=== asac_ is now known as asac [09:44] morning [11:37] asac, did you get the email to vote for the udm board? [11:37] hi [11:37] udm? [11:37] fta2: ? [11:37] hey guys [11:38] hi BUGabundo_work [11:38] asac, https://wiki.ubuntu.com/DeveloperMembershipBoard [11:39] asac, wrt https://lists.ubuntu.com/archives/ubuntu-devel-announce/2010-January/000661.html [11:40] i can't locate the email [11:42] i got a mail for the vote afaik [11:42] mybe ping dholbach to get another mail? [11:42] what's the subject ? or the from? [11:48] fta2: call for votes: Ubuntu Developer Membership Board election? /me just got a reminder from mdz. Ask him for a new mail [11:53] when was that? maybe it's in my spam folders, but it's crowded in there, i need hints [11:58] haha [11:58] fta2: how about greping for the URL? [12:00] fta2: http://pastebin.com/f2dd84321 [12:01] lets go for that [12:01] on top i want to add [12:01] a) debian/copyright.dep5-head [12:01] (thats the head of the generated copyright file) [12:01] debian/remove-dfsg.txt [12:01] -> thats src/third_party/gles_book_examples/ atm [12:02] Subject: CIVS Poll now available for voting: Ubuntu Developer Membership Board [12:02] Date: 12/24/2009 12:07:25 PM (Thu, 24 Dec 2009 06:07:25 -0500) [12:02] looks like it [12:02] http://pastebin.com/f390770a3 [12:02] err, cat debian/copyright.whitelists | pastebinit [12:02] http://pastebin.com/f390770a3 [12:02] thats the list i am currently working on to get everything that is unknown done there [12:02] http://people.canonical.com/~asac/tmp/chromium-copyright/1/ [12:02] there are the copyright files generated atm [12:03] .1 is normal source ... .2 is third_party only [12:04] ok if i commit the current changes to licensecheck for now? or wnt to preserve the old format? [12:04] from there i would want to also add the whitelist feature at least [12:04] and a wrapper that appends full icense text for all whitelisted ones like: [12:04] License: NAME [12:04] license text .... [12:04] e.g. [12:04] License: MPL-1.1 [12:05] ... [12:05] feel free to commit, my format was not meant to be compliant with anything [12:05] kk [12:05] lets move switflty then [12:05] i hope we can upload this week!!! [12:05] asac: frohes neues jahr! :) [12:05] hi sebner [12:06] fta2: are there hooks to strip stuff on orig generation? [12:06] like the book examples etc.? === BUGabundo_work is now known as BUGabundo_lunch [12:11] asac, i have hooks to remove stuff from CO, that's the best when it's a distinct svn repo (see the main DEPS file) and another hook to simply drop a bunch of directory (but it's still fetched and lives in the cache) [12:12] also hooks to remove some file types [12:12] fta2: where are those hooks? [12:12] fta2: http://people.canonical.com/~asac/tmp/chromium-copyright/1/copyright.dep5.2 [12:12] chromium/third_party/WebKit/LayoutTests/ -> thats the default WebKit license? [12:13] didn't i drop all the LayoutTests? [12:14] could be [12:14] i have to use your source [12:14] i guess [12:14] ;) [12:14] yes [12:14] it's best to start from a fresh tarball, it's way smaller than the full thing [12:15] meaning less license issues :P [12:16] hmm ok [12:17] less files, less licenses to review [12:17] fta2: do you think you could implement the whitelisting parsing/matching? idea would be to add the path/* with the license mentioned in the whitelist and then skip all the sub-directory content that has no UNKNOWN in it [12:19] asac: do you also hate crappy upstream authors with crappy/no license headers in their files? *grgrgrgr* [12:19] asac, you mean another whitelist? i already have one [12:19] hate is the wrong word ;) [12:20] fta2: committed debian/copyright.overlay [12:20] so lets not call it whitelist [12:20] its overlay [12:21] asac: kill is a too hard word :P [12:21] fta2: basically all the icu/ subtree is ICU-license [12:21] so in the copyright output we can ignore all UNKNOWN and just replace them with ICU-license [12:22] asac, could be done with the script already [12:22] asac, did you try $manually_identified ? [12:23] fta2: what does manually_identified do? [12:23] ah cool [12:23] fta2: does that inject the license if unknown? or for all? [12:24] ok seems manuall_identified always needs copyright [12:24] which isnt true ...we just want to add the license [12:26] it's used after licencecheck, but before we dig further down with