Status Meetings:2006-10-11:Log
Jump to navigation
Jump to search
12:00 <@ss> Everyone open the Agenda please.
12:00 -!- froodian is now known as froodian|sorta
12:01 <@ss> We have a lot to go through, so let's get started.
12:01 <@mento> is pinolo coming?
12:01 <@ss> I hope so. He said he would, but I'm not sure
12:01 <@ss> If not, we can still make some decisions...
12:01 <@ss> And he can read the log later.
12:02 <@ss> First off is a1
12:02 -!- Desmond has joined #camino-mtg
12:02 <@ss> We're missing a few things that we wanted, but overall, we're looking good.
12:02 <@ss> I think we need to release a1 now.
12:02 <@ss> pinkerton: ^^
12:02 < Desmond> Couldn't download Colloquy, sorry for being late.
12:02 <@ss> I say Friday.
12:02 <@hwaara> what are major blockers?
12:02 <@froodian|sorta> I agree. We need to get something out.
12:02 <@froodian|sorta> kqueue
12:02 <@ss> hwaara: Nothing really "blocks" an alpha
12:02 <@froodian|sorta> menu cleanup
12:03 <@ss> But yeah, those two things that froodian|sorta mentioned.
12:03 <@ss> We can also just release a second alpha.
12:03 <@mento> ok. no minibranch, just a tag on 1_8, right?
12:03 <@ss> Yep
12:03 <@mento> good, that's easy
12:03 <@ss> Tag, update version numbers, release.
12:03 <@pinkerton> what about the bug where dragging files to the browser doesn't update?
12:03 <@ss> It's a bug. Alphas have bugs.
12:03 <@pinkerton> ok
12:03 <@ss> It's not a beta.
12:03 <@pinkerton> then sounds good to me
12:03 <@ss> It's not feature complete.
12:04 <@ss> mento: Friday work for you?
12:04 <@mento> yup
12:04 <@ss> Awesome.
12:04 <@ss> a1 is set for Friday.
12:04 <@pinkerton> yay!
12:04 <@ss> Now, the fun stuff.
12:04 <@froodian|sorta> I'd like to land menu cleanup before that
12:04 <@ss> pinkerton: Go SR menu cleanup.
12:04 <@ss> :)
12:04 <@pinkerton> bah
12:05 <@ss> smorgan said it was enough to have his review and ardissone|away's functional
review.
12:05 <@ss> Moving on.
12:05 <@ss> l10n
12:05 <@ss> Marcello has a lot of discussion on the wiki about moving l10n stuff into m.o's CVS
12:05 <@pinkerton> what changed in that last patch?
12:05 <@pinkerton> smorgan reviewed, but now there's a new patch
12:05 <@froodian|sorta> pinkerton: just string cleanup
12:05 <@froodian|sorta> i merged two strings that shouldn't have been merged
12:05 <@pinkerton> ah ok
12:06 <@pinkerton> rs
12:06 <@ss> Everyone please go read the wiki about this.
12:06 <@ss> mento especially
12:06 <@froodian|sorta> w00t
12:06 <@ss> And pinkerton.
12:06 <@mento> i've read it weeks ago
12:06 <@ss> We need to know 1) is this doable and 2) do we want to do it
12:06 <@ss> Then, if we have a "yes" to both of those, 3) how do we do it?
12:06 <@mento> 1. yes 2. biggest issue is access to the mofo repository
12:07 <@mento> which isn't really a big issue, if they want, they could cordon off camino-l10n
access to part of the tree
12:07 <@ss> With 2, we'd just have to get Marcello and probably Ludo CVS access.
12:07 <@mento> like they've done with regular l10n
12:07 <@ss> Ah, that's true, we could definitely do that.
