Web   ·   Wiki   ·   Activities   ·   Blog   ·   Lists   ·   Chat   ·   Meeting   ·   Bugs   ·   Git   ·   Translate   ·   Archive   ·   People   ·   Donate

#sugar-meeting, 2009-01-15

Index | Today     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
10:06 gregdek w00t!
10:06 mchua \o/
10:06 gregdek LOL
10:06 OK.
10:07 looks at the TODO and presumes that it composes our agenda, more or less?
10:07 walterbender hey mchua, wanna document that in the wiki?
10:07 mchua walterbender: already heading towards there. ;)
10:07 gregdek Actually, I guess the first question is:
10:07 when is freeze?
10:08 tomeu tomorrow
10:08 unmadindu from tomorrow
10:08 tomeu http://sugarlabs.org/go/Develo[…]m/Release/Roadmap
10:08 gregdek And not only when, but, given the recent circumstances, why?  Is our calendar still right?
10:08 tomeu why not? ;)
10:08 walterbender I think we should stick to the Sugar schedule...
10:08 tomeu we want to give enough time so we can ship a stable 0.84 in the next distro releases
10:08 from my POV, that's critical for us
10:09 walterbender we need to get into sync with more than just the OLPC release cycle
10:09 so being consistent is important
10:09 tomeu because OLPC may not update their builds so often and because OLPC don't have QA again
10:09 gregdek True enough.
10:09 I'm all for a consistent schedule.
10:09 But I also want to make sure that we agree that we have the right features this time around.
10:09 tomeu also, I think we have unstabilized sugar enough for this cycle ;)
10:09 gregdek OK.  :)
10:09 tomeu gregdek: would be nice to hear what people think about it
10:10 which features are missed right now
10:10 gregdek If the consensus is clearly "DON'T TOUCH THE SCHEDULE!" then I'm fine with that.
10:10 tomeu well, I see good reasons for staying with it, but wonder if there's a very needed feature we have missed and that might not be too destabilizing
10:10 dfarning SL, strength over the last several months has been, 'plan the work, work the plan'
10:10 walterbender I need to request that the Portfoli be added to Fructose
10:11 gregdek My only point is that, now that our previous hard dependency on OLPC is broken, we have an opportunity to move the schedule if we want to take advantage of it.
10:11 walterbender and ask Sayamindu to add ti to Pootle
10:11 unmadindu walterbender: noted - will do
10:11 walterbender what would work best for Fedora?
10:11 tomeu gregdek: true, though at least me have been thinking more about F11 and jaunty
10:11 gregdek OK.
10:11 tomeu rather than OLPC
10:11 because they were changing schedules all time
10:12 so no point in hurting ourselves for nothing
10:12 gregdek Well, as it happens, the schedule as it exists now is perfectly aligned with Fedora's 3/3 freeze.
10:12 walterbender then let's go for that.
10:12 gregdek So I guess I'll just say "well done with the schedule" and move on.  :)
10:12 walterbender any comments from the Ubuntu crowd?
10:13 tomeu now that we talk about distros: http://sugarlabs.org/go/Packaging
10:13 mchua peeks at https://wiki.ubuntu.com/JauntyReleaseSchedule - looks pretty good.
10:13 dfarning debian is in a pre-release stabilization period.  So debian and ubuntu are checking each others scheduals.
10:13 tomeu dfarning: do you have poke powers over non responsive debian/ubuntu maintainers?
10:14 dfarning tomeu, limited, but yes.  Jonas is here in the room with me:)
10:14 erikos is this the developers meeting?
10:15 mchua erikos: yes.
10:15 tomeu oh, got there
10:15 erikos mchua: hey! cool
10:15 tomeu dfarning: is about abiword, we need it built with --enable-libabiword, but haven't got reply from the maintainer yet
10:16 dfarning ok, i will find and poke them--hard
10:16 gregdek So.  Has protocol in the past been to simply work down the TODO list?  Shall we proceed that way?
10:16 bertf which TODO list?
10:16 gregdek http://sugarlabs.org/go/DevelopmentTeam/TODO
10:17 tomeu gregdek: that was started on the last meeting almost one month ago and we have had holidays since then
10:17 gregdek hrms.
10:17 tomeu also, now that we start in freeze, may be better to work on trac
10:17 gregdek Is there a separate agenda somewhere?
10:17 erikos gregdek: http://sugarlabs.org/go/Develo[…]uary_15_2009_-_15.00_.28UTC.29
10:17 gregdek LOL
10:18 So I see that item 2 is "update TODO list".  :)
10:18 But since item 1 is "prep for freeze"...
