[07:37] ricotz, do you remember what cause the closure leaks in gjs tests? [07:46] darkxst, it was triggered by js port of mainloop which relied on a glib patch [07:47] http://git.gnome.org/browse/glib/commit/?id=1ce415b45bde895c31dd32e2f7ab7e7ab79e735e [07:49] hmm right, so this one I am seeing is something else then [07:50] its leaking a closure in testGObjectClass [07:51] so i guess you changed something [10:35] ricotz, no i didnt change anything, rebased patches for js17 etc, but that is all just build scripts [10:35] I suspect its a deeper issue in g-i or glib [10:37] https://github.com/darkxst/js17/tree/mozjs17-rebase [10:56] darkxst, this looks new thoug http://git.gnome.org/browse/gjs/commit/?h=wip/mozjs-188&id=ab258cc09201308aba8f3c1a956555f6f4bcc5f9 [10:57] ricotz, no, walters missed that patch when he created the wip branch [10:58] that was/is one of the original patches for my js188 port [10:58] ah ok, i didnt compared it with the patch-set [10:59] I have not touched gjs in weeks [11:00] mostly been prodding mozilla people, trying to get a release [11:00] so you are seeing this issue building the current branch against "my moz188" snapshot too? [11:01] i will just build gjs now [11:01] its more broken right now due to the new introspection checks Jasper added [11:02] in git g-i [11:02] right, just noticed [11:03] I filed a bug, but bgo is down right now [11:06] I even commented out all the closures in the test case, and that didnt help either [11:08] ricotz, anyway if you revert back to the released g-i then you will probably see it [18:10] ricotz: I filed bug 1123302 [18:10] bug 1123302 in modemmanager (Ubuntu) "Please package modemmanager >= 0.7.990" [Wishlist,New] https://launchpad.net/bugs/1123302 [18:13] jbicha, ah, i see, cyphermox has it probably in his queue already [18:14] i guess accountsservice and colord are more important get noticed ;) [18:14] last time i checked RAOF was onto colord, it has been a while already though [18:16] also osinfo, i think [18:20] for colord: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=699941 [18:20] Debian bug 699941 in libgusb "Update to 0.1.5 release" [Wishlist,Open] [18:53] ok [18:55] SonikkuAmerica: you're in Ohio? [20:09] ricotz, https://bugzilla.gnome.org/show_bug.cgi?id=693627 [20:09] Gnome bug 693627 in general "introspection errors in gjs make test" [Normal,Resolved: fixed] [20:42] jbicha, gnome-keyring http://pastebin.com/XpAHJQQC [21:06] darkxst, yeah i noticed that gjs got fixed in this regard [21:07] ricotz, the failing test happens even with js185 [21:07] altough jasper can't reproduce it [21:07] darkxst, if you want to tackle the gnome-keyring then do the whole thing and start with gcr [21:07] there is even a new lib irc [21:07] ricotz, I already did gcr [21:07] oh jbicha hasnt uploaded it [21:08] darkxst: you didn't split the library ;) [21:11] we never really came to a conclusion on that === SonikkuAmerica2 is now known as SonikkuAmerica [21:35] do you know what I need to run gnome-shell from jhbuild without hitting [21:35] Window manager warning: Log level 32: Execution of main.js threw exception: TypeError: Meta.BarrierDirection is undefined [21:39] jhbuild? What's that? [21:40] https://live.gnome.org/Jhbuild [21:41] darkxst: right, you could also ask in #debian-gnome on OFTC; of course they're trying to release wheezy with 3.4 so they probably aren't too concerned with 3.7.5 yet :) [21:43] jbicha, you need to patch out HAVE_XI23 in mutter configure [21:45] ubuntu xi has the barrier stuff patched out [21:49] hmm that doesn't seem to work [21:50] jbicha, try revert this http://git.gnome.org/browse/gnome-shell/commit/?id=fb63f48d0a0bd9b7db14400369237f8ba29da160 [21:51] and the one before it === SonikkuAmerica is now known as FatAlbert === FatAlbert is now known as SonikkuAmerica [22:00] darkxst: thanks that works, now I have to figure out who killed the usermenu... [22:01] * SonikkuAmerica sends detectives out [22:01] jbicha, mutter is supposed to fallback to the old barriers, but I suspect that it hasnt really been tested [22:02] gnome-calc http://pastebin.com/VAETCEsg [22:17] robert_ancell: we get to add an epoch to gnome-calculator [22:17] jbicha, oh, why's that? [22:17] because some maintainers can't count ;) [22:18] Hmm? [22:18] oh, someone bumped it to 6.x? [22:18] Is gcalc-tool doing something weird? [22:20] for upgrades to work, I'm pretty sure we need a transitional package and the transitional package has to have a higher version number than 6.6.2 === SonikkuAmerica is now known as GoogleSearch1832 === GoogleSearch1832 is now known as SonikkuAmerica [22:27] jbicha, just bump the xi dep in configure.ac and you are fine, no reverts needed