| micahg | bjsnider: thanks for the merge requests, could you repropose against the branches without .daily in them? | 01:32 |
|---|---|---|
| bjsnider | micahg, done | 01:52 |
| bjsnider | is there anything i can do about the issue of adding precise? | 01:53 |
| micahg | bjsnider: chrisccoulson has to do that | 01:53 |
| micahg | I think he was waiting for the builds to pass, so since he fixed trunk, I'm guessing he might have enabled it | 01:53 |
| micahg | bjsnider: and thanks, I"ll look at the branches in a bit | 01:53 |
| bjsnider | i ran a stable pbuilder build and it failed after the fix, but for reasons that don't make sense, so i'm thinking it's just my pbuilder environment | 01:54 |
| bjsnider | stable branch i mean | 01:55 |
| micahg | heh, I have to look at stable later anyways since Chromium released and I need to get it working on Lucid-Precise | 01:55 |
| micahg | or rather Chrome releases | 01:55 |
| micahg | *released | 01:55 |
| micahg | what error did you get | 01:55 |
| bjsnider | you need to get chrome working? | 01:56 |
| micahg | no, chromium | 01:56 |
| bjsnider | http://pastebin.com/B0a6WUgC | 01:56 |
| bjsnider | that's teh build failure | 01:56 |
| micahg | but there's no upstream Chromium releases, only Chrome | 01:56 |
| micahg | well, it could mean we're missing a build dep or something | 01:58 |
| micahg | or just a bug in that file | 01:58 |
| bjsnider | http://src.chromium.org/viewvc/chrome/trunk/src/chrome/build_nacl_irt.py?r1=101659&r2=106982 | 01:59 |
| bjsnider | that suggests irt.nexe is missing | 01:59 |
| bjsnider | but get-orig-source should have pulled that in | 02:00 |
| bjsnider | i checked the tarball and it isn't there | 02:00 |
| bjsnider | but the builds succeeded even after that commit, so i think this is just a quirk of my system | 02:00 |
| bjsnider | i think if the ppa build fails that's the time to worry about it | 02:01 |
| micahg | I have native client disabled on stable | 02:01 |
| micahg | it was breaking the build before | 02:01 |
| bjsnider | ok | 02:01 |
| chrisccoulson | grrrr, that stupid kde patch again | 03:41 |
| chrisccoulson | i enable it for a single day and it's already broken | 03:41 |
| chrisccoulson | oh well, at least it didn't break the precise builds :) | 03:42 |
| micahg | chrisccoulson: speaking of precise, can you enable it for chromium when you get a chance? | 04:02 |
| micahg | bjsnider: in the future, can you branch from the non-daily branches as well? | 04:13 |
| * micahg will manually merge the patches this time | 04:13 | |
| micahg | unless you're still around | 04:13 |
| bjsnider | sure i can | 04:13 |
| micahg | bjsnider: ok, you want to fix now or next time? | 04:13 |
| bjsnider | next time | 04:14 |
| micahg | ok | 04:14 |
| micahg | and please also add the new snapshot commit before the changes as well next time, I'll take care of it this time | 04:14 |
| bjsnider | what do you mean? | 04:16 |
| micahg | ideally, it shows what version the changes are made at | 04:16 |
| micahg | for build failures, stuff like this: http://bazaar.launchpad.net/~chromium-team/chromium-browser/chromium-browser.dev/revision/565 | 04:17 |
| micahg | before you do the changes | 04:17 |
| micahg | those commits don't exist in the non-.daily branches | 04:18 |
| bjsnider | you mean you want me to add something like "* New upstream snapshot: 15.0.861.0~r97996" | 04:24 |
| bjsnider | wait, he's changing the changelog and i'm not | 04:24 |
| bjsnider | i don't get it | 04:24 |
| micahg | take a look at this once LP updates it: https://code.launchpad.net/~chromium-team/chromium-browser/chromium-browser.dev | 04:25 |
| bjsnider | yes? | 04:27 |
| micahg | 568 says which revision failed, 569 is one fix, 570 is another | 04:27 |
| micahg | if we consistently fix after breakage, it makes merging into the next channel easier when the version is bumped | 04:28 |
| micahg | you just pull all the fixes down | 04:28 |
| bjsnider | ok, so you want me to add 568 in other words | 04:28 |
| micahg | right | 04:28 |
| bjsnider | no problem | 04:28 |
| micahg | thanks | 04:29 |
| micahg | bjsnider: BTW, it took me almost 5 months to get commit rights to the mozilla team bzr branches, even after that I still was learning more | 04:29 |
