[07:21] ricotz, are you seeing focus issues in overview? [07:21] with my gjs patches? [07:24] darkxst, hmm, focus issues of what kind? [07:25] (you probably want to show the patches to Jasper too) [07:29] ricotz, I can't use the mouse to click anything in overview anymore, but it worked yesterday [07:29] also alt-tab gets stuck, which also worked yesterday [07:29] Jasper has looked at my patches [07:30] mouse in overviewe and alt-tab is working here [07:30] ok, hopefully something is hosed in my jhbuild [07:30] your patches with the changed compat OBJECT [07:30] btw dash is fixed [07:31] what was it? [07:31] bug in my JS_NewNumberValue patch [07:31] have replaced that one with a wrapper [07:31] ah [07:32] have you guys been crashing in Gnome-Shell when closing apps in 13.04 dev? Every 3-4 times I close a window the whole system becomes unresponsive to input [07:32] brittyboi, nope [07:33] hrm [07:33] brittyboi, try to disable the event-sounds if they are enabled [07:33] brittyboi, can you get a backtrace [07:33] how do I do that? [07:33] I can try, but it isn't being reported in ant logs that I can see [07:33] g-c-c > audio > effects > disable [07:33] hold on I'll get you guys the bug number, i'm not the only one [07:34] i am suspecting your are hitting the eglibc bug triggered by the newer glib [07:34] https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/1094571 [07:34] Launchpad bug 1094571 in gnome-shell (Ubuntu) "gnome-shell (or session) continually crashes/locksup when closing any window" [Undecided,Confirmed] [07:35] I'll try that right now [07:35] darkxst, could you paste a new patch-set? [07:38] ricotz, https://www.dropbox.com/s/dipa498sjc5jcm9/gjs188-r2.patch [07:38] darkxst, thanks! [07:40] np [07:42] darkxst, i guess this is missing some parts [07:43] like the defintion of HAVE_JS_NEWNUMBERVALUE [07:43] ricotz you need to rerun autogen [07:43] i did [07:43] thats there +AC_CHECK_FUNCS(JS_NewNumberValue) [07:44] check the config.log then I guess [07:45] although that should fail anyway on js188 [07:45] since its removed [07:46] it is #ifndef [07:47] ricotz, did build fail? [07:47] ok, but putting JSVAL_IS_OBJECT into it is wrong [07:48] since you want the opposite here [07:48] ricotz, no, I am just cheating and using it to detect js188 [07:49] ah ok [07:50] it could just as well be #ifndef HAVE_JS_SOMEOTHER_REMOVED_FUNC [08:05] ricotz, working ok ? [08:11] not yet [08:27] ricotz, oh you need to add static back here +JSBool JS_NewNumberValue(JSContext *cx, double d, jsval *rval) (in compat.h) [08:28] (Jasper told me to remove it ;( ) [08:33] darkxst, i know this is only one part of the problem [08:33] what other problem? [08:34] linking against libgjs [08:34] ricotz, are you using jhbuild? [08:34] you should see at least warnings for gjs-console [08:34] no, i am not [08:35] oh right [08:35] while compat.h is a public installed header putting things like #ifndef HAVE_JS_NEWNUMBERVALUE in there [08:36] doesnt work [08:38] hmm right [08:38] but it should be undefined? so why doesnt it work? [08:39] either way you can just remove the #ifndef [08:40] oh I suppose its picking up js185 [08:40] if you have it installed then remove it [08:41] I am using jhbuild [08:41] which links explicitly to the mozjs188.pc [08:41] (i have the 188 headers installed so no js185 headers) [08:42] note that the header locations are the same [08:42] try to remove your systems mozjs headers [08:43] ricotz, I can't do that, I am using them! [08:44] so you only have 188 around? [08:44] no I have js188 headers [08:44] and multiple other libs [08:44] all off your ppa [08:45] ok [08:45] but I am building gjs/g-s from jhbuild [08:46] alright [08:47] i think removing the conditional stuff is better [08:47] ricotz, I basically only added it in at Jaspers request [08:47] and squash those hunks in the change build version to mozjs188 patch [08:47] i see [08:48] seems not optimal to do so [08:49] whats not optimal? [08:49] trying to support backward compat in this case [08:50] the other changes arent compatible either, so why bother to try it for those [08:51] yeh probably wont happen, given the JSHandleObject/Id stuff [08:51] darkxst, btw, it is running now for some time [08:52] without "focus-issues" [08:53] ricotz, cool, let me know any issues you do see, I think a number of the ones I am seeing are unrelated and that somehow jhbuild has hosed itself (or a broken commit has landed) [08:53] they seem more like mutter issues really [08:56] any gotta run, dinner time here [08:57] darkxst, alright, thanks for your work! [09:01] np, although it would have been much easier if the changes were actually documented ;) [09:02] which reminds me, I probably should add these to the wiki page [09:04] ^mozilla wiki [22:05] ricotz, http://git.gnome.org/browse/pango/commit/?id=48423ef827f5fcedac92e1e73c29f0bb94eeadda [22:05] can you update that? [22:06] darkxst, why? [22:08] i hope this new fontconfig will get into raring though [22:08] pango depends on it [22:08] well I will just revert that commit for now [22:08] it just depends on it to make the tests happy === john__ is now known as _j0hn [22:23] darkxst, i guess i can upload it to my staging ppa [22:28] well it seemed to build ok, after reverting that commit [22:29] it is just needed to have pango working "reliable" with multithreading [22:30] make check indicates the problems [22:30] https://launchpad.net/~ricotz/+archive/staging/+sourcepub/2922082/+listing-archive-extra [22:31] robert_ancell, hi, is this a candidate for raring yet in its early state? ^ [22:32] ricotz, I don't know enough about fontconfig/pango but we were taking a cautious approach to shipping them [22:32] ricotz, having it in a PPA and getting some confirmation it works well will help [22:33] robert_ancell, alright, will use it for some time [22:33] and maybe move it to testing ppa after [22:35] (it might solve a crash with nautilus i saw occasionally with the harbuzz/pango/gtk git-stack)