[05:41] joelesko: unfortunately, I'm out of time to do a proper review for your branch and iterate, but I will give you credit for whatever you've done for the upload, I"ll also comment on what changes I made [05:42] joelesko: we'll get into a routine for P, it's this initial huge merge that's been why I've pushed it off, future merges should be much simpler to review [05:44] joelesko: I"m even happy to have you track the beta channel if you like until we get close to the end [08:31] chrisccoulson: FYI, I just pushed up a new seamonkey head branch to be the new lp:seamonkey [09:13] chrisccoulson: you around? [09:13] micahg, yeah, i'm around now [09:14] chrisccoulson: so, I just want to confirm about seamonkey, we're switching to a single binary package with upstream defaults for bundled libs [09:14] micahg, yeah, makes sense [09:14] ok, lintian's unhappy, but not fatally so with the package, I guess we can fix that for P [09:15] I pulled in the 4 patches from thunderbird for various build failures, I hope I got them all [09:33] chrisccoulson: does your new tarball build system use /tmp? I had that when creating the seamonkey tarball [09:38] micahg, no, it doesn't use that [09:38] hmm...why did the seamonkey tarball use /tmp then? [09:38] micahg, does the new seamonkey drop the patch which renamed the binary to seamonkey2? [09:38] we never had such a patch in the distro [09:39] micahg, we did. i'm not sure why we did though [09:39] you mean the upstream binary? yes [09:42] micahg, http://bazaar.launchpad.net/~mozillateam/seamonkey/seamonkey.head/view/246/debian/patches/seamonkey-fsh.patch [09:42] that patch has gone now, which is good (it shouldn't really have ever been there) [09:42] but it means that the profile location moves [09:43] chrisccoulson: oh :( [09:43] chrisccoulson: is there a quick fix? (I can get the upload rejected) [09:44] micahg, yeah, you just need to check if ~/.mozilla/seamonkey-2.0 exists, and move it to ~/.mozilla/seamonkey if it does [09:44] (and if ~/.mozilla/seamonkey already exists, just move it to ~/.mozilla/seamonkey.old or something) [09:45] chrisccoulson: do we still need LD_LIBRARY_PATH hacks with this version? [09:46] micahg, not sure if seamonkey has been updated to use the glue to load libxul [09:47] if it has, then you don't need it. if it hasn't, then you do [09:47] ok, will leave it for now [09:47] thunderbird needs it still [09:59] chrisccoulson: hmm, the old version seems to be using the seamonkey dir as well [10:10] chrisccoulson: hmm, the old version seems to be using the seamonkey dir as well [10:10] hmm, i'm a bit surprised about that [10:11] me too, I thought that's why we changed that in the first place [10:12] micahg, what about lucid -> natty? [10:12] chrisccoulson: don't know [10:12] i know we have some releases where it's been renamed to seamonkey-2.0, as i keep seeing it in bug reports [10:13] yeah, I can check that later [10:14] the binary is renamed, but uses the seamonkey profile dir in oneiric [10:14] I guess if there's an upgrade issue, I can push out an SRU [10:15] micahg, oh, yeah, i see now [10:15] ok, that's a bit weird [10:15] firefox uses the app name as the profile folder name too [10:15] so we're ok there [10:15] sorry about the noise ;) [10:16] heh, ok, will ask for review/acceptance then... [10:16] chrisccoulson: wasn't planning on sleeping before sunrise anyways I guess :) [10:27] chrisccoulson: so, when I was generating the tarball, it couldn't find the tag for 2.4.1 in the l10n repos [10:28] for any specific locales? [10:28] all [10:28] hmmm, i just grabbed en-GB, and it's there [10:28] :(, maybe I should have the tarball rejected and try to regenerate? [10:28] possibly [10:30] also, debian/config/branch.mk should be fixed to set CHANNEL to "release", and drop "NO_AUTO_REFRESH_LOCALES = 1" [10:30] the latter is used on nightlies to prevent the list of shipped translations being auto refreshed [10:31] if you remove that, you should get a warning about the list of translations being out-of-date [10:33] chrisccoulson: I just ran ./debian/rules get-orig-source DEBIAN_TAG=SEAMONKEY_2_4_1_RELEASE and I got no warning [10:33] ah, you get it when you try to build the source package (as it checks in the clean target) [10:34] ah, ok, so I need to build the source first? [10:35] yeah [10:35] i'm just trying it here as well [10:35] No description for be at debian/build/refresh-supported-locales line 145, <$file> line 1. [10:36] micahg, yeah, this is because it doesn't yet have a list of languages :) [10:37] it will helpfully get locale decriptions from langpack-o-matic though :) [10:40] oh, you're right about only having the nightly branding [10:40] i never knew that before ;) [10:41] yeah, that's one reason I didn't care so much about non-bundled libs before [10:42] chrisccoulson: still not sure how to regenerate the locale list [10:43] micahg, debian/rules