| micahg | not saying it'll necessarily take that long, but there's a learning curve, but once you get what's going on, you can fly :) | 04:30 |
| bjsnider | like i told you earlier i'm used to just acting unilaterally | 04:30 |
| bjsnider | but i also have access to fta to ask about build failures and packaging issues | 04:31 |
| micahg | right, Ubuntu doesn't work too well like that in general, there are times to act unilaterally, but usually it's a coordinated effort amongst team members | 04:31 |
| bjsnider | he's undoubtedly going to read this irc log at some point, so -- hi fta! | 04:31 |
| bjsnider | i should have known not to edit the daily branches because i run one of the bots for another project | 04:33 |
| micahg | heh, mistakes will be made, that's why we review branches initially | 04:39 |
| micahg | heck, mistakes will be made afterwards as well, that's why we have peer review | 04:39 |
| chrisccoulson | lol @ https://twitter.com/#!/bhearsum/status/147328159996579840 ;) | 15:03 |
| === m_conley_away is now known as m_conley | ||
| chrisccoulson | bhearsum, if i download an official build of firefox 8.0.1 and then set app.update.desiredChannel to "beta", is it meant to update to the latest beta? | 16:14 |
| chrisccoulson | or is there another method to switch channels? | 16:15 |
| kbrosnan | channel switching is not supported | 16:15 |
| chrisccoulson | oh, that's probably why it offers me a 5.0 beta then ;) | 16:15 |
| chrisccoulson | well, actually, it just fails. but the update response is for 5.0 | 16:15 |
| bhearsum | you can't adjust the channel in the product | 16:16 |
| bhearsum | you can do it by adjusting defaults/pref/channel-prefs.js though - be sure to shut down the browser before changing that | 16:16 |
| bhearsum | but no, 8.0.1 _shouldn't_ give you an offer for any beta | 16:16 |
| kbrosnan | you can change the pref in nightly/defaults/pref channel-prefs.js but it will result in unpredictable results | 16:16 |
| bhearsum | our release and beta streams are completely separate now | 16:17 |
| chrisccoulson | ok, thanks | 16:17 |
| bhearsum | kbrosnan: not really unpredictible | 16:17 |
| bhearsum | maybe non-obvious though | 16:17 |
| kbrosnan | a bit impercise but mainly was saying that the switching was untested | 16:18 |
| kbrosnan | and when we did test it. it resulted in some odd states | 16:18 |
| chrisccoulson | what i was trying to figure out is that if someone installs the release, switches to beta and then decides they want to go back to release, does test pilot remain enabled? | 16:19 |
| kbrosnan | where an aurora folder/icon/label was updateing on beta | 16:19 |
| bhearsum | ah, unpredictible in terms of the state of the app afterwards | 16:19 |
| bhearsum | chrisccoulson: IIRC yes | 16:19 |
| bhearsum | that's dependent on the channel name alone, i think | 16:19 |
| chrisccoulson | just for context, we currently install test pilot in our beta builds in the application directory, which means that it gets removed if someone decides they want to use a release build again | 16:20 |
| chrisccoulson | the reason for this is that we have all users of our development release using the beta | 16:21 |
| chrisccoulson | and didn't want them all to permanently end up with test pilot | 16:22 |
| bhearsum | ah | 16:22 |
| bhearsum | hmm, i might be confusing Test Pilot with Feedback | 16:22 |
| chrisccoulson | but i'm considering installing test pilot as a distribution addon like in the official builds, so that the addon manager can update it | 16:22 |
| chrisccoulson | but that means that anybody who runs the development version of ubuntu will end up with test pilot installed until they manually remove it | 16:23 |
| bhearsum | yeah... | 16:23 |
| bhearsum | distribution addon means it ends up in the profile, right? | 16:23 |
| chrisccoulson | yeah | 16:24 |
| bhearsum | hm | 16:24 |
| bhearsum | http://mxr.mozilla.org/mozilla-central/source/browser/app/profile/extensions/testpilot@labs.mozilla.com/modules/setup.js#189 makes me think that it will be disabled when the channel is switched | 16:28 |
| chrisccoulson | bhearsum, yeah, i just tried that. the feedback button disappears :) | 16:29 |
