade him after the meeting.
moving on.....
--- bpepple has changed the topic to: FESCo-Meeting -- New Features - http://fedoraproject.org/wiki/Features/Dashboard - poelcat
bpeppleis poelcat about?
poelcatbpepple: ready to ROCK
bpepplepoelcat: you want to lead?
* jwb is here now
--- poelcat has changed the topic to: vote http://fedoraproject.org/wiki/Releases/FeatureDictionary
bpepple+1
tibbs|hDidn't we vote on the same thing pre-F8?
+1 regardless
notting+1 again.
nirik+1 again here too.
spot+1
nirikis the firefox patch going to get in? hopefully so.
tibbs|hSays "upstream state now resolved", so I'd say yes.
Only took, what, two years?
bpeppledgilmore, jwb, warren?
dgilmore+1
nottingnow all we need is for ff to use the system hunspell ;)
jwb+1
tibbs|hnotting: That's #3 in the "Optional" section.
No patch, unfortunately.
poelcatlooks like the +1s have it.. move on?
bpepplepoelcat: yup.
--- poelcat has changed the topic to: vote: http://fedoraproject.org/wiki/Features/XenPvops
nirik+1 to the ideas... 10% done and not much status makes me wonder if it's going to make it for f9... but I suppose there is still time.
spot+1, i just hope they can pull it off
warrenpage isn't loading for me
bpepple+1
tibbs|hI recall at fudcon that it wasn't decided that F9 will have 2.6.25.
* nirik wonders what that will mean for RHEL6...
notting+1
tibbs|hThis feature seems to be predicated on that.
warrenwithout being able to read the page, +1 to the name
tibbs|h, that isn't our job to worry about.
bpepplewarren: agreed.
tibbs|hIn some sense, you're correct,
f13I'm here.
tibbs|hWhat I don't want to see is "FESCo accepted this feature, so we must have kernel 2.6.25 in F9".
warrenSummary
Replacing the current forward-ported XenSource code on kernel-xen by a paravirt_ops based implementation, including dom0 support.
poelcattibbs|h: wouldn't that be a "dependency" ?
dgilmoreim saying +1
spottibbs|h: i trust davej and his team to make a sensible decision about the right kernel for F9
warrenDoes this mean the standard fedora kernel can be dom0?
spottibbs|h: i don't see this as forcing 2.6.25
jimadon't we backport stuff like this all the time?
tibbs|hWe don't see it that way; my concern is that somebody else might see it that way.
But in any case, getting rid of special xen crap and the forward porting pain is good for all.
So +1
f13+1 from me.
poelcatlooks like the +1s have it.. move on?
jwb0
warrenstandard fedora kernel can be dom0, but only if it is booted that way?
jwbwarren, ask the xen guys
spotwarren: thats outside of the scope of this feature
warrenjust trying to understand the feature
roccofrank|wwarren: http://lists.xensource.com/archives/html/xen-devel/2007-12/msg00454.html
dgilmorewarren:
Make 'new-kernel-pkg' aware of HYPERVISOR setting in /etc/sysconfig/xen
looks configureable
warrenok well, I already voted +1
dgilmorepoelcat: looks passed .  whats next
--- poelcat has changed the topic to: vote: http://fedoraproject.org/wiki/Features/GoodHaskellSupport
jwboh, another "more better" feature...
tibbs|hThis is missing "haskell packaging guidelines" in the Dependencies section.
I know he intends to submit some to FPC on Thursday.
nottinga haskell implementation of perl? whyyyy?
tibbs|hIt would be nice to say "we have a bunch of nice haskell stuff in the distro"
warrenOK well, I wouldn't see any harm in letting them try.
jwbfor the 4 haskell users?
* jwb shuts up
tibbs|hThere are really quite a few haskell users.
dgilmoreill say +1 it really cant hurt
spotehh, i'd feel better about voting +1 with guidelines done, but i'll give it the benefit of the doubt.
+1
jwb+1
tibbs|hBut I do recall that our "good ocaml support" wasn't really touted as a feature.
nirik+1
notting+1-ish
bpepple+1, also thinks this can't really hurt.
warren+1 this is just PR
tibbs|h+1
dgilmorespot: i trust that you will beat the guidelines into good shape
tibbs|hOh, we beat the guidelines nightly.
poelcatlooks like the +1s have it.. move on?
bpepplepoelcat: yeah.