get_license [12:26] hmm [12:26] seems we first try to parse licensecheck [12:26] then we just overwrite it with manual [12:26] imo that should just "fill up" unknown ones [12:27] or we need a "manual_fillup" list [12:29] http://pastebin.com/f4af8d64c [12:29] fta2: something like that ^^ [12:30] not sure if we should have the manual block as its now ... and then have an additional "fillup" one after all other options have been explored if its still unknown [12:31] asac: ham ham ham, automatic copyright recognition? NICE [12:31] http://pastebin.com/f48894ef6 [12:31] fta2: ^ [12:31] yep ;) [12:32] asac: push push push :P [12:32] asac: can you review this please? https://code.edge.launchpad.net/~dmitrij.ledkov/xulrunner/1.9.1.head-dh_xulrunner/+merge/16649 [12:32] i hope i could ;) [12:32] * asac checks [12:44] micahg: looks good [12:44] asac: ok, were my comments appropriate? [12:44] micahg: was ther some requirement? [12:44] e.g. on debhelper version etc.? [12:44] e.g. will this just work in hardy too? [12:44] he said not, I haven't looked [12:45] that's what I was wondering [12:45] micahg: yes. thats all fine [12:45] thanks [12:46] * micahg is checking syntax now [12:47] asac: dh --with xulrunner will only work with karmic and up [12:47] idk how the scripts use it now [12:47] yes. but you can always just run dh_xulrunner manually i guess [12:48] ? [12:48] yeah [12:52] asac: ok, so do I approve now and push to the other 2 branches? [12:52] yes [12:52] thanks [12:53] micahg: can you add the xulrunner fix for thumb2 too to xul 1.9.1 branch too? [12:53] one sec [12:53] where is it now? [12:54] https://bugs.edge.launchpad.net/ubuntu/+source/xulrunner-1.9.1/+bug/488354 [12:54] Ubuntu bug 488354 in firefox "NS_InvokeByIndex in xptcinvoke_arm.cpp is not Thumb-2 safe for Lucid" [Unknown,Confirmed] [12:57] asac: can I merge through the UI? [12:57] * micahg has never done this before? === BUGabundo_lunch is now known as BUGabundo_work [13:01] micahg: no. afaik you have to do that locally [13:01] asac: k, then I set to merged? [13:01] micahg: that automatically happens if you push afaict [13:01] maybe double check [13:01] ah, ok [13:02] asac: do you want me to pull the thumb patch from mozilla or arm? [13:02] from arm [13:02] the mozilla patch is bogus afaik [13:02] we should update it [13:04] asac: should I prepare a lucid changelog entry for 1.9.1.6? [13:05] yea [13:05] is that today= [13:05] ? [13:06] no, 3.5.7 is soon, but I figured you could push this to Lucid so the dh_xul merges could happen [13:08] asac: actually 3.5.7 is tomorrow [13:08] yes [13:08] so, should I just prepare it as 3.5.7 [13:09] hmm. lets upload 1.9.1.6 today to lucid if we have those patches [13:09] k [13:19] asac: do I need to list all the files added in teh changelog? [13:22] also, is this a decent patch name: fix_bz532198_lp488354_not_thumb2_safe.patch [13:24] actually, changed patch name to bz532198_lp488354_ns_invokebyindex_not_thumb2_safe.patch === cyphermox_ is now known as cyphermox [13:28] asac: how does this work, I dropped a patch for 1.9.1.7, but I still need it for 1.9.1.6? [13:30] hola everybody! [13:30] micahg: i would fork a branch from before the patch was dropped [13:30] then cherry-pick the commits i want from after that [13:30] and make a release [13:30] then merge that into head again [13:31] will TB3 go into lucid? [13:31] dholbach: yes :) [13:31] micahg: rock [13:31] dholbach: soon, I just have to add the bugs targeted and make sure it builds [13:32] dholbach: frohes neues! :) [13:32] sebner: and the same to you [13:32] thanks a bunch micahg [13:32] :) [13:32] dholbach: if lazy asac only would upload it :P [13:32] leave poor asac alone! :) [13:34] * sebner thinks poor asac makes to many holidays :P [13:48] asac: I can't figure all this out right now, if it can wait till tonight, I can do it [13:48] eekk [13:48] forgot to lookup my TB3 bug :( [13:49] micahg: wanna try to reproduce it ? [13:49] BUGabundo_work: not right now, maybe later [13:52] ok [13:52] if u need an GApps account ping me [13:52] i'll create one [13:53] micahg: sure [13:53] asac: sure it can wait? [13:53] micahg: can you add the patch to current .head? maybe i will do a release today if i get to it (though unlikely) [13:54] asac: for arm? [13:54] for 1.9.1.7 or 1.9.1.6? [13:56] I'll be back in about 45 minutes === jtv1 is now known as