10:18 bertf when is a good time to discuss etoys packaging? should I care?
10:18 gregdek bertf: As in, distro packaging?
10:18 bertf gregdek: yes
10:19 gregdek bertf: I can't speak for others, but I'm going to push hard to make this happen in Fedora, and the Fedora/OLPC meeting is at 1pm Eastern US time today.  In about 2.5 hours from now.
10:19 I suspect that a lot of the work for Fedora packaging will help clear the way for others.
10:19 So, wanna come to that meeting?  1pm, #fedora-olpc?
10:20 tomeu bertf: perhaps dfarning can help with etoys packaging in debian/ubuntu?
10:20 bertf gregdek: sorry, no,. I'm on kids-watch today
10:20 gregdek Ah, ok.  Well, watch the mailing list for updates.
10:20 bertf gregdek: gavin was working on fedora packaging
10:20 gregdek Yes, but not moving too quickly, from what I last saw.  I need to sit down with him.
10:20 Figure out what the roadblocks are.
10:21 Anyway, that's for later.  :)
10:21 So... um...
10:21 dfarning tomeu, bertf, will poke again.  We resolved most of the issues a few months ago.
10:21 gregdek ...is everyone ready for feature freeze?  :)
10:21 bertf dfarning: great
10:21 silbe regarding packaging in Debian: how long are you going to support (i.e. react to bug reports) 0.82?
10:21 gregdek Is there a list of proposed features with statuses, etc. anywhere, or no?
10:22 cjb appears, hello
10:22 tomeu silbe: that's a good subject, can you add it to the agenda?
10:22 hi cjb
10:22 silbe lenny might ship with 0.82, so Debian users are going to be using that for quite a while if they're using the distro-provided packages
10:22 tomeu: how?
10:22 tomeu silbe: http://sugarlabs.org/go/Develo[…]uary_15_2009_-_15.00_.28UTC.29
10:22 "Topics"
10:22 gregdek: we used to, but as OLPC changed plans, we ditched it and moved to the TODO list
10:22 gregdek Hi cjb.  Good to "see" you.  :)
10:22 erikos gregdek: that is what we have in the todo list (of pending features at least)
10:23 gregdek All right, then.
10:23 So maybe the best way of assessing where we are for freeze is actually going down the TODO list, after all?
10:23 erikos the list needs to be updated though
10:23 tomeu but we have good info on the old feature pages
10:23 erikos gregdek: yup
10:24 gregdek Okey doke.
10:24 "Making PS a bit more controllable from the control panel".
10:24 morgs?  You around?
10:24 tomeu for the curious: these links lead to the old feature pages: http://sugarlabs.org/go/Develo[…]ase/Roadmap#Goals
10:24 erikos gregdek: we have a patch we are about to sort out for that
10:25 silbe tomeu: done
10:25 gregdek erikos: So it's "in" for FF?
10:26 erikos gregdek: yes - i just added the ticket number
10:27 gregdek OK, so I understand:
10:28 The goal of adding the ticket is so that we can track it through the release cycle?
10:28 does not know our trac workflow at all, apologies.
10:28 erikos gregdek: yes, our release notes does contain the closed ticket numbers
10:29 gregdek Got it.  Thanks.
10:29 OK, then.
10:29 tomeu from the changelogs
10:29 gregdek Moving on:
10:29 "Cp language section (display the language options in the original language: English, Espanol, Deutsch)".
10:29 unmadindu?
10:29 erikos gregdek: we have a patch - needs to be reviewed
10:30 unmadindu patch for review
10:30 gregdek looks at ticket.
10:30 erikos unmadindu: did you test performance wise on the XO?
10:30 unmadindu erikos: not very encouraging - I got a lag of a few seconds (3-4)
10:31 erikos unmadindu: for drawing the combobox?
10:31 gregdek What's our patch review process?  Just a +1 from any developer?
10:31 unmadindu erikos: no - for getting the translations
10:31 tomeu unmadindu: can we do those computations lazily? I thought TreeModel supported something like that
10:31 unmadindu tomeu: I:l check
10:31 erikos gregdek: yes that is about to be correct
10:31 tomeu gregdek: let me look up in the wiki
10:32 gregdek LOL
10:32 :)
10:32 tomeu gregdek: http://sugarlabs.org/go/Develo[…]ntTeam/CodeReview
10:33 gregdek Good stuff, folks.
10:34 All right, so from a process perspective, does a patch have to be "submitted" by Feature Freeze, or successfully reviewed by Feature Freeze... or do we care?
