[00:00] fta: did you create xul/ffox branches yet? [00:00] yes [00:00] maybe we want to start using the new approach: trunk is always trunk ;) ... and then when they spin of branches we do branches ;)? [00:01] dont think it matters much ;) [00:01] fta: how about renaming our ffox to 3.5/3.6 finally? [00:01] fta: what do I need to make this vids play on chromium ? http://p.bugabundo.net/ [00:01] at least we should get it done before we release 3.5 as stable ;) [00:01] (or even 3.6 ?;) [00:02] BUGabundo, it plays fine for me [00:02] we have to manage to get to a point for karmic+1 where we can do major version upgrades for security i guess [00:02] I mean, do I need to enable anything? like codecs? [00:02] otherwise we (I) will have endless pain [00:03] backporting stuff [00:03] in /etc/chromium-browser/default i have CHROMIUM_FLAGS="--enable-plugins" [00:03] fta: I prob don't === asac_ is now known as asac [00:14] $ bzr push --remember lp:~mozillateam/firefox/firefox-3.7.head [00:14] Using default stacking branch /~mozillateam/firefox/firefox-3.2.head at lp-45207824:///~mozillateam/firefox [00:14] Created new branch. [00:14] asac, ^^ ?? [00:17] asac: any test for me on MM/NM ? I didn't manage to try the diff [00:21] fta: whats the question. the stacked branch? [00:22] thats because if you dont say anything, bzr will try to stack on top of the branch that is associated with current development focus series in launchpad [00:22] stacked branch just meanst that you dont upload everything to launchpad. only those commits not in the stacked branch [00:22] ment to make things like branching faster on first upload [00:23] i know what a stacked branch is, but i don't know where it's getting that information from [00:30] fta: it defaults to what i said. otherwise you can specify it on command line explicitly somehow [00:30] but i didnt get that working when i tried [00:31] but could be that it was just broken at that point as it was quite hot [00:31] BUGabundo: thats ok. the patch wouldnt work anyway. it basically was the other way around [00:31] BUGabundo: please sudo apt-get build-dep modemmanager ... then git clone the modemmanager upstream branch and run the following commands: [00:31] sh autogen.sh --prefix=$HOME/local_mm [00:32] make [00:32] make install [00:32] (if last command fails because of permissions, just run sudo make install, but ry without first) [00:32] then i give you a patch that adds more debugging output [00:32] * BUGabundo awakes [00:32] BUGabundo: cloning mm is: git clone git://anongit.freedesktop.org/ModemManager/ModemManager [00:33] cd ModemManager [00:33] http://paste.ubuntu.com/253405/ [00:33] ... then the autogen stuff and so on from above [00:33] asac: calm down [00:33] fta: looks good [00:33] I'm still adding a source to my source.list [00:33] BUGabundo: the whole procedure will take 3 minutes [00:33] ;) [00:33] (after installing build-deps) [00:33] i mean, the bot is ready [00:34] fta: kick it off ;) ... or does it have a --dry-run feature ;) [00:34] nope [00:34] was kidding [00:34] fta: howa bout renaming our ffox branches? [00:35] do we want to defer that until this is set up? [00:36] yes, esp. that i'm not there this w-e [00:36] ok lets defer it then [00:39] * BUGabundo waits for aptitude to finish updating [00:43] #include "third_party/zlib/contrib/minizip/unzip.h" [00:43] #include "third_party/zlib/contrib/minizip/zip.h" [00:43] well done ;) [00:43] grrr, we don't have that in our system zlib dev [00:43] minizip sounds like a different project? [00:44] apt-cache search minizip [00:44] zlib-bin - compression library - sample programs [00:44] odd [00:44] ;) [00:44] http://www.winimage.com/zLibDll/minizip.html [00:44] winimage? [00:44] Now, with version 1.23 and 1.14 of zLib, the Minizip library is inlucded in the contrib/minizip