jtv [14:03] micahg: for arm yes. [14:03] to .head [14:15] fta2: ping [14:19] fta2: I'm having a problem using ppabot to build mplayer, the version has mplayer_2:1.0~rc3... but it seems that bzr-builddeb is trying to find mplayer_1.0~rc3... instead. It correctly created an upstream tarball name mplayer_2:1.0...orig.tar.gz, but bzr is looking for the wrong file. === ripps is now known as ripps|sleep === \vish is now known as mac_v === mac_v is now known as \vish [16:43] asac: whats our next step to get major version upgrades done? looks like something about embedding reverse dependencies? [16:46] asac: probably can't do arm patch for another few hours [16:49] kk [17:02] ripps|sleep, hm, on the top of my head, i don't know, i never tried a version with an epoch. i need to dig into the code, but not now, i'm busy /w work === yofel_ is now known as yofel [18:39] ccheney: the firefox part is properly dealt with. the main work is required for xulrunner rdepends. [18:39] ccheney: one thing is to get webkit backported and renamed [18:42] ccheney: so try to get karmic webkit built in hardy ... then if that builds [18:42] rename the soname to be non-conflicting to what we have in hardy [18:43] e.g. we want to upload it as libwebkit-karmic...so or something [18:43] also the package name needs to be adjusted [18:43] same has to happen for all the newer stuff needed for the epiphany browser from karmic [18:43] stage all that in a ppa ... until you can build epiphany from karmic against all those renamed libs [18:43] ccheney: understood? [18:47] ok [18:48] !time [18:48] Information about using and setting your computer's clock on Ubuntu can be found at https://help.ubuntu.com/community/UbuntuTime - See https://help.ubuntu.com/9.04/serverguide/C/NTP.html for information on usage of the Network Time Protocol (NTP) [18:48] webkit for hardy due to needing it for epiphany, right? [18:49] ah yea i think i see now :) [18:49] the other packages that actually need xulrunner we just backport xulrunner for them later? === Nafallo_ is now known as Nafallo [19:16] * ccheney wonders if epiphany actually needs all those b-d, lots are missing entirely for hardy not even old versions of them [19:24] asac: new epiphany needs new glib,gtk,webkit,etc is that ok to upgrade all of that? [19:25] ccheney: why new glib? [19:26] everythig not glib and gtk sounds definitly expected [19:27] in any case. its not about upgrading, but about making a new package with a different name that we can roll out in parallel for those libs [19:28] its in the source dependencies (not just part of debian packaging) [19:28] ok [19:28] glib-2.0 >= 2.19.7, gio-unix-2.0 >= 2.19.7, gtk+-2.0 >= 2.16.0, gtk+-unix-print-2.0 >= 2.16.0 [19:28] hmm. the debian package says 2.18 [19:28] i dropped the introspection packages from b-d and it seemed to not complain about them missing [19:29] and 2.16 for gtk [19:29] what is in hardy? [19:29] 2.12 [19:29] yeah. introspectoin is probably ok to drop [19:29] the debian control file may be buggy [19:29] probably [19:29] the source is definitely looking for 2.19/2.20 in any case [19:29] so you could check what happens if you relax the depends [19:29] in configure [19:29] and just build [19:30] ok will modify configure also :) [19:30] if it build fails ... then its obvious ... otherwise we should check with epiphany upstream what exactly they require [19:30] i removed all the b-d in control but wasn't sure what would happen if i did it to configure [19:30] ok [19:30] ccheney: maybe start with webkit [19:30] that alone most likely has some requirements we might not be able to fulfill [19:30] or did you already check that? [19:30] looks like i need to backport webkit and soup and relax the configure checks to get it to try to build [19:31] it needs libsoup-gnome-2.4 which appears to not be in hardy as well [19:31] libsoup2.4 itself is though, i guess libsoup-gnome must be some sort of bindings or something [19:34] ccheney: maybe libsoup-gnome-2.4 is just a forked package ... or it was added in later versions [19:34] or a .pc file split up [19:35] (if we are lucky) [19:35] otherwise we have to check again ;) [19:35] yea looks like it was added to libsoup2.4 in later versions [19:36] * ccheney will backport both then see what happens with the rest of the packages [19:36] eek soup was glib 2.21 [19:37] er /was/wants [19:37] * ccheney hopes this doesn't spiral too much [19:39] ccheney: libsoup was added? [19:39] err i mean: .pc file was added? [19:39] sure there are also headers added and not just a split up? [19:43] looking into more detail now, there are definitely separate pc and libraries now, not sure if it was split from the previous libsoup library though [19:45] different headers too :\ [19:49] wow backporting gnome related things is pita :) [19:49] new libsoup2.4 wants libproxy since apparently its needed for new gnome stuff as well [19:53] hmm [19:53] we should probably check if all that are new features [19:53] and maybe do a special build without those :( [19:55] *sigh* ... thats what i mean ;) [19:55] all gnome should stop this "i am a library and its great to require all new libs too" [19:56] asac: probably not going to be able to do arm patch till after work [19:57] i ripped out proxy from configure check [19:57] but it failed to build due to apparently new glib names: G_SOCKET_FAMILY_IPV6 [19:58] which was added in 2.22 [19:58] ccheney: yeah. you can fix that by backporting just that define. G_SOCKET_FAMILY_IPV6 = GLIB_SYSDEF_AF_INET6 [19:59] or just put it in the soupo .c file that needs that [19:59] ok [19:59] /usr/include/glib-2.0/gio/gioenums.h: G_SOCKET_FAMILY_IPV4 = GLIB_SYSDEF_AF_INET, [19:59] /usr/include/glib-2.0/gio/gioenums.h: G_SOCKET_FAMILY_IPV6 = GLIB_SYSDEF_AF_INET6 [20:08] ccheney: was that everything needed so far? === mac_v is now known as \vish [20:27] it then needed the definitions for GLIB_SYSDEF_AF_INET6 as well [20:28] new glib seems to mostly build on hardy as is, but it failed in a weird manner [20:30] debuild: fatal error at line 1247: [20:30] debian/rules build failed [20:30] not sure if that is some sort of cdbs issue or what as there aren't that many lines in the rules file [20:30] glib doesnt have those GLIB_SYSDEF_AF_INET6 ? [20:30] hmm [20:32] hmm actually glib doesn't use cdbs but its not got that many lines either, very strange [20:32] asac: it seems to define it somehow in the new version but not sure how to backport it, it seems to be generated in configure [20:34] and its not defined in the hardy version, no [20:34] let me check what you say ;) [20:36] seems it represents a magic number 23 on windows and maybe 10 on linux [20:36] er for INET6 [20:36] hmm [20:36] so maybe i can just use the 2 and 10 for INET and INET6 in my patch [20:36] no gioenums.h? [20:36] in hardy version? [20:36] was that a separate package maybe? [20:37] ccheney: ? [20:37] gioenumtypes.h [20:37] but not for the inet stuff [20:37] iirc gio was first added around hardy timeframe [20:37] so probably wasn't fully done yet [20:37] so #define GLIB_SYSDEF_AF_INET AF_INET [20:37] so #define GLIB_SYSDEF_AF_INET6 AF_INET6 [20:37] ok [20:38] so #define GLIB_SYSDEF_AF_UNIX AF_UNIX [20:38] i would guess [20:39] yea [20:39] i pulled the full enum [20:40] ccheney: so thats unblocked? [20:43] trying to find the proper place to stick it now [20:43] everything seems to be trying to use it, heh [20:44] i guess soup-portability is a good location and then stick that header include in the main set of includes so it is defined everywhere it needs to be [20:47] more failures now that i fixed that part :( [20:47] needs GInetAddress [20:47] which is in 2.22 also [20:48] a bunch of functions in that section of glib [20:50] * ccheney attempts to build glib2.0 again to see if it will work this time [20:51] * ccheney tries to remove parallel option to see if it helps [20:52] it should build according to its build-depends but fails :( [20:52] ccheney: i really dont think we wwant to build glib2.0 newer [20:52] we should try to avoid that [20:53] it seems to not link [20:53] /root/glib2.0/glib2.0-2.22.3/glib/gtestutils.c:1307: undefined reference to `__abort_msg' [20:53] due to that [20:53] so glib needs libsoup? [20:53] well we need to backport the whole inet section from glib then or somehow get around needing it [20:53] soup needs glib [20:53] ok ... so why do we want a new glib? [20:53] soup needs the new inet stuff from glib, more than just the defines [20:53] i really hope we can avoid gtk and glib [20:54] and epiphany needs soup-gnome [20:54] functions too? [20:54] hmm [20:54] yeah [20:54] yea the functions [20:54] what functions are those? [20:54] * ccheney wonders if there is a soup-gnome from before