10:34 In other words, is this feature sufficiently "in"?
10:35 begs everyone's pardon for learning on the job.
10:35 erikos gregdek: the code should be in git
10:35 tomeu gregdek: they aren't yet in, but we hope they will be pushed before the freeze
10:35 oh, it is already? /me shuts up
10:36 erikos tomeu: no, it is not in already - i was replying to the process question
10:36 gregdek erikos: To be clear.  Do you mean this code is NOW in git, or this code SHOULD be in git to be "frozen", but is not yet?
10:36 OK, thanks.  :)
10:36 Then we need to get it into git, and since unmadindu wrote the patch, he can not own the process of getting it in.
10:36 So who's reviewing this patch and putting it in?
10:37 erikos gregdek: i am here for reviews today
10:37 gregdek Awesome.
10:38 erikos i will assist to get morgan's #142 and unmadindu's #51 in
10:38 gregdek OK.  Brilliant.
10:38 Moving on:
10:38 RTL support in Sugar.
10:38 unmadindu?
10:39 unmadindu bits and pieces still missing. there is a patch from Khaled, but marcopg is not happy with that
10:39 gregdek How does that affect freeze?  Is the code currently in place sufficient to do *something* useful?
10:40 erikos unmadindu: yeah, we did not really work on it in the last months, right?
10:40 unmadindu erikos: yes - it requires some invasive changes all over the board
10:40 gregdek (And is there any place where the functionality is described?)
10:40 unmadindu gregdek: it is mostly done - only bit missing is that some icons do not show up as mirrored
10:40 gregdek: it[ on the OLPC trac
10:40 erikos unmadindu: the '>' ones etc?
10:41 unmadindu erikos: yes
10:41 tomeu gregdek: basically, there are some UI elements that should follow the language writing order
10:41 gregdek Ah, ok.
10:41 "RTL" == "right to left".
10:41 :)
10:41 tomeu gregdek: we need to change that following a runtime variable that depends on the language active in tha tmoment
10:41 right ;)
10:42 gregdek So this looks like incremental improvement where we can get it.
10:42 Seems like the feature may be "done enough" such that fixes during the cleanup can responsibly be considered "bugfixing"?
10:42 unmadindu gregdek: yes, but the missing bits require significant invasive changes, if done the correct way
10:42 gregdek Ah.
10:43 unmadindu distros can use Khaled[ patch if they want
10:43 erikos yeah, i guess unlikely to get in today :/
10:43 tomeu gregdek: the right "fix" would be involving developers from those locales
10:43 unmadindu that is pretty safe
10:43 gregdek And the wrong "fix"?
10:43 unmadindu gregdek: we already have that
10:43 (not in git though)
10:43 gregdek That is Khaled's patch?
10:43 unmadindu yes
10:43 tomeu put some code in now that may not be enough but destabilize stuff
10:44 gregdek I'm new here, so I will rely on everyone else's thoughts...
10:44 tomeu unmadindu: I'm afraid khaled's patches won't apply too cleanly now
10:44 but we can give it a try
10:44 unmadindu tomeu: I can take care of that
10:44 tomeu ok, nice
10:44 unmadindu tomeu: but do we want that in upstream ?
10:44 erikos i think the first move would be, to move the patches to d.s.o
10:45 tomeu unmadindu: I'm not sure
10:45 gregdek Yes.
10:45 erikos the trak i mean
10:45 gregdek Even the dirty patches you don't like should be upstream.
10:45 tomeu unmadindu: I miss marco's opinion on this one
10:45 unmadindu tomeu: I think it may be best to leave it to the distros - eg: if RTL is a priority for OLPC, they patch it in their own package
10:45 tomeu unmadindu: if we had someone who was able to tell us how much important it will be for us in the next 6 months, we may be able to take a better decision
10:45 gregdek Does the RTL affect anything for the distros outside of Sugar itself?
10:46 erikos unmadindu: i thought RTL is important to sugar in general?
10:46 unmadindu tomeu, erikos: I think it is pretty important to OLPC
10:46 gregdek: http://dev.laptop.org/ticket/6562
10:47 tomeu hmm, if we switch to some other canvas for 0.86, we have to rewrite those parts anyway
10:47 unmadindu oh - are we considering that ?
10:47 tomeu unmadindu: we have been considering that since before using hippo-canvas ;)
10:47 unmadindu oh, lol :)
10:47 tomeu unmadindu: hippo is not being maintained, so...
