=== rickspencer3 is now known as rickspencer3-afk [00:48] asac [00:48] you there? [11:47] asac: Hello [11:47] hi [11:47] quite busy so dont run away ;) [11:47] wait ... :-P [11:47] ok [11:48] asac: ping me when you're free please [11:51] asac: chmsee 1.0.6 can't be merged from Debian because it must be build against xulrunner 1.9.0 (not .1), and the change you done to chmsee is build-dep on xulrunner-1.9.1 [11:58] AnAnt: why does that mean it cant be merged? most likely the change i did just needs to be applied similarly [11:58] asac: because chmsee cannot compile against xulrunner-1.9.1 [11:59] AnAnt: but it worked for the package i uploaded. the merge should rebase and adjust whatever changes i did so it builds against 1.9.1. makes sense? [11:59] asac: I tried that here, and it didn't work, I got compile error, some sort of configure script failed saying that it cannot find libxul-whatever <= 1.9.0.99 [12:00] and I just asked upstream developer (lidaobing), and he confirmed that chmsee 1.0.6 cannot be compiled against xulrunner 1.9.1 [12:01] AnAnt: then the patch (or wahtever i did needs to be properly adjusted/merged) [12:01] ok [12:02] well. previous version couldnt be compiled too i guess [12:02] hang on, it seems I can patch chmsee 1.0.6 can be patched [12:02] merging involves ensuring that previous patches are not lost [12:02] ok, I will work on it with upstream [12:02] so dont just drop my patch [12:02] asac: it seems that I won't have to ! === Jazzva__ is now known as Jazzva [12:03] doh, how do I use quilt in a cdbs package ? [12:04] man quilt [12:05] asac: could you review nspluginwrapper 1.3.0 :)? I've been using it for about a week or so, and it's OK (meaning it's not crashing anymore than 1.2.2). [12:05] yes [12:06] we can upload that i think do we know what changes were done there? [12:06] do we have latest svn or release? [12:06] maybe latest svn is better ? [12:06] asac: we have latest 1.3.0 (released as tar.gz). I don't know the state of latest svn [12:06] AnAnt: what i mean is: there is nothing special about quilt wrt packaging or not packaging. the cdbs-edit-patch thing is not use. you just use quilt directly [12:10] asac: I mean what should I include in rules file [12:12] AnAnt: dpkg -L quilt | grep cdbs.*mk [12:12] yeah found it, thanks [12:13] asac: do you want me to update to the latest svn_ [12:13] ? [12:14] Jazzva: no. just review what has happened in between for now would be great [12:15] morning asac :) [12:15] hi [12:23] asac: some improvements, afaics. last commit was on 17. january. [12:38] ok [12:39] lets get 1.3.0 up then and review if the patches are worth taking [12:39] 3 patches are already submitted upstream, but no comment from upstream yet [12:40] asac: this should be the link to the branch http://code.launchpad.net/~jazzva/nspluginwrapper/1.3.0 [12:43] asac: ok, would you sponsor this: LP 416346 [12:44] Launchpad bug 416346 in chmsee "Candidate release chmsee 1.0.6-1ubuntu1" [Undecided,New] https://launchpad.net/bugs/416346 [12:56] i will check this later [12:56] * asac on a MIR rush [13:03] AnAnt: did you get the xsplash bug mail? [13:11] mac_v: oh, you're here, yes, thanks [13:12] mac_v: didn't check the link yet === jdstrand_ is now known as jdstrand [13:12] :) [15:06] seems someone sponsored chmsee already [15:08] Jazzva: you didnt base on current lp:nspluginwrapper? [15:23] asac: Hmm, I did, but I uncommitted changes related to 0ubuntu6. [15:23] those are revisions 54-56 [15:24] and then applied diff from the archive. [15:24] brb, need to go outside and buy cigarettes. [15:35] Hello asac around ? [15:36] Hve you tested this...? http://blogs.gnome.org/dcbw/2009/07/10/unwire-with-networkmanager/ [15:37] have [15:38] Also tested with gnome-bluetooth