=== m_conley_away is now known as m_conley === xplinscott is now known as mlinscott [05:32] Morning === m_conley is now known as m_conley_away [06:07] Good morning [06:11] Good morning pitti! [06:13] hey RAOF, how are you? how's the jet lag? [06:13] Mitigated by the extreme length of time it took to actually get back :) [06:13] (I got back home *yesterday*) [06:16] ! [06:16] RAOF: how's that? you didn't fly with TheMuso? [06:16] No. I was flying with Kate and William and Ian. [06:16] QF 10 to Melbourne, then Melbourne→Hobart [06:17] So I spent Saturday night in a hotel outside Heathrow, QF 10 got in to Melbourne at 7pm Monday which was just in time to miss the last plane to Hobart for the day. So I spent Monday night in a hotel opposite Melbourne airport, and then got back Tuesday morning. [06:18] It was *totally awesome* [06:19] hmm, spending the night in a hotel for a change! [06:20] we really should do a sprint on a camping site one of these days! [06:21] I was saying that on Friday. [06:21] We should have an open-air sprint. Somewhere nice and sunny, but not too sunny. [06:24] with a forest and a lake, too [06:30] Yes. [06:30] It would obviously need to be *in* the forest, because working in the direct sunlight sucks. [06:46] Morning pitti. [06:46] RAOF, pitti, yeah I agree, an open air sprint would be nice. [06:56] wow, LibreOffice on armel, almost 5 days and still going... [06:59] good morning [07:02] bonjour didrocks [07:05] hey pitti! how do you feel? [07:05] didrocks: much better, thanks! [07:05] almost no pain any more, and I can eat bananas again [07:07] pitti: waow, seems you were quick to recover! You are quite lucky :) [07:08] didrocks: fortunately it was relatively easy indeed; no horizontal teeth or other jaw-breaking actions :) [07:19] RAOF: hey, it seems that the nvidia driver is not kind with me from my Monday upload [07:20] RAOF: I had to remove my xorg.conf (two monitors), rebooting just display on my external screen, and when I start nvidia-settings: http://paste.ubuntu.com/808272/ which is the laptop screen [07:21] didrocks, hello [07:23] didrocks, you missed the soname bump of librhythmbox [07:23] hey ricotz, how are you? [07:23] didrocks, thanks i am fine [07:23] ricotz: oh? didn't see it in configure.ac. Should have been already in latest snapshot [07:23] hoping you too? [07:23] yeah, I'm fine, thanks! [07:23] it got bumped on release 4 to 5 [07:24] *release, [07:25] ricotz: indeed, tweaking that. Thanks [07:25] yw [07:53] pitti: looking for the slow "shutdown" issue, seems that users are mainly pointing network-manager as being guilty [07:53] (reported my findings there: https://wiki.ubuntu.com/DesktopTeam/Meeting/2012-01-24) [07:54] didrocks: interesting, thanks! [07:55] yw :) [08:19] good morning everyone [08:19] hey chrisccoulson [08:19] hey chrisccoulson, how are you? [08:19] hi pitti, how are you? [08:19] hi didrocks, i'm good thanks. and you? [08:20] chrisccoulson: i'm fine, thanks :) [08:20] chrisccoulson: quite well, thanks! almost no pain any more, and I can eat bananas again :) [08:20] pitti, pain? what happened? [08:20] ah, that's good :) [08:20] bryce: got two wisdom teeth pulled yesterday [08:20] ahh [08:21] pitti, was it the lower wisdom teeth? [08:21] chrisccoulson: both on left side; right side will follow in a few weeks [08:22] pitti, good to hear you're recovering; I had mine out when I was in high school [08:23] same here, got them pulled out almost 10 years from now. [08:35] yeah, i had my bottom ones removed a few years ago [08:36] i actually quite enjoyed it, because i was given midazolam whilst they were being removed ;) [08:36] chrisccoulson: that must be a hell of a drug; I didn't particularly enjoy when they crowbared them out, felt like my jar was getting torn off [08:39] pitti - yeah, they had to cut my gums and drill the side of the socket out so that they could pull them out sideways (I've still got one of my top wisdom teeth in the way) [08:39] pitti, ow, sucks :-( [08:39] but i didn't really mind whilst they were doing it :) [08:40] chrisccoulson: *shudder* [08:40] guess I was lucky. Doctor said, "Breathe deeply," and next thing I knew I was in my Dad's car being driven home with wads of cotton in my mouth. [08:40] oh,you had a general anaesthetic? [08:41] yup, don't know exactly