12:07 -!- Pinolo has joined #camino-mtg
12:07 <@mento> we should talk to someone at mofo about that
12:07 < Pinolo> hi
12:08 <@ss> And here's Pinolo
12:08 <@mento> hey marcello
12:08 < Pinolo> sorry for being late
12:08 <@ss> Pinolo: We're talking about l10n right now. :)
12:08 <@pinkerton> yeah talk to mofo and see how they want to handle it, sounds like a good
idea to me
12:08 < Pinolo> right
12:09 < Pinolo> regarding adissone's doubts
12:09 <@ss> So, we can do it.
12:09 <@ss> Now is "how"?
12:09 <@mento> build system changes aren't really a big deal
12:09 <@ss> My biggest question is, would we just release one Camino?
12:09 <@mento> concern is size
12:10 <@hwaara> yes, please
12:10 <@ss> Right
12:10 <@ss> The size is about 4 MB in the dimmidge right now.
12:10 <@mento> yeah, we inflated more than that when we unified
12:10 < Pinolo> do you think for the beginning only me and ludo can check-in?
12:10 <@hwaara> I think it's worth it, to have one build for everyone
12:10 < Pinolo> I don't think keeping a smaller english camino is a problem
12:10 <@hwaara> not many other respectable mac apps separate US and "everyone else"...
12:11 < Pinolo> it's probably a matter of how you present this to the world
12:11 <@mento> well, going forward, if we integrate localizations, we should be producing
localized nightlies
12:11 <@hwaara> I think it also makes it easier for people who accidently download the US build
because it's the first thing they se
12:11 <@mento> seems retarded to release two nightlies each day
12:11 <@pinkerton> i'm a bit wary of upping our DL to 20MB
12:12 <@ss> mento: I don't necessarily agree.
12:12 <@ss> Because what if some strings are missing?
12:12 <@ss> The nightlies would be half-assed for l10n
12:12 <@mento> then that l10n is busted but the app as a whole is ok
12:12 <@ss> And there _will_ be strings missing.
12:12 <@mento> you want to do en-only l10ns? then we're limiting nightly-level l10n testing,
that sucks
12:12 <@ss> No doubt
12:12 < Pinolo> yes. that might be the case
12:12 <@ss> We add new ones too often.
12:13 <@ss> Hm
12:13 <@mento> for releases, i'm still not convinced that we want a single
12:13 <@froodian|sorta> we'll have to answer a lot of feedback
12:13 < Pinolo> what about localizing only branch builds
12:13 <@froodian|sorta> about broken strings
12:13 <@pinkerton> i'm with mento
12:13 <@froodian|sorta> I still answer feedback about the broken 1.0.2
12:14 <@mento> can we move more stuff out of nibs and into strings files?
12:14 <@mento> make the nibs not per-l10n?
12:14 <@froodian|sorta> i believe it's possible
12:14 < Pinolo> mento: moving stuff to strings is risky for the different lenght of strings
12:14 <@froodian|sorta> i don't know how
12:14 <@mento> right, that's why people usually do nibs :(
12:15 < Pinolo> we do a lot of box adjusting when localizing
12:15 <@froodian|sorta> if we don't want to have unified release builds, i'm not sure i see the
point
12:15 <@froodian|sorta> in the whole thing
12:16 -!- mode/#camino-mtg [+o Desmond] by froodian|sorta
12:16 -!- mode/#camino-mtg [+o Pinolo] by froodian|sorta
12:16 <@Pinolo> froodian: the point is to automate some way the build of ML
12:16 <@Pinolo> nowadays I build it
12:16 <@ss> Yeah, there's definitely a point to it either way.
12:16 <@froodian|sorta> ok
12:16 <@Pinolo> where "build" means that I launch a bunch of scripts and chech some stuff
12:17 <@Pinolo> but before I have to locally collect all of the l10ns
12:17 <@Pinolo> moreover, to send the Ml pkg I need to be on broadband
12:17 <@ss> So, let's do this.
12:18 <@Pinolo> yeah
12:18 <@ss> 1) Talk to MoFo about a partition for l10n to check into
12:18 <@Pinolo> which are next steps
12:18 <@ss> 2) Setup our build system to release nightlies with the l10n stuff
12:18 <@ss> 3) Keep our releases separate for the time being.