directory. [00:44] step one done [00:44] BUGabundo: my own step 1? [00:44] the build-dep? [00:44] yes [00:44] !info zlib1g-dev [00:45] now git [00:45] zlib1g-dev (source: zlib): compression library - development. In component main, is optional. Version 1:1.2.3.3.dfsg-13ubuntu1 (karmic), package size 159 kB, installed size 380 kB [00:45] so do we have 1.23? [00:45] sudo apt-get install git-core [00:45] You need the source code of Zlib (zlib123.zip or zlib-1.2.3.tar.gz. Fo [00:45] bash: git: command not found [00:45] bah [00:45] what a version mess ;) [00:45] !info zlib-bin [00:45] zlib-bin (source: zlib): compression library - sample programs. In component universe, is optional. Version 1:1.2.3.3.dfsg-13ubuntu1 (karmic), package size 49 kB, installed size 132 kB [00:45] no idea [00:45] i assume we have 123 aka 1.2.3 aka 1.23 ;) [00:46] seems the 123 name is for old windows crap ;) [00:46] guess no long names supported;) [00:47] so i can't drop the zlib sources from the tarball :P [00:47] zlib-1.2.3.3.dfsg$ ls contrib/minizip/ [00:47] ChangeLogUnzip ioapi.c iowin32.c Makefile minizip.c mztools.h unzip.h zip.h [00:47] crypt.h ioapi.h iowin32.h miniunz.c mztools.c unzip.c zip.c [00:48] feels like we need to package the headers up somewhere [00:48] let me check [00:48] they are not shipped in -dev [00:49] i see that ;) [00:49] just want to check why [00:50] step two brgan [00:50] fta: they build it, but dont make install it for some reason in rules [00:50] git complete [00:51] $(MAKE) [00:51] $(MAKE) CC=$(DEB_HOST_GNU_TYPE)-gcc OPT="$(CFLAGS)" -C contrib/minizip [00:51] but [00:51] just [00:51] $(MAKE) prefix=$(CURDIR)/debian/tmp/usr install [00:51] install -d debian/tmp/lib [00:51] prefix? [00:51] sounds bogus [00:51] ;) [00:51] DESTDIR [00:51] folks dont use prefix for that ;) [00:51] indeed [00:51] might not matter, but for libtool and stuff it hurts [00:52] making [00:52] pfft [00:52] cat contrib/minizip/Makefile | pastebinit [00:52] http://paste.ubuntu.com/253416/ [00:52] its not ment to be installed as it seems ;) [00:53] its not even a lib or something. just a binary. [00:53] /usr/bin/install -c -m 644 77-mm-ericsson-mbm.rules 77-mm-zte-port-types.rules '/lib/udev/rules.d' [00:53] /usr/bin/install: cannot remove `/lib/udev/rules.d/77-mm-ericsson-mbm.rules': Permission denied [00:53] /usr/bin/install: cannot remove `/lib/udev/rules.d/77-mm-zte-port-types.rules': Permission denied [00:53] make[3]: *** [install-udevrulesDATA] Error 1 [00:54] guess I need sudo [00:54] so chromium rips stuff out of existing binaries now [00:54] BUGabundo: yeah [00:54] done [00:54] BUGabundo: ok after that you should be able to start the build modemmanager like sudo ./src/modem-manager --debug (after stopping NM, killing mm etc. as usually) [00:55] so basically similar to how you did before ... just with ./src/modem-manager ... instead of just modem-manager [00:55] ok [00:55] BUGabundo: check that it works and then wait for 2-3 minutes ;) [00:55] have to run somewhere first [00:56] brb [01:04] BUGabundo: so wget http://paste.ubuntu.com/253420/plain/ [01:04] and apply the file like: patch -p1 < FILENAME [01:04] (in ModemManager dir) [01:04] make; sudo make install [01:11] BUGabundo: wait a sec [01:12] http://paste.ubuntu.com/253422/plain/ [01:12] rather use that one [01:12] the other fails to build ;) ... hackish me [01:12] BUGabundo: still there ;)? [01:26] fta: did anyone ever complain about our move to migrate daily users to 3.5? [01:26] except the one who catched the temporary issue? [01:28] nope [01:28] odd [01:29] maybe noone uses < jaunty [01:29] or it was a perfect job ... what i cannot yet believe [01:29] https://edge.launchpad.net/builders/ [01:29] too bad ... armel dead [01:29] :/ [01:30] i guess we will never get tbird out [01:30] but all dailies ;) [01:30] makes me a bit happier [01:30] the armel situation frightens me though for sure [01:31] fta: hmm. actually i think we need to put 3.0 in the daily for the < karmic users to get fully migrated [01:31] now i am not even sure what state they are in [01:31] ;) [01:33] i think we dont have enough distinct icons for dailies anymore [01:33] shiretoko branding for 3.5 ... minefield for 3.6 ... 