what it was but it sure did the trick [08:45] chrisccoulson: oh man, prepare for a world of pain [08:45] hi smspillaz [08:45] or rather, they injected so much local anaesthetic into my mouth when they did that that I couldn't eat or speak for about 3 days [08:46] heh :) [08:55] chrisccoulson: ubuntu-fr has a worse blackout FYI: http://www.ubuntu-fr.org/ [08:56] hey smspillaz [08:56] didrocks, nice! [08:56] it's a shame we don't officially participate in this ;) [08:57] yeah [08:57] http://linuxfr.org/ is nice too [08:57] lol [08:57] a SOPA-compatible image :) [08:59] smspillaz: do you know where we are with the "transparency regression" in the SRU? (so that we can upload a new version) [09:00] hey desktopers ;-) [09:00] salut seb128 [09:00] bonjour seb128, ca va? [09:00] lut didrocks, pitti [09:01] c'est l'heur de boulot? [09:01] hey seb128 ! [09:01] how are you? [09:01] pitti, en effert ! [09:01] ████████ seb128 [09:01] or "l'hour"? [09:01] didrocks: I had a quick look at the code and it doesn't look like any recent changes would have directly caused it. Looks like a nux state leak, I'll need to talk to jay [09:01] effet [09:01] pitti, "l'heure" [09:01] ah, merci [09:01] chrisccoulson, can't read that but hey to you ;-) [09:01] lut agateau [09:01] chrisccoulson: did you just say something inappropriate to seb128 :) [09:01] everybody is awake today it seems ;-) [09:01] smspillaz: ok, I think we will have to hold on the SRU because of it [09:01] chrisccoulson: luckily sopa is here to protect us [09:01] guh :/ [09:01] heh :) [09:02] didrocks: we should just remove that option, though I guess we can't do that in an sru [09:02] chrisccoulson, wth! you don't respect the coc there! [09:02] didrocks: sucks, because that branch should really boost performance [09:02] lol [09:02] pitti: there is a regresssion in the unity SRU ^^ (if you change the transparency for the panel in unity, it affects the decoration transparency (and some windows transparency as well) [09:02] ;-) [09:02] smspillaz: not really an option in a stable release [09:02] didrocks: ah, thanks for the warning; so we'll wait for a followup? [09:02] didrocks, how do you change that option? is that only ccsm? [09:02] didrocks: yeah I know, thats why it sucks [09:03] seb128: yes [09:03] "who cares"? [09:03] seb128: ccsm only, so not officially supported [09:03] but people will hate on us so ... [09:03] but a lot of people seems to do that, even at canonical and I got pinged a lot [09:03] like users who are able to shot themself in the feet with ccsm can deal with it [09:03] ok... [09:03] I can add a workaround in compiz for it, but thats about it [09:03] well, nobody made the link with the unity option apparently [09:03] they thought something changed into compiz :) [09:03] well if we stop on any option provided in ccsm that will be sport ;-) [09:04] seb128: that options that *we* provide [09:04] (for unity) [09:04] didrocks, I don't count ccsm as "provided" [09:04] well, if you take the blame, I'm ok ;) [09:04] no, that's fine, I just wish we would stop shipping ccsm ;-) [09:04] people will use gconf-tools, they are sneaky! [09:04] ccsm is not really the problem problem [09:05] *really the problem [09:05] didrocks, if they use gconf tools they are really on their own and deserver breakages they get though ;-) [09:05] smspillaz: I'm fine with the workaround, if it can be in unity as well :) [09:05] didrocks: yeah its just ... ugh, hang on [09:05] (for the SRU only) [09:05] smspillaz, the problem is the number of untested or not well maintained options we expose to users [09:05] well, anyway [09:05] pitti, how do you feel today? [09:06] pitti, did you manage to sleep normally? [09:06] seb128: indeed [09:07] didrocks: I can't test it directly right now, but unityshell.cpp:449 in trunk can be changed to [09:07] glPushAttrib(GL_VIEWPORT_BIT | GL_ENABLE_BIT | [09:07] GL_TEXTURE_BIT | GL_COLOR_BUFFER_BIT | GL_SCISSOR_BIT); [09:07] erm [09:07] glPushAttrib(GL_VIEWPORT_BIT | GL_ENABLE_BIT | GL_COLOR_BIT | [09:07] GL_TEXTURE_BIT | GL_COLOR_BUFFER_BIT | GL_SCISSOR_BIT); [09:08] is there a web page somewhere which describe the changes between each values of debian/compat? [09:08] *value [09:08] smspillaz: ok, thanks, I will try to deal with that and get back to you :) [09:08] agateau, man