12:18 <@hwaara> sorry, maybe I'm too late
12:18 <@mento> ok, on the size issue
12:18 <@ss> My main concern is outdated nibs.
12:18 <@hwaara> but can we do one nib, but still localize menus?
12:19 <@hwaara> I agree with doing en-only nightlies and shipping one final release
12:19 <@mento> if we convert all of the keyedobjects.nib files to text format
12:19 <@mento> they'll compress better
12:19 <@mento> but load more slowly
12:19 <@hwaara> right
12:19 -!- jcraig has joined #camino-mtg
12:19 <@hwaara> with the nibtool plist importer?
12:19 <@mento> yeah, or even plutil, same thign
12:20 <@Pinolo> mento when you say load do you mean startup or during Camino's run?
12:20 <@mento> whenever the nib is first loaded
12:21 <@mento> some at startup, some as camino is running
12:21 <@Pinolo> only first time
12:21 <@mento> right
12:21 <@Pinolo> so, for example when you first open a prefpane
12:21 <@mento> right
12:21 <@mento> shouldn't really be a perceptible difference, really
12:22 <@ss> So, in theory, we could have a unified release as well, depending on size.
12:22 <@Pinolo> I'm ok with all you decide about nibs/strings. we should test if that affects
the effort of localizing
12:22 <@mento> might turn the 4MB balloon into a 3MB balloon
12:23 <@Pinolo> ok
12:23 <@ss> Better... not ideal, but better.
12:23 <@ss> So, another final question is
12:23 <@ss> l10n of the license?
12:24 <@Pinolo> can't we compress everything only for packaging and then decompress only once
at first run?
12:24 <@ss> Is that possible? We have some licenses we can steal from MoFo/Co
12:24 <@mento> Pinolo: i'd like to avoid playing games like that. what if the app is
read-only? (on a cd? installed by another user?)
12:24 -!- mode/#camino-mtg [+o jcraig] by Desmond
12:24 <@Pinolo> mento: ok
12:24 -!- kreeger has joined #camino-mtg
12:25 <@mento> Pinolo: if you want to move in that direction, we should provide a .pkg or some
other installer. not really where i'd like to go if we don't need ot.
12:25 -!- mode/#camino-mtg [+o kreeger] by Desmond
12:25 <@mento> do NOT translate the license.
12:25 <@Pinolo> well...
12:25 <@ss> Heh
12:25 <@mento> mofo might have a jp translation, we can use that
12:25 <@ss> mento: I'm saying take the translations that already exist
12:25 <@mento> right, we can use those
12:26 <@Pinolo> I heard there are translations but at MozillaItalia they were discussing their
poor quality
12:26 <@Pinolo> at least for italian
12:26 <@ss> http://www.mozilla.com/legal/eula/
12:26 <@ss> They may be poor quality, but they're legal
12:26 <@Pinolo> technically it should be feasible
12:26 <@ss> And we can't break that legality, I believe.
12:27 <@mento> if the translations are bad, take it up with mofo
12:27 <@mento> they presumably commissioned those translations and are sufficient for eula
purposes
12:27 <@ss> We'd also need to s/Corporation/Foundation
12:27 <@ss> And I can talk to Frank and make sure that's okay.
12:27 <@Pinolo> I just read it
12:27 <@mento> co/fo and firefox/camino
12:27 <@ss> Right
12:27 <@Pinolo> it's ok, since it says that it's not legal ;)
12:28 <@ss> Alright, moving on?
12:28 <@ss> We can have follow up stuff next week on this
12:28 <@ss> To see where we are
12:28 <@ss> Tab scrolling was landed on the trunk
12:28 <@ss> Everybody say "yay" for Desmond
12:28 <@froodian|sorta> w00t
12:28 <@froodian|sorta> yay Desmond
12:29 <@Desmond> Thanks froodian|sorta :)
12:29 <@ss> Tinderbox upgrades are going to start taking place. :D
12:29 -!- smorgan has joined #camino-mtg
12:29 <@ss> Tell coop he's amazing if/when you see him.