| bhearsum | nice! | 16:29 |
| chrisccoulson | so that might be ok then. i don't think it would be the end of the world if people who are currently running precise end up with test pilot installed in their profile :) | 16:29 |
| kbrosnan | the button disapears, afik test pilot test may still run if the study allows that version | 16:30 |
| chrisccoulson | and it obviously doesn't matter for people who opted in to the beta on our stable releases either | 16:30 |
| chrisccoulson | kbrosnan, i think that's probably ok. unless you can think of a reason why it shouldn't run :) | 16:30 |
| kbrosnan | nope, just enumerating possibilities | 16:31 |
| chrisccoulson | thanks | 16:31 |
| chrisccoulson | ok, i'm going to switch it to being a distribution addon in ubuntu then, so that people don't keep reporting bugs about being pestered to install the latest version of test pilot, when they can't :) | 16:32 |
| kbrosnan | btw damn cool that you have test pilot setup | 16:32 |
| chrisccoulson | we try to offer an experience that closely matches the upstream one, with the exception that it's delivered through our package manager instead :) | 16:32 |
| bhearsum | and that's why you guys are so awesome | 16:33 |
| chrisccoulson | heh, thanks ;) | 16:33 |
| alex_mayorga | Is the lack of about:crashes on firefox from https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa intentional? | 16:55 |
| chrisccoulson | alex_mayorga, yes, we don't submit crash symbols for those builds. we do that for release + beta though | 16:56 |
| alex_mayorga | chrisccoulson: what's the rationale? | 16:57 |
| chrisccoulson | alex_mayorga, too much data ;) | 16:57 |
| alex_mayorga | chrisccoulson: so if I hope to help Mozilla I should switch to http://nightly.mozilla.org/ ? | 16:58 |
| chrisccoulson | alex_mayorga, yes, you can do that for nightlies | 16:58 |
| alex_mayorga | chrisccoulson: what's the function of "Ubuntu Mozilla Daily Build" then? | 17:00 |
| chrisccoulson | alex_mayorga, so we don't get any nasty surprises at beta :) | 17:00 |
| chrisccoulson | of course, you can still run it if you wish, but without the crash reporter for the time being :( | 17:01 |
| alex_mayorga | what would it take to have about:crashes? perhaps that can be my 1st patch ever ;-) | 17:01 |
| chrisccoulson | alex_mayorga, well, it's just a matter of turning on the crash reporter. but then the debug symbols have to go to mozilla | 17:02 |
| chrisccoulson | we do that for release and beta already, but the amount of data we'd be submitting would be crazy if we did it for nightly builds too | 17:03 |
| chrisccoulson | (1 build / release, 2 architectures / build for Firefox + Thunderbird, nearly every day) | 17:03 |
| alex_mayorga | I see | 17:04 |
| chrisccoulson | that ends up being around 400MB (compressed) every day | 17:04 |
| chrisccoulson | and those need to be stored somewhere ;) | 17:05 |
| alex_mayorga | have there been talks with mozilla if they want that? | 17:08 |
| chrisccoulson | no, but i'm not sure there are enough users of our nightly builds to justify it :) | 17:09 |
| chrisccoulson | of couse, i could be wrong | 17:09 |
| alex_mayorga | any way to obtain a "guesstimate"? | 17:12 |
| chrisccoulson | i think launchpad has an API for counting downloads | 17:12 |
| chrisccoulson | i'm not sure of a better way though | 17:13 |
| bjsnider | micahg, we have an el problemo with the new chromium 18: http://code.google.com/p/chromium/issues/detail?id=107585 | 19:37 |
| bjsnider | i think it might be fixed quickly though so we should sit on it | 19:38 |
| bjsnider | something is awry with this commit: http://bazaar.launchpad.net/~chromium-team/chromium-browser/chromium-browser.head/revision/845 | 20:55 |
| bjsnider | after that dbg packages stopped being created properly | 20:55 |
| micahg | I don't see anything getting installed there in debian/rules | 21:00 |
| bjsnider | i think i'll revert all of that and run it in pbuilder to see if a proper dbg package is created | 21:04 |
| micahg | well, we don't want breakpad, the only thing that looks like it might help is the linux symbols stuff | 21:05 |
| === m_conley is now known as m_conley_away | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!