debhelper [09:09] agateau, look for "COMPATABILITY LEVELS" [09:09] agateau, it has the details of each version [09:09] seb128: oh thanks, I didn't scroll down enough [09:09] yw [09:11] seb128: it is a nice description of the mess I am in :/ [09:11] seb128: I am still working on sorting the installation dir of lightdm tools [09:12] agateau, ok, sorry but the update is taking time, my bzr skills suck, I need didrocks help [09:12] seb128: the only way to get libexecdir not to be /usr/lib/lightdm and thus pkglibexecdir not to be /usr/lib/lightdm/lightdm is to set debian/compat to 9... [09:12] seb128: but then we enter the world of multiarch... [09:12] agateau, the packaging is derivated from the 1.0 serie and we want to switch to 1.1 but they diverged, I'm not sure how to deal with it [09:13] seb128: meaning the tools are then installed in /usr/lib/$arch-dir/lightdm [09:13] hum, "nice" [09:13] seb128: which makes it painful to call them from postinst [09:13] agateau, you need a postinst.in generated from the rules or something [09:14] wow [09:14] I'm wondering why it went force and back for pkglibexecdir [09:14] but yeah, that's non trivial [09:14] from /usr/lib/lightdm to /usr/lib and then now back to /usr/lib/lightdm [09:14] good question [09:14] drop him an email ;-) [09:14] seb128: can't we just get the tools to install in /usr/bin or /usr/sbin? [09:14] that would be much simpler [09:14] seb128: it's a tooling issue, not robert's (the patch is mine, after a debian guy told it changed from /usr/lib/lightdm to /usr/lib in the tooling) [09:15] which was right at the time :) [09:15] agateau, what problem do you try to solve? [09:15] like is there anything not working now? [09:15] I want to be able to call lightdm-set-defaults from unity-greeter and lightdm-kde-greeter postinst [09:15] out of the dir being theorically wrong on disk from an unedeed subdir? [09:15] well you can do that today [09:16] it does not work right now because the postinst look for the script in /usr/lib/lightdm, not in /usr/lib/lightdm/lightdm [09:16] just fix the scripts? [09:17] you mean use "/usr/lib/lightdm/lightdm/lightdm-set-defaults" ? [09:17] that would be super ugly, no ? [09:18] especially since it has to be duplicated in all lightdm greeters [09:18] well, I guess we could install the utility in sbin [09:19] that's what I think as well [09:19] this should be done upstream, right? [09:19] I don't like much putting non-user-commands in the path [09:19] but you could argue that lightdm-set-defaults is a tool admin could be wanting to run as well [09:20] upstream> better yes [09:20] /usr/sbin is not in the path by default, so that won't affect non-root users [09:21] we already install the greeters in /usr/sbin, which makes even less sense imo [09:21] ok [09:21] let's do that then [09:22] ok, thanks, I am going to make a merge-request for this then [09:22] (think you need to change casper, ubiquity and some other sessions postinst IIRC ;)) [09:22] oh fun [09:23] didrocks: maybe on the packaging side we can ship a transition script in /usr/lib/lightdm? [09:23] didrocks: on the other hand, they are already broken right now [09:23] didrocks: so it won't be worse [09:24] agateau: well, it's quick enough to fix, just look at the impacted places [09:24] I can help you if needed [09:24] didrocks: I mean: sure I want to fix these, but it may not be worth setting up a transition script [09:25] agreed, if it's broken today :) [09:25] I assume it is, since precise lightdm package does not put the script in the right dir [09:26] * didrocks wonders how daily image starts then [09:27] that is weird indeed [09:28] we are not broken [09:28] only the ppa version has the wrong dir no? [09:28] the precise version should be ok? [09:28] * didrocks dpkg -L [09:29] seb128: ah, correct! [09:29] agateau, didrocks: I will workaround it from the packaging for 1.1.1 [09:29] i.e put it back to /usr/lib/lightdm [09:29] precise still has 1.0.6 [09:29] /usr/lib/lightdm/lightdm-set-defaults [09:29] didrocks, yeah, that's the correct location [09:29] yeah, that was my latest patch [09:29] something changed meanwhile on the toolchain again [09:30] ? [09:30] didrocks, agateau: http://bazaar.launchpad.net/~lightdm-team/lightdm/trunk/revision/1358 [09:30] that's your issue right? [09:30] hum no [09:30] similar but for the guest session [09:30] seb128: if you are into "beautiful" fixes, I did that yesterday: http://pastebin.com/bYRx0n6i [09:31] seb128: as ugly as it is, it works :) [09:31] but I guess there are better ways to do that [09:31] so, we were using libexec_PROGRAMS before [09:31] I assume passing --libexecdir=/usr/lib to configure would be enough [09:32] agateau, yes, I was going to say [09:32] that's a less hackish workaround ;-) [09:32] then, we got: https://bugs.launchpad.net/lightdm/+bug/861371 [09:32] Launchpad bug 861371 in lightdm "use pkglibexecdir instead of libexecdir for lightdm-set-defaults" [Low,Fix released] [09:32] ah here we go [09:33] so, in fact, it's assuming we are using dh9 [09:33] but now dh9 move pkglibexecdir to arch-based, isn't it? [09:33] mvo: good morning! sorry, was offline yesterday [09:34] I love this part: "(well in fact in now use multi-arch dirs but it's irrelevant here)." [09:34] mvo: I don't know how to get a real gobject pointer from a pygi wrapper [09:34] didrocks, we are not on v6 though [09:34] v9 [09:34] as long as you don't try to call the script, it's indeed irrelevant :) [09:34] seb128: indeed, we discussed with robert and he planned to move to it :) [09:34] mvo: if you need real GI bindings for gtkspell, perhaps you can open a bug for it? or do you get along with the ctypes hack for now? [09:34] well, let's put it in sbin [09:34] with a compat symlink to the old location in the packaging [09:35] agateau: maybe ask slangasek? there is maybe another autotools variable for installing in /usr/lib/ without being arch-linked? [09:35] or sbin :) [09:36] I wonder how other packages handle binaries installed in multi-arch dirs, I looked on my system and there are a few of them. I am afraid the answer will be "generate the scripts which calls those binaries" :/ [09:36] * agateau likes the sbin solution :) [09:37] agateau, well usually stuff in libexecdir are not meant to be called by others [09:37] pitti: its quite cool, hash(gobject) is the pointer to the gobject (which makes sense). the ctypes hack works, but it seems a much better idea to annotate the gtkspell, I guess thats all thats needed and looks like its not too much work [09:37] I will try to look into it later today [09:38] so, we need to do: the greeters, gnome-session, casper and the xubuntu settings package (maybe some edubuntu ones?) as far as I see [09:38] mvo: hash(object)> heh [09:38] didrocks: if we set up a symlink as seb128 suggested we don't have to do them all at once, though [09:38] agateau: indeed, but still, better to drop it before the end of the cycle :) [09:38] didrocks: agreed [09:39] ctypes really allows for crazy stuff, its fun (and unforgiving, each mistake -> SEGV) [09:40] didrocks, gnome-session? [09:40] didrocks, where,why does it set the lightdm greeter? [09:41] seb128: maybe it calls another bin from /usr/lib/lightdm? [09:42] seb128: it sets the lightdm default session [09:42] same tool :) [09:42] ah right [09:42] oh right [09:42] well in any case I'm going to do 1.1.1 with the old path still [09:42] I don't want to mix update and transition [09:42] pitti: closed the "focus" nautilus issue as ubuntu10 I uploaded this morning now unconditionnaly opens in a new window (which makes more sense, thinking about it) [09:42] we can do another revision which moves the bin to sbin and add a symlink [09:43] didrocks: thanks [09:43] but I would like robert_ancell to ack the move to sbin first [09:43] pitti: for the crash, it's the one I discussed with seb yesterday, quite tricky, and mixing libunity issue and nautilus one [09:59] smspillaz: are you sure about GL_COLOR_BIT? got a FTBFS/ error: 'GL_COLOR_BIT' was not declared in this scope === chrisccoulson_ is now known as chrisccoulson [10:01] didrocks: oh, hmm, its a little more involved than that actually [10:01] I'll try it out later today [10:01] ok [10:02] I would say GL_CURRENT_BIT but I'm not keen on that since iirc it was broken on nvidia [10:25] didrocks, agateau: ok, lightdm uploaded and vcs updated, sorry it took a while [10:26] didrocks, thanks for the help sorting the merge issues [10:28] seb128: thanks! looking at it right now [10:31] seb128: gdmflexiserver and lightdm-guest-session-wrapper are still in /usr/lib/lightdm/lightdm, that is going to