3.7 [01:33] granpardiso fo 3.0 ;) [01:33] we need to put the version on it [01:33] good that we can edit the icons ;) [01:34] maybe we want to patch the number in during build by getting it automagically ? [01:35] eh? [01:36] gasp http://launchpadlibrarian.net/30345461/buildlog_ubuntu-karmic-i386.firefox-3.7_3.7~a1~hg20090814r31579%2Bnobinonly-0ubuntu1~umd1_FAILEDTOBUILD.txt.gz [01:36] Applying patch awesome_browser_branding_install.patch [01:36] patching file browser/installer/unix/packages-static [01:36] Hunk #1 FAILED at 273. [01:38] convert -font helvetica -fill black -pointsize 30 -draw 'text 70,25 "3.6"' firefox-3.6.png firefox-3.6-new.png [01:38] guess thats not the best approach ;) [01:39] * asac wonders if one can teach quilt to dump the .rej files after failure so one can see how bad it is in the build log [01:41] * asac checks out the latest branches [01:55] fta: 3.7 depends on 1.9.2-dev [01:56] * asac looks at the packages patch [01:56] bad [01:56] yep, missed that one [01:56] fixing [01:58] asac, ^^ fixed [01:58] asac, #442 [01:59] woww, 442 commits for a single package :P [01:59] ok let me update then before committing patch fix [01:59] yeah. things getting old ;) [02:00] now i cannot test build anymore ;) [02:00] hehe [02:01] fta: so do we stop documenting diverged patches in changelog for 3.7+ ? [02:01] i mean unless for things that are getting released in to the archive [02:01] we should, it's useless for the end users [02:02] bzr log is enough [02:02] ok i agree. thats why i am asking [02:02] lol chromium, 284 commits in less than 280 days [02:03] maybe i should give it a rest and stop improving it [02:04] sometimes going slower, but bigger steps is more efficient. yes. [02:04] especially to prevent burnout ;) [02:05] xulrunner-1.9.3 amd64 is ready, if it's the one you need [02:05] but there also is the risk, that you get distracted by all the new time and the fall behing ;) [02:05] fta: no. i already committed stuff [02:09] fta: ok 3.6 seems to build ... /me should have looked at the real state [02:10] why dont they display the puzzle piece on frontpage ;) [02:10] 3.6? [02:10] fta: wanna kick 3.7 again= [02:10] ? [02:10] https://edge.launchpad.net/~ubuntu-mozilla-daily/+archive/ppa [02:10] i looked there and thought 3.6 had the same brokenness [02:10] but it was wait depend [02:10] yes, it's been like that for months [02:10] i know [02:11] i was just a bot and thought they probably landed stuff on both branches [02:11] most likely we will see the same breakage in a few days there too [02:13] so bugabundo probably failed ;) [02:16] http://code.google.com/p/chromium/issues/detail?id=19157 [02:23] fta, you there? [02:23] sort of [02:24] fta: would you mind pastebining your ~/.dput.cf file? [02:24] thx for rekicking 3.7 ;) [02:24] hmmm. why did todays prism build just fail on karmic? [02:25] run-mozilla.sh: Cannot execute ../../dist/bin/shlibsign. [02:26] why is that in there? [02:26] nss leftover? [02:27] pace_t_zulu, http://paste.ubuntu.com/253445/ [02:27] old style probably [02:29] fta, ty [02:33] asac, Estimated repository size: 9.0 GiB (90.41%) of 10.0 GiB [02:33] moz? [02:33] yes [02:33] yeah. we should bump it [02:35] anyone ever complained ... yet? [02:37] about what? [02:37] resource consumption ;) [02:42] no [02:47] fta: I think I fixed prism for karmic daily (couldn't sleep, so I got back to that) [02:47] Jazzva: why did it fail ? [02:47] what does $(wildcard $(MOZ_XPI_FILE)) do? [02:47] Jazzva: probably checks if a file exists [02:47] if it doesnt exist it returns empty [02:48] otherwise the filename [02:48] because of change in mozilla-devscripts "$(MOZ_XPI_FILE)" => "$(wildcard $(MOZ_XPI_FILE))" [02:48] or a list of filenames if the XPI_FILE would be a pttern [02:48] let me check [02:49] well, I noticed prism build on jaunty (which has m-d 0.12), and failed in karmic, which has 0.14, and I looked into diff and played around a bit, and found that after returning to $(MOZ_XPI_FILE) it builds locally [02:49] s/prism build/prism built/ [02:50] :/ [02:50] Jazzva: maybe this would help too: http://paste.ubuntu.com/253451/ ? [02:51] asac: I'll try that now [02:54] all my attempts to make the chromium tarball smaller are ruined after a few days.. http://www.sofaraway.org/ubuntu/tmp/chromium-tarballs-1.png [02:54] fta: "all" is kind of an exaggeration [02:55] asac: not working... [02:55] in nature nothing is a straight line [02:55] Jazzva: same issue? [02:55] mhm [02:57] asac, I drop more that 50% of the files now [02:58] Jazzva: http://paste.ubuntu.com/253453/ that works for me (e.g. make echo) [02:58] and i completely skip some huge deps (not counted in the 50%) [02:58] Jazzva: o sorry. remove the $(..) from the calls [02:59] e..g just $(call XPI_FILE) [02:59] but shouldnt matter [02:59] i mean the previous worked too [02:59] http://paste.ubuntu.com/253454/ [02:59] Jazzva: what is MOZ_XPI_FILE in prism? [02:59] asac: http://paste.ubuntu.com/253455/ here's the output :) [02:59] Jazzva: you need to put the file in the PWD named as Makefile [03:00] or use make -f FILENAME echo [03:00] hmm [03:00] but you have 1.2.3. ;) [03:00] odd [03:00] yes... [03:00] make --version [03:00] GNU Make 3.81 [03:00] though, when I type "man sh", it shows me dash's man page... [03:00] but I thought dash is default [03:00] oh [03:01] works in dash too [03:01] Jazzva: is that the output from the last thing i pasted? [03:01] now I'm confused... [03:01] asac: yes [03:01] http://paste.ubuntu.com/253454/ ? [03:02] http://paste.ubuntu.com/253454/plain/ ;) [03:02] Jazzva: oh [03:02] asac: http://paste.ubuntu.com/ [03:02] Jazzva: you need a file called COPYING in the current dir ;) [03:03] * asac feels completely off ;) [03:03] hmm... ok, that was it :) [03:03] wildcard COPYING doesnt give you anything if the file doesnt exist [03:03] which is the principal idea [03:03] Jazzva: so what is MOZ_XPI_FILE in prism? [03:03] refractor.xpi, [03:04] it's specifically named so in debian/rules, so xpi.mk should use that... [03:04] and does that exist? [03:04] Jazzva: maybe a typo? [03:04] asac: well, it should exist once it's build (which is before unzip) [03:04] *built [03:04] but does it exist in top level dir? [03:04] asac: let me check [03:05] asac: no, it's in ./dist/xpi-stage/refractor.xpi [03:06] http://paste.ubuntu.com/253458/ [03:06] Jazzva: we should do that [03:06] hmm. so thats why its broken everywhere [03:06] i thought we copy it to top level dir in some kind of post-build rule in rules [03:06] asac: well, at least now it's only broken in karmic :) [03:07] Jazzva: but thats odd [03:07] maybe < karmic it failed silently? [03:07] and we just dont have refactor xpi packaged? [03:07] asac: prism built ok in PPA in < karmic [03:08] yes. but it could just have failed silently somehow [03:08] prism is more than the .xpi [03:08] the .xpi is just a supplemental thing [03:08] Jazzva: so refractor.xpi:: build/prism [03:08] check that out [03:09] that basically should take care that its copied to top level dir [03:09] you sure its not in there? [03:09] (in debian/rules) [03:09] asac: unzip -d temp-xpi-SeaQkIOD ./refractor.xpi ... that's from jaunty log... [03:09] yes [03:09] but does it exist in karmic too? [03:10] it should [03:10] i mean we copy it