refresh-supported-locales TARBALL=/path/to/tarball LANGPACK_O_MATIC=/path/to/langpack-o-matic should do it [10:43] which you can get from lp:langpack-o-matic [10:44] micahg, i just found another issue with the tarball though [10:44] create-tarball needs this - http://bazaar.launchpad.net/~mozillateam/thunderbird/thunderbird-trunk.head/revision/457?remember=455&compare_revid=455 [10:44] i wish i had time to review this too :) [10:46] hmm, well, it seemed to build the mail files as well, but I can regenerate and see what happens [10:46] ah, but it might have pulled the tip instead of the tag of the branch... [10:47] yeah, that's what it's doing here [10:50] chrisccoulson: so, I need to redo the tarball before redoing the locales? [10:51] micahg, yeah, i think so. what does suite/config/version.txt contain in your tarball? [10:51] if it's 2.4.1, then it's ok [10:51] but i've got 2.4 here, which is the current tip [10:52] micahg, actually [10:52] it's 2.4.1 [10:52] it has updated to the correct version here now [10:53] perhaps --comm-rev isn't actually needed? [10:53] i got the impression that client.py just did a "hg update" without that [10:53] I think it's the default maybe? [10:53] ok, that's ok for now, but we should pass --comm-rev in future builds, just to make sure we really do get the right version :) [10:54] ok, I'll include that change in the upload [10:57] thanks [10:57] yeah, i think it might have been a stroke of luck that we got the right version :) [10:58] it does just do "hg update -r default" [10:58] micahg, ok i managed to create a tarball here with all the translations in it [10:59] chrisccoulson: so, I need a new tarball? [10:59] micahg, if they're not in the current one [10:59] perhaps we can leave this until the next version though [10:59] seeing as it's already accepted anyway [10:59] yeah, they seem to be there, so I think we're ok [11:00] I'm testing a build now to make sure that the langpack binaries aren't empty, then I'll upload [11:00] did you manage to generate everything ok? [11:00] yeah [11:01] * micahg adds a README.Source note [11:03] micahg, yeah, i also changed this in firefox recently to make it all a single step - http://bazaar.launchpad.net/~mozillateam/firefox/firefox-trunk.head/revision/1002 [11:04] I'll have to fix that later [11:04] so now, running debian/rules refresh-supported-locales will update the control file too [11:18] * micahg wishes Firefox 4+ supported -jX [11:37] chrisccoulson: build failed :( *** No rule to make target `/run/shm/seamonkey-2.4.1/build-tree/mozilla/l10n/be/suite/installer/unix/README [11:38] micahg, ah, you'll need to leave that off for now then [11:38] which? [11:38] the problem is that the EXCLUDE list in create-tarball hasn't been updated for seamonkey [11:38] so it's missing the suite/ folder for each locale [11:39] ah, ok, well, we'll fix for P [11:47] * micahg -> zzz [12:22] BTW, thunderbird isn't happy moving 20k messages between IMAP folders... [12:22] 23006 micah 20 0 9337m 4.6g 15m S 0 29.6 182:01.85 thunderbird-bin === m_conley_away is now known as m_conley [13:18] chrisccoulson: ping [13:28] anyone reproduces this https://code.google.com/p/chromium/issues/detail?id=99835&thanks=99835&ts=1318339615 [14:27] micahg: that's great news and I understand about the merge. [14:28] micahg: what do you mean 'track the beta channel'. I would like to learn and help. [14:31] hi m_conley [14:31] hi joelesko [14:31] thanks for helping out with seamonkey :) [14:31] chrisccoulson: hey! How was your weekend? [14:31] m_conley, yeah, not too bad thanks [14:31] and you? [14:31] your welcome. Thanks for the advice and providing the initial package. [14:31] chrisccoulson: it was good! Thanksgiving in Canada - so lots of feasting. Ate way too much. [14:32] heh :) [14:32] chrisccoulson: Just wanted to let you know that I've got a newer version of my EDS contacts add-on ready: 0.3.7a [14:32] chrisccoulson: fixes an annoying authentication loop problem [14:32] m_conley, excellent. does it work with 8.0? :) [14:32] i was going to ping you about that [14:32] i need to update it in our beta PPA :) [14:33] chrisccoulson: yes, it should - I'll need to bump compatibility though, hang on... [14:33] thanks [14:35] chrisccoulson: k, compatibility bumped. [14:36] thanks [15:03] So, am atrting to test my TB extension on TB8 Beta1. But the linux builds from Mozilla don't seem to work with libnotify. Are these built differently to the PPA? If so is there a comm-aurora, beta and central PPA anywhere? Am specifically interested in TB not FF [15:03] s/atrting/starting [19:08] joelesko: each upstream beta release (like 2.5.0 beta 2), we could upload to the next release once it opens if you like [19:19] * micahg wonders why people don't stick around long enough to get an answer... === chrisccoulson_ is now known as chrisccoulson === m_conley is now known as m_conley_away