glib 2.22 [20:55] i dont think so. i remember soup being bad API wise at some point [20:55] for gtk apps etc. [20:55] soup-gnome got added in jaunty [20:55] ok [20:56] http://pastebin.ubuntu.com/351426/ [20:57] that is the current error list for soup [20:58] so what glib do we have in hardy? [20:58] glib 2.16.6 gtk 2.12.9 [21:01] if i reply to a message in gmail, can i make the reply appear in TB 3? [21:01] i know the other way is possible [21:01] mbana: use imap [21:02] that's so vague. and i'm ready using imap i think. google doesn't support pop, iRC. [21:03] it supports pop afaik [21:04] yep supported pop first then later added imap support [21:04] ccheney: so what do we need? [21:04] gresolver.h [21:04] ginetaddress.h [21:04] ginetsocketaddress.h [21:04] gunixresolver.h [21:04] gsrvtarget.h [21:05] hmm. what is that? [21:09] looks like dns service records for local services on the domain [21:10] hmm [21:10] i'm using imap asac [21:10] yeah [21:10] just checked [21:10] ccheney: ok. so those are enough for soup gnome? ;) [21:10] e.g. do they actually build ;)? [21:11] well to get past that part of the failure, how should we go about adding all those though? [21:11] copy those into soup and adjust the build procedure? [21:22] ccheney: i think copying all as static funcs or something could work [21:22] depends where they are used in soup [21:23] so far looks like they isolate it to that one source file [21:23] so that might work out [21:23] ccheney: to one? thought we need at least three .h ... [21:23] from the build failure you pasted [21:24] ripps|sleep, indeed, my bot doesn't support epochs, yet. 1/ your get-orig-source should not generate the tarball with the epoch - that epoch is only in d/changelog, 2/ i should teach both update-pkg & sync-ppa how to react to an epoch. [21:24] ah you mean isolate consumption [21:25] ripps|sleep, i just wonder if i should add a parameter in the per-package conf file to specify that epoch, or just re-use the one found in d/changelog [21:26] yea [21:27] ok then try that. maybe pull in func by func until you have enough to build it ;) [21:50] asac: around? [21:51] asac: can you give me a python snipped for determing the emid of an install.rdf file using rdflib? [21:54] bdrung: where did that stuff end up? [21:54] i can see rdflib anywhere in moz-devscripts ;) [21:55] asac: in a install-xpi python script (in m-d) [21:55] last commit is 281 [21:55] ever pushed that? [21:55] asac: not yet committed [21:55] asac: look at the dh_xul-ext branch (will land there) [21:55] but its uploaded? [21:56] not yet [21:56] hmm [21:56] thought you already uploaded it [21:56] i will give you a diff [21:56] asac: i will get rid of xpath first and write the man pages [21:57] * asac checks out dh-xulext branch [21:57] asac: and apply this patch: http://paste2.org/p/594222 [22:01] bfiller: http://paste.ubuntu.com/351454/ [22:01] try that [22:01] if not i hvae to try and error here [22:01] bdrung: ^^ [22:01] bfiller: unping [22:05] asac: thanks. it works. [22:06] asac: can you have a look at the other TODO too? [22:14] hehe commenting out the proxy in configure didn't make it go away :) just made it fail later [22:16] is libproxy the proxy configuration soft thing that uses javascript? [22:17] bdrung: dont see any TODO in that file [22:17] maybe push what you got ;) [22:17] not sure [22:17] asac: paste line 55 [22:18] " libproxy is a lightweight library which makes it easy to develop applications proxy-aware with a simple and stable API. [22:18] asac: i have to check it before committing [22:18] it appears to b-d on webkit too (wtf) [22:18] bdrung: simplify this? [22:18] thats what i did initially [22:18] thats ok [22:18] * ccheney wonders why a proxy needs a web browser [22:19] asac: you gave me a query for the EMID (line 107), not for the target apps [22:21] bdrung: sorry i am dense today ;) [22:21] in line 56 there is a query. thats not enough? [22:23] asac: in line 56 we need only the target apps and not the min and max version [22:23] how high is your alcohol level? [22:25] zero [22:25] you should be more verbose [22:26] just first line of the query doesnt work? [22:27] asac: first line = "SELECT ?id ?max ?min" or what? [22:27] err ... that query works [22:27] you an ignore max/min [22:27] its optional output [22:27] can [22:28] grr webkit also needs glib 2.21.3+ [22:29]