10:47 gregdek My take, which is uninformed in this case but generally my philosophy: (a) imperfect code is better than no code; (b) open a ticket against 0.86 so you don't lose track of it; (c) even dirty code should be upstream.
10:48 erikos this looks like marco's opinion: http://dev.laptop.org/ticket/6562#comment:13
10:48 unmadindu would cautiously agree with gregdek ;)
10:49 gregdek marco advocates for the "right" way.  No time for that now, so the only open question is, do we fix it the "wrong" way in the short term, or not at all?
10:50 Which is a function of how painful it will be to fix once we've done it the "wrong" way.
10:50 tomeu but it's not only 6562, right?
10:50 gregdek A question that I, as a non-developer, cannot answer.  :)
10:50 unmadindu tomeu: that[ the only major RTL bug left, as per my last conversation with Khaled
10:50 tomeu so that patch is not so big
10:51 erikos gregdek: i would argue for not fixing this in a buggy way
10:51 unmadindu tomeu: no, it[ pretty simple
10:51 tomeu I thought it was adding reverse() calls to all horizontal hippo boxes
10:51 erikos gregdek: the situation has not changed to when we punted it for 0.82
10:52 gregdek What else is involved in the feature besides olpc6562?  Anyone?
10:52 is worried we're taking up too much time with this one... still a number of features to go through...
10:52 erikos gregdek: i think we can discuss it again after the meeting
10:52 gregdek tomeu: Gut feeling -- dirty fix, or no fix?
10:53 Don't think.  Answer.  :)
10:53 tomeu no gut feeling here
10:53 gregdek Well, right now, the only firm opinion is -1.
10:53 From erikos.
10:53 tomeu but the lazy guy in me says no fix
10:53 gregdek Which means, unless someone says PUT IT IN, it's out.
10:53 LOL
10:54 All right.  It's out.  Will punt to 0.86.
10:54 tomeu we really need a champion for these issues
10:54 erikos will update the wiki
10:54 gregdek Already did.
10:55 tomeu: We'll work on that.  :)
10:55 Next: NetworkManager.  erikos?  From the notes, looks like using the native gconf stuff is a fix that won't make it in?
10:56 erikos gregdek: yes - i would need to write that today :(
10:56 gregdek Ah well.  Rome wasn't built in a day.
10:56 tomeu erikos: do we really need it?
10:56 erikos tomeu: oh, it is not that important, but would be my first update
10:56 gregdek My take as a user: it's an annoyance, but not a show stopper.
10:56 tomeu ok, 0.86 sounds good
10:57 gregdek updates.
10:57 Next: logout option.  http://dev.laptop.org/ticket/8141
10:58 unmadindu?
10:58 silbe will sugar 0.84 work without NM as well? upstream says not to use it at all if it cannot cope with the given use case
10:58 oh, sorry, too late
10:58 unmadindu the solution proposed there relies on an older version of GDM
10:59 I don think it will work with the GDM for the next version of GNOME
10:59 erikos unmadindu: what solution do you mean exactly?
10:59 gregdek A long dispute with Eben, heh.
11:00 unmadindu erikos: eben wants something like fast-user-switch in GNOME
11:00 erikos: that is still not there for the new GDM, when I last checked
11:00 gregdek The use case: me switching from Sugar to GNOME easily on my F11 system.
11:01 unmadindu yes
11:01 erikos unmadindu: oh, ok - read that now
11:01 unmadindu: for now - we can just apply the patch we have in F10
11:01 gregdek The simple answer would be "give me a logout button," but eben doesn't like that.  :)
11:01 unmadindu erikos: yes - that would be the solution I propose
11:02 erikos gregdek: we did that in F10; i think that is absolutely fine
11:02 unmadindu fast-user-switch-applet communicates with GDM via a socket
11:02 gregdek Me too.
11:02 So eben is overruled.  :)
11:02 unmadindu that is not there in the newer GDM (yet)
11:02 gregdek Which means:
11:02 erikos gregdek: eben can give us a nice icon if he wants to :)
11:02 gregdek 1. patch for "logout" for F10 is ported to current release;
11:02 eben gregdek: What does it mean to logout if you're just running Sugar itself?
11:03 gregdek Speak of the devil, and the devil appears.    :)
11:03 silbe will it support regular logout, as well (for usage on a normal desktop)?
11:03 gregdek 2. user-switching is punted to 0.86.
11:03 Is that right?
11:03 tomeu eben: I think we should only show the logout option when sugar has been launched from gdm
11:03 unmadindu gregdek: I would highly recommend that
11:03 tomeu eben: I think we can know that
11:03 eben tomeu: Ah, in that case I'm fine with it.