12:29 -!- mode/#camino-mtg [+o smorgan] by froodian|sorta
12:29 <@Pinolo> wow
12:29 <@kreeger> Pinolo: what are you working on?
12:29 <@ss> Our review queue is a bit larger than most, but it's going alright.
12:30 <@kreeger> Desmond: yay
12:30 <@ss> Anyone have anything else before we go into specific bugs?
12:30 <@kreeger> (sorry i haven't been at my desk)
12:30 <@ss> kreeger: Follow along on the wiki.
12:30 !sand.mozilla.org ss invited thebot into the channel.
12:30 -!- thebot has joined #camino-mtg
12:30 <@ss> bug 350331
12:30 < thebot> ss: Bug https://bugzilla.mozilla.org/show_bug.cgi?id=350331 maj, --, Camino1.1,
nobody@mozilla.org, NEW, Pages sometimes fail to display when loaded; resizing
the window fixes display
12:31 <@ss> We have a regression range there, but we need someone to actually test backing out
patches.
12:31 <@pinkerton> i thought someone did
12:31 <@froodian|sorta> didn't torben just post something about that?
12:31 <@ss> Oh, heh
12:32 <@pinkerton> and i was glad to see that mofo minus'd the bug
12:32 <@kreeger> im having trouble pulling up the agenda on cb.o
12:32 <@ss> pinkerton: I talked to them.
12:32 <@ss> kreeger: WFM
12:32 <@kreeger> i think it is the wifi here
12:32 <@ss> mento: So that's your bug.
12:33 <@ss> I'm poking you now because of it. :)
12:33 <@ss> And the sooner they have a patch, the better probably. Given Fx2 and all, they'll
want a low risk patch for 2.0.1, yadayadayada
12:33 <@mento> eh. that's weird. ok
12:33 <@ss> mento: I'm assigning that bug to you. :)
12:34 <@mento> fine
12:34 -!- kreeger2 has joined #camino-mtg
12:34 <@ss> bug 356242
12:34 < thebot> ss: Bug https://bugzilla.mozilla.org/show_bug.cgi?id=356242 nor, --, ---,
stuart.morgan@alumni.case.edu, NEW, Make CHBrowserListener an informal protocol
12:34 -!- kreeger has quit [Quit: kreeger]
12:34 <@ss> smfr already said it was fine to do
12:34 <@ss> pinkerton: Any complaints from you?
12:34 <@ss> Gah, you and your food.
12:34 <@pinkerton> reading
12:34 < kreeger2> makes since
12:35 <@ss> Oh, you're really here...
12:35 < kreeger2> no nead to implement things there
12:35 <@pinkerton> how about this
12:35 <@hwaara> yeah, that makes sense
12:35 <@hwaara> or we could use delegates
12:35 < kreeger2> i.e. KeychainService
12:35 <@pinkerton> instead of making an informal protocol, break it into chunks
12:35 <@pinkerton> some formal, some informal?
12:35 <@hwaara> pinkerton: popupBlockDelegate?
12:35 <@hwaara> etc
12:35 <@pinkerton> just a suggestion, i dunno if that would work
12:36 <@smorgan> pinkerton: I'll look at that
12:36 < kreeger2> feedDetectedDelegate , etc.
12:36 <@hwaara> smorgan: what do you think about using delegates instead?
12:36 <@smorgan> What do you mean using delegates?
12:37 <@hwaara> actually, since there's only one object listening to gecko, I suppose that
wouldn't make sense
12:37 <@hwaara> an informal protocol sounds good
12:38 <@ss> Next?