from SVN but with no succes [15:40] not yet tested. how does it fail? NM not seeing it? [15:42] Hello asac, it just dont give me this option...http://ubuntu-pics.de/bild/22501/3_bt_pan_1kQDA9.png [15:45] asac: I'm back, if you need any more info. [16:03] Jazzva: on a call. ~30min [16:21] and me... :)... lost in the jungle ;) [17:18] asac ??? [17:43] sorry ... was on a call ... will be back in 2 hours ... far tooo hot to stay inside any minute longer [17:43] plun: Jazzva: ^^ [17:45] asac: no prob. I'm studying (taking a pause atm), so I might be a bit unresponsive later. Ping me when you come back, and I will see the message. Have fun... [19:50] Mook_sb, http://launchpadlibrarian.net/30582954/buildlog_ubuntu-jaunty-amd64.songbird_1.4.0~a~svn20090820r14593-0ubuntu1~usd1~jaunty_FAILEDTOBUILD.txt.gz gasp, did something change? [19:51] fta: hmm, we moved to bz2 instead of gz for xulrunner bits [19:54] looking at my rules.. [19:59] Mook_sb, i used your make-xulrunner-tarball.sh [19:59] -d [19:59] hmm [20:00] sh ./make-xulrunner-tarball.sh $(XUL_SRC_DIR)/compiled/xulrunner-$(XUL_RELEASE)/dist/bin $(SRC_DIR)/dependencies/linux-$(MACHINE)/xulrunner/$(XUL_RELEASE) xulrunner.tar.gz [20:01] should I just change the last arg? [20:01] it'a using $TAR xzvhf, so... that doesn't sound quite right; poking our build guy at the moment [20:03] fta: :( he's using http://publicsvn.songbirdnest.com/vendor/trunk/xulrunner/make-xulrunner-tarball.sh instead, which uses -cjvh [20:03] (but yes, change the last arg too) [20:03] ok [20:04] sorry about that :( [20:26] asac: when do you plan to release -devscripts 0.15? [20:32] bdrung_: let me think ;) ... hmm [20:32] there was something we still needed? [20:32] anything you would like to land first? [20:33] asac: i want to implement the xpi-build target (as alternative to MOZ_XPI_BUILD_COMMAND) [20:33] asac: if you want to release soon, i can implement it today [20:34] asac: is there any progress i having a policy for debian version names? [20:34] bdrung_: can we review if something prevents us from using it outside of cdbs? [20:34] from what i see we have xpi-install and xpi-clean [20:34] which you can just call [20:34] if you use debhelper etc. [20:35] give that those xpi- targets appear to become top level rules ... i wonder if we should name it a bit different [20:35] e.g. no xpi-build (which would imply you should call that on your own?) [20:35] or maybe implement xpi-build and provide a hook rule [20:35] like xpi-build-impl/PACKAGE ? [20:38] asac: short away [20:48] applet discussion: http://pastebin.ubuntu.com/256499/ [20:48] (just for me) [20:59] Mook_sb, hm, nope, not good enough. http://launchpadlibrarian.net/30585187/buildlog_ubuntu-karmic-amd64.songbird_1.4.0~a~svn20090820r14593-0ubuntu1~usd2_FAILEDTOBUILD.txt.gz [20:59] tar -j -x -p -f /build/buildd/songbird-1.4.0~a~svn20090820r14593/build-tree/songbird/dependencies/linux-x86_64/xulrunner/release/xulrunner.tar.bz2 -C /build/buildd/songbird-1.4.0~a~svn20090820r14593/build-tree/songbird/compiled/dist/xulrunner [20:59] bzip2: (stdin) is not a bzip2 file. [20:59] tar: Child returned status 2 [20:59] so something more is needed [21:16] asac: back to -devscripts. [21:16] asac: what do you mean with "can we review if something prevents us from using it outside of cdbs?" [21:21] fta: sorry, was lunching/meeting; I don't understand that bzip2 error, but line 40 of the new make-xulrunner-tarball.sh should be echoing the command line, so maybe it's ending up using the old one? [21:32] night ! o/ [21:38] Mook_sb, hmm [21:38] ./tools/scripts/make-xulrunner-tarball.sh [21:38] ./dependencies/vendor/xulrunner/make-xulrunner-tarball.sh [21:38] yes, should be using the vendor/ version [21:39] http://paste.ubuntu.com/256521/ [21:39] the