11:03 tomeu silbe: yeah, that's the idea
11:03 gregdek tomeu: Could you, or someone, make that change today?
11:04 Probably a one-liner?
11:04 erikos tomeu: yeah, that sounds right
11:04 unmadindu tomeu: I can modify the patch if you want to do that
11:04 eben I'm not against the option...I'm against showing it if it is meaningless in a given context.
11:04 silbe tomeu: ok, thx
11:04 tomeu unmadindu, erikos: do you already know how to check if sugar was launched from gdm?
11:04 eben Just add it to the XO menu (which I'm still working on a patch for)
11:04 walterbender we need to detect the context somehow
11:04 unmadindu tomeu: check for GDMSESSION I think
11:05 tomeu unmadindu: can you take that? I would like to work on evince, as the gnome API freeze was already three days ago
11:05 erikos so yeah, i think the two remaining items are: find out if we have been launched from gdm, the icon we use
11:05 unmadindu tomeu: sure - will send in patch by midnight today (IST)
11:05 gregdek So.  The goal for THIS RELEASE is to take the "show logout" patch for F10, tweak it so it only displays if the session was launched from GDM, and commit.  Today.  If the answer is "yes," then we need to know who owns it.
11:05 tomeu unmadindu: awesome, I can review, etc
11:05 gregdek unmadindu: this is your item?
11:05 unmadindu can someone give me an icon ?
11:05 gregdek: I need a icon
11:05 looks at eben
11:05 erikos eben: ^^^
11:06 eben I can make an icon...are there any standard suggestions?
11:06 gregdek eben seems hesitant.  :)
11:06 tomeu eben: dunno :/
11:07 eben: an arrow exiting from a square?
11:07 silbe what about kdm et. al?
11:07 those wouldn't set GDMSESSION
11:08 unmadindu silbe: I:l see if I can handle KDM as well, but no promises ;-)
11:08 eben Hmm....we could use the stroke/fill metaphor here, maybe....a white outline of an XO?
11:08 gregdek eben: The international "exit" sign?  The little dude running out of a burning building?  :)
11:08 Anyway, I think we've got this one nailed.  Ready to move on?
11:08 mchua eben: http://www.manucornet.net/GNOME/logout.png
11:08 gregdek "screen issues like those in the control panel".  tomeu?
11:09 tomeu gregdek: plan to work on that today
11:09 erikos gregdek: i think this can be seen as a bugfix as well ;p
11:09 tomeu have my eeepc updated
11:09 true
11:09 I don't plan to do anything fancy there
11:09 erikos ok, cool
11:09 tomeu mainly reducing padding or adding scrollbars
11:11 erikos ok, think we can move on then
11:11 gregdek tomeu: So what's the work?  Is this "feature" or "bugfix" or what?  Any exposure to anything else?
11:11 tomeu gregdek: can be considered bugfix
11:11 gregdek Sounds like small tweaks, not really "feature-ish", low risk?
11:11 Okey doke.
11:11 tomeu right
11:11 gregdek Moving on.
11:12 erikos font size on other resolutions: have not worked on that
11:12 gregdek So nothing for 0.84?
11:12 Push to 0.86?
11:12 tomeu I would say bugfix as well
11:12 gregdek Oh, is it?
11:12 erikos yup
11:13 gregdek erikos?
11:13 tomeu should be a matter of tweaking one value in one file
11:13 gregdek OK.
11:13 Gotcha.
11:13 tomeu takes time testing in different hw and setups, that's most of it
11:13 gregdek Well, that's what the entire beta process is for.  ;)
11:13 Next: control panel resizing.
11:13 tomeu guess so, we'll have quite a bit of time
11:14 gregdek erikos?  control panel resizing?
11:14 erikos gregdek: bugfix as well
11:14 gregdek ok/
11:14 "Better separation of XO features in the control panel so they can be removed for distros".  morgs?
11:14 erikos the next one, i would be really fancy to get in
11:15 since it will effect string freeze
11:15 tomeu oh, true
11:15 eben resizing?
11:16 erikos this effects as well the power section in the control panel
11:16 tomeu eben: calculating the right size, I think
11:16 based on the length of the text
11:16 gregdek Wait...
11:16 erikos i just wonder how we can make this customizable by distros
11:16 gregdek ...I'm confused.  Which issue are we discussing?  control panel resizing, or separation of XO features in control panel?
11:16 tomeu erikos: I guess the right thing to do here is to autodetect the hardware
11:16 both
11:17 confusing, eh? ;)
11:17 erikos gregdek: our design just interrupted again ;)
11:17 gregdek :)
11:17 OK.