12:38 <@ss> bug 306396
12:38 < thebot> ss: Bug https://bugzilla.mozilla.org/show_bug.cgi?id=306396 min, --, Camino1.1,
stuart.morgan@alumni.case.edu, ASSI, Don't show wyciwyg URLs in location bar
12:39 <@ss> Basically, pinkerton & mento... any reason we should use nsIURIFixup instead of our
own homerolled version?
12:39 * pinkerton shrugs. sharing code is good
12:39 <@mento> because it's already written and it works
12:40 <@ardissone|away> smorgan: what all do we have to pull in to use it?
12:40 <@smorgan> Do we want user:pass hiding, or should that pref change for Camino?
12:40 <@smorgan> nsIURIFixup, or whatever it's called
12:41 <@smorgan> I'm not clear on what kind of cost we incur from pulling in a new component
12:41 <@ardissone|away> dylibs/xpts, or just #import stuff?
12:41 -!- ardissone|away is now known as ardissone|food
12:42 <@smorgan> This is where my moz internals knowledge breaks down. It's one of those
things with a magic FOO_CONTRACT_ID value.
12:44 <@smorgan> Which as I understand it we have to list in some registry of things we want,
and maybe make project changes, etc. Somebody who knows this stuff should
chime in...
12:44 < kreeger2> smorgan: chances are it is already in a dylib that we copy
12:45 <@smorgan> As an aside, is there some set of documents somewhere that actually explains
any of this stuff?
12:45 < kreeger2> smorgan: looks like it is sitting in base/docshell
12:45 < kreeger2> im not sure which dylib that is w/o looking at the makefiles
12:46 <@hwaara> we already have nsIURIFixup, I think? I think we actually use it?
12:46 < kreeger2> "@mozilla.org/docshell/urifixup;1"'
12:46 < kreeger2> is the contract ID
12:46 <@smorgan> Oh, we do. How did I miss that
12:46 <@hwaara> http://landfill.mozilla.org/mxr-test/mozilla/search?string=nsIURIFixup&find=%2Fcamino%2F&findi=&filter=&tree=mozilla
12:46 <@ardissone|food> so, do we want user/pass hiding?
12:47 <@ardissone|food> i like seeing them, personally
12:47 <@ardissone|food> it makes it clear I'm actually having to use them
12:48 <@Desmond> Am I still connected to the Internet?
12:49 <@smorgan> Ah, I missed it because we hard-code the CONTRACT_ID value, and I was looking
for that constant
12:49 <@froodian|sorta> Desmond: yes
12:51 <@Desmond> Excellent.
12:52 < kreeger2> got to run l84
12:53 < kreeger2> er l8r
12:53 -!- kreeger2 has quit [Quit: Chatzilla 0.9.75 [Firefox 2.0/2006100319]]
12:53 <@ss> Next?.
12:53 <@ardissone|food> so
12:53 <@ardissone|food> do we want user/pass hiding?
12:53 * ss knows little about this stuff.
12:53 <@ardissone|food> that's what we have left on smorgan's bug, right?
12:53 <@hwaara> user/pass hiding?
12:53 <@ardissone|food> i vote no hiding, but it it's a pref I can switch, I don't care
12:54 <@ardissone|food> when urls contain usernames/passwords
12:54 <@smorgan> We'll just do the default for now then
12:54 <@hwaara> oh
12:54 <@smorgan> It doesn't really have to be decided at this moment
12:54 <@ss> /m/m
12:54 <@ss> Whoops.
12:54 <@hwaara> doing it in URIs is inherently unsafe..
12:54 <@ss> Yeah, that sounds good.
12:55 <@ardissone|food> next?
12:55 <@ss> bug 350732
12:55 < thebot> ss: Bug https://bugzilla.mozilla.org/show_bug.cgi?id=350732 tri, --, Camino1.1,
nobody@mozilla.org, NEW, Going back with popup bar on screen truncates scroll
view
12:58 * ss looks around.
12:58 * ardissone|food wonders if everyone is eating?
12:58 <@froodian|sorta> everybody can repro, right?
12:58 * ss hasn't tried.