other should be dropped then [21:39] +one [21:40] yes, it should [21:47] okay, the stale scripts removed upstream. [21:48] thanks [22:02] asac, do you know which tool/site uses our customized user agent? [22:17] fta: what do you mean? [22:17] bdrung_: i mean that folks tend to think that mozilla-devscripts is only for cdbs. so i want to be sure it works well with debhelper and other approaches ;) [22:18] asac, i remember that at some point, we stopped mentioning ubuntu in the UA and it broke some stats, like ubuntu loosing popularity or something [22:18] asac: ah, ok. i can test it if it would work with simple debhelper [22:18] fta: i think all http logs have that user agent [22:19] and some do analysis [22:19] netcraft is probably one of them [22:19] fta: if you ask if it would make sense to do something similar for chromium, then i would say yes [22:19] hmm.. there was a bug and a lot of attention [22:19] just do it in the same syntax that firefox uses [22:20] so websites that really need to parse it can at least learn [22:20] for me the problems incurred are worth the win [22:20] but thats a subjecting pov [22:20] asac: https://bugs.launchpad.net/ubuntu/+source/pwdhash/+bug/416651 [22:20] Ubuntu bug 416651 in pwdhash "Please merge pwdhash 1.7-3 from Debian unstable (main)." [Wishlist,Confirmed] [22:20] for the distro having the distro stuff in there is definitly beneficial [22:21] fta: there is the shiretoko bug [22:21] but at least some of those sites actually hate the word Linux ;) [22:21] you cannot help those [22:21] asac, well, chromium won't do it apparently [22:21] bug 246775? [22:21] Launchpad bug 246775 in firefox "user string agent doesn't identify ubuntu" [High,Fix released] https://launchpad.net/bugs/246775 [22:22] asac: you could say "good os" instead of "Linux" :) [22:22] fta: that was the bug when we dropped it (accidentially actually) [22:22] i moved stuff to or from ubufox and forgot to commit one half iirc [22:22] asac: 3G still broken :( [22:22] BUGabundo: definitly not in the same way. so you need to be more specific [22:23] like --debug logs etc [22:23] oh ok [22:23] on next boot and upgrade [22:23] I have to downgrade to old version for it to work [22:23] wait till all three are there: nm + applet + mm [22:23] hmm [22:23] i need the log then [22:23] the bug you filed is fixed ... pretty sure [22:24] wait for the next batch of logs :) [22:24] at least the behabior is *exaclty* the same [22:29] i hav the feeling you are stuck with something old you built [22:29] on your own ;) [22:29] but well. file a bug and we can check that in depth [22:29] have to go now .. and play a game [22:30] eheh [22:30] ok [22:30] go play [22:30] and let us suffer :) [22:30] asac, http://code.google.com/p/chromium/issues/detail?id=17095 [22:32] fta: what is the full user agent atm? [22:33] seems they explicitly want patches in downstreams [22:33] Mozilla/5.0 (X11; U; Linux i686; en-US) AppleWebKit/532.0 (KHTML, like Gecko) Chrome/4.0.203.0 Safari/532.0 [22:35] thx ... looks like the epiphany webkit string [22:36] they land a fix now ... not sure it goes into webkit directly though [22:36] k bbl [22:38] heyyy that looks like my android agent!!!!!! [23:01] asac: my idea of having a xpi-build target is not needed. the build/$(MOZ_EXTENSION_PKG) target could be used for long build commands. e.g. use http://paste.ubuntu.com/256553/ instead of http://paste.ubuntu.com/256554/ [23:01] that should be documented === rickspencer3 is now known as rickspencer3-afk [23:17] asac: i tried to combine xpi.mk with dh 7, but it failes. here's an example rule for adblock-plus: http://paste.ubuntu.com/256563/ [23:18] asac: if someone can hack in perl, following simple rule would be cool: [23:18] %: [23:18] \tdh --with xpi $@ [23:34] bdrung_: just override_dh_auto_install: [23:35] bdrung_: just override_dh_auto_install: xpi-install [23:35] doesn t work either? [23:35] why do you think that override_dh_auto_install: is the right hook? [23:35] asac: doesn't work either [23:36] asac: why not override_dh_auto_install? [23:37] asac: xpi-install is equivalent to "make install", isn't it? [23:45] bdrung_: i had no idea about dh 7 ... but looked at code. either i miss something or its not extensiable at all [23:45] it fails because it doesnt know about xpi-install target and because its run with catch-all rule %: [23:45] it thinks it must be a sequence [23:45] so we need to add a sequence somehow ;) [23:46] didnt see how thats doable from looking at /usr/bin/dh ;) ... fail [23:46] man dh [23:46] %: is only used if there is no other rule [23:46] asac: whats a "sequence"? [23:47] well. look at the code [23:47] i get "xpi-install is not a sequence" [23:47] a sequence is a sequence of commands [23:47] in dh code its basically the full list of dh_ commands [23:47] in some order [23:47] sequence = command for the rule [23:47] and then done in a way that you can override them [23:48] asac: have you seen bug 416627 [23:48] Launchpad bug 416627 in firefox-3.5 "Firefox 3.5.3pre: Restart launches /usr/bin/firefox regardless of the original launch command." [Undecided,New] https://launchpad.net/bugs/416627 [23:48] well. the error comes from dh [23:48] so the catch all rule is called ... dont ask me why [23:48] micahg: yes i am aware of that bug [23:48] its tricky thing [23:49] ok, does have to do with making it the default? [23:49] the patch for that should probably be adapted for non-default builds [23:49] asac: make[2]: *** Keine Regel, um »xpi-install« zu erstellen. Schluss. [23:49] we have a patch that restarts using that path so you can restart firefox even though the firefox version/directory change [23:49] d [23:49] dh xpi-install [23:49] dh: Unknown sequence xpi-install (choose from: binary binary-arch binary-indep build clean install) [23:50] yes, but shouldn't it use the original command like /usr/bin/firefox-3.5? Shouldn't the symlink just change? [23:50] asac: because xpi-install is missing, dh xpi-install is run. so why is xpi-install missing? [23:52] asac: Medium -> Triaged -> Assign to you? [23:52] or is this something I might be able to fix in the patch? :) [23:56] 00:49 < bdrung_> asac: make[2]: *** Keine Regel, um »xpi-install« zu erstellen. Schluss. [23:57] 00:49 < asac> we have a patch that restarts using that path so you can restart firefox even though the firefox version/directory change [23:57] 00:49 < asac> d [23:57] 00:49 < asac> dh xpi-install [23:57] 00:49 < asac> dh: Unknown sequence xpi-install (choose from: binary binary-arch binary-indep build clean install) [23:57] 00:50 < asac> thats with MAKE .... [23:57] 00:50 < asac> your make is probably wrong [23:57] 00:50 < asac> you ened to do make -f $(CURDIR)/debian/rules xpi-install [23:57] 00:52 < asac> or use ...: xpi-install [23:57] 00:52 < asac> that has the same error afaict [23:57] 00:53 < asac> so solution is to make -f /usr/share/.../xpi.mk xpi-install ;) [23:58] (05:52:05 PM) micahg: asac: Medium -> Triaged -> Assign to you? [23:58] (05:52:48 PM) micahg: or is this something I might be able to fix in the patch? :) [23:58] micahg: the problem is that we have a script [23:59] so we cannot create a link to the right binary and then $0 would be correct [23:59] you can't see where the original call was from? [23:59] no [23:59] but i have an idea [23:59] ok, then can we just set it to launch based on version? [23:59] creating a /usr/bin/firefox-real -> /usr/lib/firefox-VERSION/firefox [23:59] and runing that [23:59] from the firefox script