11:17 erikos tomeu: eben yup
11:17 gregdek So are these changes invasive, or not?  Both bugfixes, or not?
11:17 I can tag them both "bugfix".
11:18 If it's actually true.  :)
11:18 tomeu well, changes will be limited to that part of the code
11:18 specially the second
11:18 gregdek It's your release, folks.  :)  You tell me.
11:18 tomeu wonder if we should just take out the power panel if it isn't the XO
11:19 that should be easy an safe
11:19 erikos tomeu: yeah that could be fixed like that
11:19 tomeu gregdek: so bugfix
11:19 erikos tomeu: About my computer is a bit harder
11:19 gregdek So that's a very specific issue: "power panel".
11:19 tomeu "hide power panel if not XO"
11:19 erikos tomeu: but only the power panel is a bugfix
11:19 gregdek The more general question, "Better separation of XO features in the control panel so they can be removed for distro", is unaddressed.
11:20 tomeu erikos: yeah, though it's also a matter of hiding stuff when not in a XO, right?
11:20 gregdek: yeah, that would be the right fix, more sofisticated and probably less safe
11:20 gregdek hrms.
11:20 walterbender and adding a few things (logout)
11:21 erikos tomeu: yeah, i guess we can remove parts of the About section when not on an xo
11:21 tomeu gregdek: we can fix those as bugs past the freeze, and take care of them properly for 0.86
11:21 erikos walterbender: well that would mean making the palette customizable
11:22 gregdek Seems like it's dead easy to do a lot of simple hacks, and unclear complexity about how to "isolate" code paths based on "whether I'm an XO" or "whether I'm an eee" or other as-yet-unclear properties.
11:22 erikos tomeu: no
11:22 tomeu: we need to change the name of the section
11:22 tomeu erikos: ok, that can be done now, right?
11:22 erikos eben: is 'About my computer' ok with you?
11:22 tomeu just change the string unconditionally
11:22 gregdek Seems like there's a lot of complexity right under the surface here.  Spidey sense tingling.
11:22 erikos tomeu: jup - i just make sure it gets not lost
11:23 eben erikos: yeah, that's fine
11:23 tomeu gregdek: the right fix would be to have a generic way to autodetect hw capabilities
11:23 erikos eben: or 'About my machine' :)
11:23 tomeu gregdek: that can wait for 0.86
11:23 eben erikos: nah, computer is better
11:23 gregdek tomeu: OK.  I'll alter the feature description to say that.
11:23 tomeu for now, it's enough to hide some UI components based on the existence of one file in the file system
11:23 erikos eben: ok, taken!
11:24 gregdek Updating...
11:25 erikos gregdek: you can assign it to me
11:25 gregdek OK.
11:26 "Adding buddies not tracked in the PS."
11:26 Looks like we've already agreed to defer this?
11:26 tomeu 0.85
11:26 yeah, haven't started work on it
11:26 would be really cool, but...
11:26 gregdek Okey doke.
11:26 So...
11:27 ...what's the practical difference between saying "0.85" and "0.86"?
11:27 I don't want to confuse things unnecessarily.
11:27 We will track everything we defer against 0.86, although we will obviously do the work in 0.85.  Yes?
11:27 tomeu not sure
11:28 erikos i think it make sense to call things a 0.86 feature
11:28 tomeu I guess work in 0.86 is only stabilizing what was done in 0.85
11:28 gregdek So when we say "defer to 0.86", we mean "work in 0.85, release in 0.86".
11:28 All right.
11:28 updates wiki again.
11:29 Moving on:
11:29 Continue FT implementation in Gabble.  cassidy?
11:29 Or anyone?
11:30 cassidy gregdek: no progress atm. I've been pre-empted to work on others (tubes related) tasks. Someone else is suppose to start to work on it soonish
11:30 gregdek OK, so deferring to 0.86.
11:31 And then the activities.sl.o stuff is separate.
11:31 cassidy anyway, Sugar can still implement FT. The option would be just disabled if the CM doesn't support it
11:31 gregdek Got it.
11:31 cassidy we provide API to see if the CM supports FT or not
11:32 tomeu still need to implement that
11:32 cassidy in Empathy we greyed the FT button if it doesn't
11:32 Sugar could do something similar
11:32 gregdek Pardon my ignorance -- what is FT?
11:32 cassidy file transfer
11:32 gregdek Ah, ok.
11:33 So.
11:33 That's the feature list.  Any that we missed?