12:59 <@ss> But smorgan, you finally reproduced it. So do we need to talk about it?
12:59 <@Desmond> I am watching the football
12:59 <@Desmond> http://news.bbc.co.uk/sport1/hi/football/internationals/6039750.stm
12:59 <@smorgan> I couldn't in my oldish branch, but could in a built trunk
12:59 <@froodian|sorta> smorgan: you can repro the bug, yes? it's just not where you'd expect
it when debuggin
12:59 <@Desmond> Can people access the video?
12:59 <@Desmond> Sorry for OT
12:59 <@froodian|sorta> ah
12:59 <@froodian|sorta> Desmond: ->#camino
12:59 <@smorgan> I'm just not seeing how it happens
12:59 -!- hwaara has quit [Quit: hwaara]
12:59 <@smorgan> I don't know why it's on the agenda, so don't look at me
13:00 -!- hwaara has joined #camino-mtg
13:00 -!- mode/#camino-mtg [+o hwaara] by ChanServ
13:00 <@froodian|sorta> yeah, i don't think there's anything to say about it
13:00 <@ss> Alright, next.
13:00 <@hwaara> last time I tried, I couldn't repro it
13:00 <@hwaara> I was talking to ardissone|food about it
13:00 <@ss> bug 355080
13:00 <@froodian|sorta> I like option 2
13:00 < thebot> ss: Bug https://bugzilla.mozilla.org/show_bug.cgi?id=355080 enh, --, Camino1.1,
nobody@mozilla.org, NEW, prettify about:config warning
13:00 <@hwaara> so, can someone give a simpler testcase on that bug?
13:00 <@ardissone|food> i like option 1
13:01 <@smorgan> hwaara: There is one in the bug. Go to a scrolling page, go to the popup
test, then go back
13:01 <@hwaara> smorgan: last time I wanted to debug it, I couldn't repro with bugzilla &
popuptest
13:02 <@froodian|sorta> well, we can talk about that some other time, since there are no
decisions to be made about it specifically
13:02 <@froodian|sorta> imo
13:02 <@ardissone|food> ?
13:02 <@ardissone|food> 1vs2 is not a decision?
13:03 <@froodian|sorta> that = the popup blocker bug
13:03 <@ardissone|food> ah
13:03 <@ss> I like 2, I think.
13:03 <@froodian|sorta> i like the extra whitespace
13:03 <@froodian|sorta> especially since it takes up the whole content pane
13:03 <@froodian|sorta> i think 1 would look small and alone in the center of the page
13:04 <@ardissone|food> 1: https://bugzilla.mozilla.org/attachment.cgi?id=241287&action=view
13:04 <@ardissone|food> 2 = https://bugzilla.mozilla.org/attachment.cgi?id=241288&action=view
13:04 <@ardissone|food> pinkerton: is 2 ok with you?
13:05 <@pinkerton> looking
13:05 <@ardissone|food> the spacing above the line is what i find odd, fwiw
13:05 <@pinkerton> 2s fine
13:05 <@ss> Oh, and one more thing (when pinkerton makes his decision)...
13:06 <@ss> Are we all okay with waiting to land Keychain stuff til after a1?
13:06 <@ss> I'm leery of any potential regressions.
13:06 <@ardissone|food> yes
13:06 <@pinkerton> sounds good
13:06 <@ss> Awesome.
13:06 <@ardissone|food> 1.2 = keychain + tabs stuff ;)
13:06 <@ardissone|food> as a quick follow-up to 1.1 ;)
13:06 <@froodian|sorta> there's been discussion about that...
13:06 <@froodian|sorta> which we shouldn't have now
13:06 <@froodian|sorta> imo
13:06 * ss strongly believes there won't (and shouldn't) be a 1.2. But that's for late.
13:06 <@ardissone|food> right
13:06 <@ss> later*
13:07 <@ss> Thanks everyone for coming
13:07 <@ss> Have a great day/night/evening
13:07 <@ss> See you all next week