cause troubles I am afraid [10:31] agateau, why? [10:32] seb128: aren't there other packages calling them... mmm, maybe not actually [10:32] agateau, the first one has its dir injected to the path by lightdm and the second is called by lightdm [10:33] agateau, well user switching and guest session still work for me [10:33] seb128: yw ;) [10:33] seb128: right, so moving them would actually cause trouble... you are correct [10:33] ;-) [10:33] well it wouldn't cause trouble, well at least not trouble over a restart [10:33] but the currnet path doesn't create issues either [10:35] seb128: I am currently looking into moving stuff to sbin, but we can't move gdmflexiserver there. I was thinking about adding a --with-gdmflexiserver-dir to ./configure, do you think it makes sense? [10:36] agateau, well, ideally we would drop that stuff and get a dbus service equivalent [10:37] it doesn't hurt where it's stored at the moment I would bother trying to change that [10:37] seb128: I am not exactly sure what gdmflexiserver does, I just know it must not be in $PATH by default :) [10:37] if I was to invest time on it that would rather be to deprecate the concept, and have a cross dm dbus service rather than a wrapper with gdm in the name ;-) [10:38] makes sense :) [10:38] agateau, it's an utility which can be used to give commands to gdm [10:38] like "switch to this user" [10:38] lightdm emulates it because that's what gnome-screensaver use to switch user [10:39] well having it in the path wouldn't be an issue, but it would make lightdm and gdm conflict since they would ship the same file [10:39] seb128: ok, makes sense [10:40] one other way to "fix" it would be to have a standalone source,binary with a gdmflexiserver wrapper [10:40] so gdm, lightdm, etc could depends on that [10:40] the wrapper would have the path location [10:40] other dms would rename theirs to i.e gdmflexiserver.gdm [10:42] and the wrapper would call the appropriate server based on the currently running dm? [10:42] right [10:43] would be nicer indeed [10:43] I don't like much how they did it for lightdm [10:43] they basically inject /usr/lib/lightdm to PATH [10:43] indeed [10:44] interestingly, there is a package named kdm-gdmcompat, which diverts gdmflexiserver, calling kdmctl if kdm is running and the original gdmflexiserver if not [10:44] that's a bit similar to your solution [10:44] right [10:44] we should consolidate all that ;-) [10:44] didrocks: do you happen to have a way to reproduce bug 901689 ? [10:44] Launchpad bug 901689 in pygobject "oneconf-service crashed with signal 5 in g_object_newv()" [Medium,Confirmed] https://launchpad.net/bugs/901689 [10:45] pitti: no, that's part of the issue in pygobject, everytime I tried to reproduce them, I don't get anything :/ [10:45] ok [10:45] I have no clue how to shake this thing [10:49] didrocks: might be another "only happens on shutdown" errors then? [10:49] if it's reproducible and happens to many people, I'm interested in looking into it [10:50] but not much that I can do without a reproducer :( [10:50] the stack trace doesn't help there, one needs the python code [10:50] pitti: agreed, can totally be a "happens on shutdown" [10:50] but it doesn't seem to be that important after all? [10:53] pitti: if it's the case, no it's not. The thing is that i have various (but different pygobject crashes) and I really would like to help there [10:53] * didrocks needs to find time to look at the code === DBO is now known as jsmith === jsmith is now known as DBO [11:20] Hi all. [11:20] hey Sweetshark [11:21] pitti: bug 917153 is looking fun [11:21] Launchpad bug 917153 in libreoffice "failed to upgrade from oneiric to precise: /usr/lib/libreoffice/program/unopkg.bin: error while loading shared libraries: libicule.so.48: cannot open shared object file: No such file or directory" [High,New] https://launchpad.net/bugs/917153 [11:21] missing dependency? [11:21] -common postinst needs unopkg (from -core), but -core depends on -common [11:22] didrocks, interesting comment on the nautilus bug, it does segfault when you try to use the ql but no nautilus dialog is open [11:22] didrocks, which I can confirm here [11:22] didrocks, which might be a different issue than the bookmark refresh one you mentioned [11:22] possible solution: make unopkg an package