11:33 Any not reflected on that list that we need to discuss?
11:33 tomeu nope
11:33 gregdek All right.
11:34 The only one of those features that highlighted collaboration was the File Transfer feature, and that got deferred.  Should we be concerned about that?
11:34 tomeu gregdek: we have FT on salut (LAN)
11:35 gregdek Oh, is that new for 0.84?
11:35 Already reflected somewhere?
11:35 tomeu also, if a distro shipping sugar 0.84 happens to ship a gabble with FT, FT in sugar will work with that as well
11:35 without any extra coding
11:35 gregdek: in git ;)
11:35 though I blogged about it when I was working on it
11:35 gregdek tomeu: Perhaps we should add a trac ticket for the release notes?  :)
11:36 Since that appears to be our process?  :)
11:36 tomeu gregdek: if you find time to play with it and can file tickets would be nice, I'm sure the UI could be more polished
11:36 gregdek: I don't think we have now a process for tracking what to put in the release notes
11:36 gregdek Hmm, ok.
11:36 Need to think about that, then.
11:36 tomeu perhaps a single trac ticket would be enough?
11:36 gregdek Probably.
11:36 tomeu to list all the stuff needs to be in the release notes?
11:37 or a wiki page, not sure which advantage would have trac
11:37 gregdek In the meantime, we're at 1:40 meeting time.  Plenty of time to discuss release management process later.  :)
11:37 erikos tomeu: well we have the release notes from each development release
11:37 tomeu: and there it does show up
11:38 gregdek: i did make sure all the development release notes were accurate - to build the final release notes from
11:38 tomeu erikos: yeah, but I guess greg meant to note it down now so we remember later
11:38 gregdek erikos: Brilliant.
11:38 tomeu oooh
11:38 erikos: git it, you are right
11:38 s/git/got
11:38 erikos tomeu: git it too ;p
11:38 gregdek Any other business for today's meeting?
11:38 tomeu yeah
11:39 c)
11:39 c) How long to support "old" versions of Sugar (Debian lenny might ship 0.82)
11:39 silbe: I would say depends on the interest and the available resources
11:39 like with any distro or sugar release
11:40 erikos how does gnome handle that?
11:40 (i mean they have other resources, but in general)
11:40 tomeu no idea
11:40 gregdek No idea.
11:40 erikos and what does 'support' mean exactly
11:41 gregdek "Will you please backport this fix?"
11:41 tomeu maybe are distros the oens that support that stuff?
11:41 silbe well, mostly reacting to bug reports
11:41 gregdek It's gonna be a case-by-case basis, I think.
11:41 cjb likes the {Fedora,Ubuntu} {N, N-1} idea
11:41 tomeu silbe: so that's a business of the bugsquad
11:41 ?
11:41 silbe and for activity authors probably making stuff that's only in newer releases optional
11:41 gregdek I don't think we're mature enough to have much policy around this question yet.
11:41 cjb and other distros can package, of course, but we'll make sure we don't break deps for those
11:42 erikos gregdek: yeah :/
11:42 tomeu cjb: though is that good for upstream projects like gnome and sugar?
11:42 cjb tomeu: hm?
11:42 silbe not sure what your bugsquad does exactly
11:42 cjb tomeu: oh, maybe I wasn't clear
11:42 I mean that Sugar should make sure it always works on the latest and one-before-that of Fedora and Ubuntu
11:42 erikos silbe: being responsive to bug reports and triaging them
11:43 cjb i.e. we refuse new deps that would break any of those
11:43 tomeu cjb: oh
11:43 cjb: sounds good
11:43 gregdek Here's what the bugsquad SHOULD do: (a) they should test latest released versions and file bugs, which should be fixed primarily in development.  (b) they should test betas and file bugs, which should be fixed primarily in development.
11:43 silbe erikos: do they import patches as well?
11:43 cjb I guess there's a question about whether the "latest" Fedora would count as F10 or F11
11:43 probably F10
11:43 gregdek Yep.
11:43 erikos silbe: no, they need to communicate that then with the devs
11:44 silbe so policy for bugsquad is not to support older releases ("which should be fixed primarily in development")?
11:44 tomeu cjb: yeah, we develop in F10 and intrepid for F11 and jaunty, and depend on the platform we develop in
11:44 gregdek silbe: It's not bugsquad's job to figure out where the fixes go.  That's development's job.
11:44 silbe: And whether *development* chooses to backport a fix is a different question.
11:45 cjb tomeu: but should we support F9 too, or just F10/F11?