right there in the rules [03:10] hmm [03:10] maybe the dependency does not properly trigger [03:10] oh yeah [03:10] asac: yeah, it was copied to top-level in jaunty [03:11] i think i know whats going on [03:11] the wildcard evaluates to NOTHING [03:11] and hence the copy never happens [03:12] I see... [03:12] stamp-unzip-%: $(XPI_FILE) [03:12] that one does not trigger the copy from the prism rule because at that time the .xpi does not even exist [03:14] so, to remove wildcard from XPI_FILE, and to test if that file exists in stamp-unzip-%, and to give error if it doesn't? [03:15] i think we should use a different hook in prism [03:15] that will fix it [03:15] and then also think about what to do with the $(XPI_FILE) requirement in xpi.mk [03:15] http://www.flickr.com/photos/36015194@N00/2260529558 [03:16] Jazzva: try to hook that copy thing into common-post-build:: [03:17] instead of refractor.xpi:: [03:17] you dont need the build/prism part then [03:17] nice fox [03:18] must be cold there ;) [03:19] ok, trying that now [03:41] Jazzva: ok i am off. we can check that tomorrow. but i think the common-post-build:: target is the way to go [03:42] asac: ok... i'm off too. at least we know why it fails :). [03:42] good night [03:42] yeah [03:42] 'night [04:29] fta, still around? === ripps_ is now known as ripps [11:04] asac: do you have time to merge the new upstream release branch of adblock-plus? [20:26] hi all [20:26] asac, I've sent a mail for you [20:26] asac, feel free to read it when you have some free time [22:38] boas [22:38] asac: I'm back to good old really old working version [22:38] :) [23:16] BUGabundo: did you get the log at least? [23:16] yes [23:17] forgot to upload it :) [23:17] good [23:17] where? [23:17] pff [23:17] well for my defence, I was networkless [23:17] :))) [23:17] to the bug report, of course [23:17] id? [23:17] no idea [23:17] was going to ask you the same [23:17] * BUGabundo turns to FF awesome bar [23:18] why do you think i remember your bug id more than you ;) [23:18] bug 413333 [23:18] Launchpad bug 413333 in modemmanager "[huawei] MM wont connect, keeps nm.applet spinning" [Undecided,New] https://launchpad.net/bugs/413333 [23:18] bug 412570 [23:18] Launchpad bug 412570 in modemmanager "[novatel] modemmanger does not connect to 3G network, icon keeps spinning. " [High,New] https://launchpad.net/bugs/412570 [23:18] luckily modemmanager bug list is short ;) [23:18] https://bugs.edge.launchpad.net/ubuntu/+source/modemmanager/+bug/413333/ [23:18] Ubuntu bug 413333 in modemmanager "[huawei] MM wont connect, keeps nm.applet spinning" [Undecided,New] [23:18] correct [23:18] BUGabundo: there is no new log [23:18] I filed it on MM?? [23:19] yes [23:19] wasn't it on NM? [23:19] no [23:19] asac: ill upload it NOW [23:19] there is no new log BUGabundo [23:19] ah ok [23:19] asac: I had no network [23:19] until now [23:19] and was way all day [23:19] driving karts [23:19] and taking pics with my android [23:19] just because of this modemmanager thing? cant believe it ; [23:19] ah ;) [23:19] asac: http://p.bugabundo.net/ [23:21] use ogg [23:21] upload [23:21] asac: I wish Android provided that codec [23:21] not even PNG :( [23:21] kart on sand? [23:21] wow [23:21] we only have tarmac things here ;) [23:21] oh thats miniature ;) [23:22] right [23:22] remote controled cars [23:22] BUGabundo: thats not a log with my patch [23:22] some of those go as fast as 80KM/H [23:23] yeah ... but wouldnt be a sport for me ;) [23:23] -rw-r--r-- 1 bugabundo bugabundo 5.2K 2009-08-15 01:00 /home/bugabundo/Desktop/mm2.log.txt [23:23] it must be [23:23] its the one I ran last night before bad [23:28] asac: is it not correct? [23:28] doesnt feel correct [23:28] BUGabundo: did you capture all output on console? [23:28] no [23:28] that's the log from MM [23:29]