--- poelcat has changed the topic to: vote: http://fedoraproject.org/wiki/Features/CommonLispController
bpepple+1
nirikso this is standardizing the already shipped lisp implementations?
notting+1, seems harmless
jwbisn't this a packaging guideline?
spotinstall a script /usr/lib/common-lisp/bin/<impl>.sh
i don't like that.
thats a terrible place to put scripts
tibbs|hThis is really sort of a "lisp packaging guidelines" thing.
jwbthis isn't a feature
tibbs|hShouldn't it go through FPC?
jwbtibbs|h, yes
spotand yes, this should be a FPC guideline
nirikyeah, should go thru the package comittee
tibbs|hNot that I really want to deal with it....
warren+1 to the idea that it should be standardized, but FPC decides the standard
tibbs|hSpot, ack on the "terrible place to put scripts" thing.
warrenterrible, but is it wrong?
spoti would say yes.
dgilmorei say -1 as a feature.  push it to FPC and get it as a guideline
jwbwhat dgilmore said
spotsame here. -1 as a feature, drive it through the FPC
warrenOK
f13-1
however
this could come back once the FPC approves the guidelines and all the packages fall in to conformity, this could be a feature for better Lisp whatever.
warrenI might agree that it is terrible, but it isn't wrong.
f13a lineitem on release features.
jwbthere is such thing as too much PR...
spotf13: i could see that, as an item on a list of other Lisp improvements done for F9
f13but our dozens of Lisp fans might want to feel special.
* poelcat realizes we don't have a procedure for features that are voted down... how should we communicate this to the feature proposer?
poelcatpublic/private mail?
jwbspot, can you get ahold of anthony green and tell him to go through FPC?
poelcat, would have to be public to some degree. meeting minutes will reflect we nak'd it
poelcatjwb: fair enough
warrenDid we agree on the earlier idea that *somebody* would have to draw the line between features and minor features?
poelcatwarren: we said there would be a line, we didn't say who
i'd assume that is FESCo
poelcathttp://fedoraproject.org/wiki/Features/Policy#head-2f8da702f8ab153647862bc942d9b0e28a080206
--- poelcat has changed the topic to: vote: http://fedoraproject.org/wiki/Features/IMDesktopIntegration
tibbs|hThis is a little sketchy on details for something that's 30% complete.
notting+1. watching discussions of IM UIs should be entertaining
dgilmorethis is very vague
spotthis is a good idea, but painfully vague.
jwbi'd say defer the vote
spot"build an infrastructure to bring up the input method on demand" "make things happen, magically"
nirikyeah, more details would be good. Also, anything that has the string 'gnome' should also mention what that means for KDE and other DE's
jwbnirik, it does sort of
"apply to the QT applications" "apply to the X applications"
dgilmorewe would need to carry the qt3 patch i think
qt4 could get upstream
jwb:(
tibbs|hYes, qt3 is closed for that kind of thing.
JuanCarlosLinHi all!
tibbs|hI'd be surprised if qt4 doesn't already have something of this nature.
dgilmoreit quite possibly does
tibbs|hBut, still, +0 from me without at least some additional detail.
bpepplepoelcat: could we have the owner of this feature provide a bit more information to us before we vote on it?
dgilmorewe would need to look at wx-widgets and all the other gui tool kits more so than desktops
+0 for now
poelcatbpepple: okay
dgilmoreid like to see something more concrete
bpepplepoelcat: great, thanks.
--- poelcat has changed the topic to: vote: http://fedoraproject.org/wiki/Features/freeIPA
dgilmore+1
spotdoes this involve any customization to make sso to the freeIPA server easier for Fedora clients?
warrenwait!!
tibbs|hI'm doing python-tgexpandingformwidget, so that's another dependency down.
spotas is, it just looks like "put freeipa packages in Fedora."
notting+1
warrenoh, nevermind.  I thought this was the crack ideas to integrate the *other* IM to the desktop
+1
dgilmoreall of fedora-ds has been approved now
jds2001    
f13warren: telepathy?
nirik+1
bpepple+1
* nirik wonders if he is the only one who sees IPA and thinks of a nice hoppy beverage. :)
tibbs|hI think that's sort of the point.
f13+1
bpepplenirik: nah, I was thinking the same thing.
dgilmorenirik: you might need one after looking at it
warrentibbs|h, knowing details of this particular integration, it only makes it better without changing parts that it cannot integrate.
jds2001nirik: no :)
poelcatwe need one more "+1" to move on
tibbs|hWhat I really want is some documentation for migrating a home-rolled configuration to this.
nirikI think they should get packages in asap so we can get testing done before f9.
tibbs|hBut that's way out of scope here.