11:45 gregdek Which will probably be a function of:
11:45 silbe ok, so this topic is more about development than about the bugsquad
11:45 gregdek (a) severity; (b) available hands; (c) age.
11:45 erikos silbe: yes
11:45 tomeu cjb: I don't think 0.84 should be able to run in F9
11:45 cjb ok
11:45 so what I said about
11:45 n, n-1 released versions isn't tue
11:45 true
11:45 tomeu cjb: GNOME and other deps are converging quickly to what we want
11:46 cjb we're instead saying that we'll support the *latest* released version of Fedora and Ubuntu, but not the previous ones
11:46 tomeu cjb: well, F10 will ship with 0.82, but I don't think 0.84 will depend on anything not in F10
11:46 cjb there was a guy in #sugar yesterday asking for help running jhbuild in F9
11:46 tomeu cjb: yeah, more like that
11:46 jhbuild is different
11:46 cjb and he seemed to think it was unreasonable that we didn't support F9 any more
11:47 tomeu well, we may support jhbuild in F9 quite easily
11:47 cjb because it was only released ~7 months ago
11:47 tomeu he just needs to compile inside jhbuild gtk, etc
11:47 cjb yeah, I think he did get it working eventually
11:47 tomeu but we need someone who runs that distro to maintain the sysdeps file updated
11:47 gregdek Yeah, jhbuild is really a separate topic.
11:48 silbe and what about 0.84 and following on the latest debian release (i.e. lenny)? soon unsupported because too old?
11:48 erikos i think we can say that our 'support' is really driven by resources
11:48 tomeu but about supporting stuff in general, we can support more thing if more people jump on the train
11:48 gregdek Since the point of jhbuild is to be able to rebuild your entire dependency map, we need to make sure that jhbuild actually does that.
11:48 tomeu if we are half a dozen cats, then we cannot support all the distros in the world
11:49 gregdek: there's a xml file that specifies what needs to be built and what needs to be installed in each distro release
11:49 gregdek: we need people running older releases to maintain those files
11:49 gregdek tomeu: Yep.
11:49 Lemme take that process on.
11:50 tomeu gregdek: like this: http://git.sugarlabs.org/proje[…]eps/fedora-10.xml
11:50 gregdek I'm about to run jhbuild for F10 myself.  It'll be the first time I've done jhbuild in over a year.  :)
11:50 tomeu gregdek: nice, await a flood of tickets ;)
11:50 silbe gregdek: good look. squeakvm.org is down ATM :-/
11:50 gregdek LOL!
11:50 Ah, squeak.  The perpetual breaker of jhbuild.  ;)
11:50 silbe s/look/luck/
11:51 gregdek The more things change, the more they remain the same.
11:51 tomeu gregdek: any idea when people will start testing 0.83 in rawhide?
11:51 bertf silbe: use --skip squeak
11:51 gregdek tomeu: When we ask them to?  :)
11:51 silbe bertf: i just used "skip module" manually. fortunately only some meta package depended on it
11:52 gregdek All right, gents, unless there's any more pressing business, I've got to call this meeting and prep for my next one.  Any objections?
11:52 tomeu gregdek: any luck finding packagers?
11:52 we are going to release today or tomorrow
11:53 gregdek tomeu: I've been focusing more on OLPC packagers, tbh.
11:53 tomeu gregdek: hmm, couldn't be the same?
11:53 gregdek tomeu: Could be, yes.
11:53 I will bring it up in the next meeting.
11:53 tomeu oh, wait
11:53 gregdek Who is currently doing the packaging?
11:53 tomeu gregdek: OLPC packagers won't be putting 0.83 in rawhide?
11:53 erikos gregdek: me
11:53 gregdek It should just be a matter of taking over spec files, yes?
11:53 tomeu that's exactly what we want
11:54 I thought the plan of record was dumping OLPC-4 and working only in rawhide
11:54 so the next OLPC release would be based on F11
11:54 gregdek That is correct.
11:55 tomeu ok, so that's awesome
11:55 we are all happy :)
11:55 gregdek And my next meeting to discuss that is in less than an hour, so I've got to go.  :)
11:57 silbe @all: thanks for the infos (and sugar in general, of course). got to go as well, shop closing soon.
11:57 tomeu silbe: see you
11:57 erikos silbe: hope to have better answers to your support request soon
11:57 gregdek So then:
11:57 Meeting ends in 5...
11:57 4...
11:57 3...
11:57 2...
11:57 1...
11:57 1/2...
11:58 #endmeeting

Index | Today     Channels | Search | Join

Powered by ilbot/Modified.
Webmaster