+1.
spot+1
warrenI must highly suggest approving this feature, it really is harmless if they fail.
dgilmorenirik: i can yell at them to do so
--- poelcat has changed the topic to: http://fedoraproject.org/wiki/Features/AddingCrossCompilers
spotthis is missing packaging guidelines, and I wouldn't even think of approving it without them.
f13same here
tibbs|hHaven't we been there before?
spotalso, the targeted cross toolchains are vague, i'd rather see specific combinations targeted
f13and I think we've been down the packaging guidelines many times before, with stalemates.
dgilmorethis will need  some big disclaimers that the cross comilers cannot be used to create a secondary arch
tibbs|hStalemates as in Ralf says it's bad; Ralf won't say what's good.
spottibbs|h: i bet if someone wrote guidelines based on ralf's existing packages, he'd think they were good.
tibbs|hBut, yeah, +0 without some guidelines.
nirikyeah, needs guidelines. I would love to see things like tools for rockbox devel, or openwrt devel available tho. ;(
warrenopenwrt would be cool
f13or the n800 devel
tibbs|hAnd n810 devel.
warrenHow to approach this?
f13somebody needs to create packaging guidelines that don't suck
tibbs|hPeople who want it should talk and work something out.
warrenAsk them to suggest a packaging standard and review that standard?
dgilmorewarren: better than openwrt would be a fedora for wrt  but openwrt would be cool
nottingfedowrt?
dgilmorenotting: yup
tibbs|hSo that's it for features?
poelcatAcceptance of this feature is gating on someone else creating packaging guidelines? Is that fair to the feature proposer?
warrendgilmore, but cross compilers are not for secondary arches?
poelcator is there a better way to put it?
jwbwell
dgilmorewarren: they are not permitted
f13this feature has no chance of acceptance without existing packaging guidelines.
warrenI would suggest, people who care about it must 1) write a proposed standard 2) implement that proposed standard 3) ask FPC to review the standard and implementation
tibbs|hWe don't care who provides those guidelines.
dgilmorefor cross compilers i say we ask the leaders of it to propose packaging guidelines when they are through we can re evaluate
tibbs|hThe feature proposer could do it, or he could get someone else to do it, or he could coordinate an effort.
poelcatfair enough
Note: I'm also maintaining a list of proposed features with incomplete feature pages in the middel/bottom of the Dashboard page. http://fedoraproject.org/wiki/Features/Dashboard
dgilmoretibbs|h: sure  but he should drive it forward
poelcatI don't think it is a fair use of FESCo time to bring these up for a vote if all the requested information on the template is not present or has not been marked N/A or TBD.
bpepplepoelcat: seems reasonable.  anything else on Features?
poelcatthat's all for today
warrenwe already spent too much time discussing possibly minor features like lisp
bpepplepoelcat: great, thanks.
tibbs|hWe need to decide on gcc43 soon, though.
jwbwhy isn't that in the buildroots yet
nottingand xulrunner seems to be happening even without our decision :)
tibbs|hIt sort of is, just not the default tag.
f13jwb: because jakub is on vacation
jwb: and putting it in the buildroots now is a terrible idea.
warrenthey are trying to discover + fix gcc43 issues even before it goes in?
jwbwhy
tibbs|hwarren: Yes, they are.
f13His plan is to have it in the buildroots once he's back and can handle questions and debugging build issues.
warrenjwb, because jakub can very rapidly FIX issues created by the new compiler
poelcatf13: when is he back?
f13plus we didn't want it to be in the way of getting buildfixes in for the alpha
tibbs|hJakub did a mass rebuild, and it's easy to get test builds with it by scratch building against a different release tag.
jwbwarren, fix the compiler or fix other packages?
f13poelcat: should be later this week or early next week.
jwb: both
warrenjwb, both
jwbi find the latter to be horseshit
_maintainers_ should fix their packages
warrenjwb, well, he's good at both.
jwboh, i'm not doubting that
tibbs|hSome of the needed fixes are nontrivial and nonobvious.
jwbso what happens if jakub dies?
tibbs|hIt is an act of extreme courtesy to do things the way he's doing it.
warrenand he of course can prioritize his own time
jwbfedora just withers?
don't get me wrong, jakub rocks
but progress should not hinge on one guy
warrenjwb, compiler martyr?
tibbs|hHow is it depending on just one person?
f13jwb: not at all, but with jakub around things can be fixed much faster
jwb: because you and I know full well that /many/ of our maintainers are /not/ programmers.
as much as that sucks, that's the way it is.
jwbyeah, i'm on a tangent
ignore me
nottingthat being said, i think 'waiting until after alpha' is a perfectly fine reason
niriknext topic?
bpepplenotting: +1
--- bpepple has changed the topic to: FESCo meeting -- Free discussion around Fedora
bpeppleanything people need to discuss before wrapping up for today?
jds2001k - cna we get some traction on my bug workflow proposal?
* nirik was just going to mention that.
bpepplejds2001: I was going to add that to the schedule for next week.
jds2001http://fedoraproject.org/wiki/JonStanley/BugWorkflow
ok, we can table til then.
jwbnotting, not to be an ass but i think it would have been better to get it in before alpha
oh well. not everything works out nicely
bpepplejds2001: yeah, I figured we should give it at least a week to give folks to weigh in before bringing it here for a vote.
tibbs|hSo we are making slow progress on merge reviews.
warrenHow is cooperation from RH engineers?
tibbs|hSpotty.
* f13 looks at spot
tibbs|hSome folks are right on it and those get closed out quickly.
From others you get no response.
* nirik was lucky and got good responses back from the last 2 he did.
tibbs|hI have considered just making trivial fixes like fixing non-utf8 or improper directory ownership.
f13tibbs|h: I wouldn't stop you
dgilmorenirik: did you re-review that package
f13"that" package?
nirikdgilmore: which one?
dgilmorenirik: the odd one
f13gnome-settings-daemon?
if so, I did the re-review
dgilmoreyeah that one
f13it was fine
nirikf13 reviewed it
dgilmoref13: :)  thanks ifigured it was fine  but found the review odd
tibbs|hSo how many packages are we adding per day?
tibbs|hIt seems like it's more than ten, which is crazy growth.
f13oof, too many
dgilmorewhere is the list of merge reviews outstanding?
nirikI hate to call people on that kind of thing, but we do need more than one person in a review.
tibbs|hdgilmore: http://fedoraproject.org/PackageReviewStatus/NEW.html
has all of the reviews; just search for "Merge".
jds2001btw, as of last night we had 790 package reviews in new state
tibbs|hhttps://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=426387
buggbotBug 426387: low, low, ---, Jason Tibbitts, NEW , Merge reviews to be completed for F9
jds2001oh, you already got something :)
tibbs|hThat has the hitlist for F9.
dgilmoref13: https://bugzilla.redhat.com/show_bug.cgi?id=226493
buggbotBug 226493: medium, medium, ---, Nobody's working on this, feel free to take it, NEW , Merge Review: tix
JuanCarlosLinOk, can we  start the meeting for North & South America Ambassador?
* nirik wonders if we have any stats graphed from bugzilla... would be nice to see trends. reviews closed, reviews added, bugs added, etc.
jwbJuanCarlosLin, we still have 1 minute :)
bpeppletibbs|h: that looks better than I thought it would.
jwboops no more
* jds2001 working on that
warrenThat's a great way to limit meeting length =)
tibbs|hbpepple: Yeah, we're not doing too bad.
dgilmorebpepple: lets wrap up
bpeppleJuanCarlosLin: yeah, just give me a minute to wrap up.
warren(It isn't like we can't possibly have multiple meeting channels.)
tibbs|hAnd there are still several easy tickets on the hitlist.
* bpepple will end the meeting in 60
warren"Sorry, we can't build more meeting rooms here."
JuanCarlosLinOk...Hi all, please state your names
nirikjds2001: let me know if you need anything... great that you are taking on the bugzilla beast. ;)
* bpepple will end the meeting in 30
bpepple will end the meeting in 15
rMenezessorry, I'm late
bpepple-- MARK -- Meeting End

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