itself and depend on it from -core and -common. do you have a better idea? [11:22] seb128: hum, even with the version I gave to you yesterday? [11:22] didrocks, yes [11:22] that's the one I'm running [11:23] the one which spawns new dialogs [11:23] pitti: do you have a better idea? [11:23] Sweetshark: ah, cyclic dependency [11:24] Sweetshark: either this, or make it work also if the postinst didn't run yet [11:24] didrocks, it might explain the number of duplicates, I guess people try to use the ql to open nautilus when no dialog is open yet [11:24] seb128: oh you're right [11:24] * didrocks run nautilus and attach gdb [11:24] seb128: not sure it's the same stack than what we had yesterday though [11:24] seb128: is there a duplicate handy with ubuntu10 ? [11:26] pitti: well, -common cant really do much without -core anyway (an extension to a product that isnt installed yet?!), but at some point in time, the postinst has to run. [11:26] didrocks, no but the stacktrace seems the same than bug #917821 here [11:26] Launchpad bug 917821 in nautilus "nautilus crashed with SIGSEGV in g_file_is_native()" [High,Confirmed] https://launchpad.net/bugs/917821 [11:26] hum, that is really weird then :/ [11:27] Sweetshark: does -core really need -common that much then? you might be able to drop that from Depends: to recommends;? [11:27] * didrocks installs dbg there [11:27] pitti: one could move the registering of extension to the -core postinst though. [11:27] didrocks, well with activate_bookmark_in_menu_item -> activate_bookmark_by_quicklist [11:28] didrocks, but the top 3 functions are still the same ones [11:28] yeah, that's what is weird, I would expect it to crash when I get the application [11:28] pitti: I guess -core is half of a metapackage thus I would need to add -common then to all the stuff that depended on -core before [11:29] s/depended on -core/depended only on -core/ [11:30] * Sweetshark loves debtree btw for this ... [11:31] hehe [11:31] * pitti -> lunch, bbl [11:31] Sweetshark: if -common is just meant to provide the arch: all data, then moving the registration into -core sounds fine as well [11:32] Sweetshark: but then you still need to break the dependency loop [11:32] apt breaks those in an arbitrary position, so you won't know which one will be configured eventually [11:33] oh, there is no loop as -common doesnt yet depend on -core [11:40] seb128: the stack is different there [11:40] which makes sense :) [11:40] didrocks, is it? [11:40] what do you get? [11:40] seb128: http://paste.ubuntu.com/808450/ [11:41] but this is due to ubuntu10, not ubuntu9 which is the quicklist race I bet [11:41] didrocks, hum, weird, I guess 10 has both issue and I ran into the other one [11:42] seb128: yeah, the other issue isn't fixed [11:42] so you totally can have it, but with a window opened or not [11:42] didrocks, I've tried to valgrind without much luck [11:42] I asked kamstrup for an idea on the quicklist failure [11:42] as we more or less know why and that there is no magical way to fix it on the nautilus side alone :) [11:43] (well, meaning we need a proper fix) [11:43] looking at the other issue now [11:44] Urgh. Unity's alt-tab now hates my dual-head setup, it seems. [11:46] seb128: confirmed, I can get both issues there [11:46] didrocks, http://pastebin.ubuntu.com/808459/ [11:46] didrocks, there is an invalid unref call somewhere [11:48] seb128: hum? it seems quite different from the stacktraces we have [11:48] as it doesn't crash on a gobject_unref [11:50] didrocks, well, valgrind showed that error, it might be yet another issue, or the invalid unref might be what lead to the segfault a bit later [11:50] btw when running under valgrind and use the ql I get some of those "GLib-GObject-WARNING **: invalid unclassed pointer in cast to `GtkWindow' [11:50] " [11:50] when you have new opened window, isn't it? [11:51] yes [11:51] yeah, I think I can fix this issue quite easily [11:51] it would help if I didn't get screens and screens of [11:51] ==6934== Conditional jump or move depends on uninitialised value(s) [11:52] ==6934== at 0x521C650: ??? (in /usr/lib/i386-linux-gnu/libpixman-1.so.0.24.0) [11:52] ==6934== by 0x5208E28: ??? (in /usr/lib/i386-linux-gnu/libpixman-1.so.0.24.0) [11:52] ==6934== by 0xE3BC743: ??? [11:52] wth libpixman! [11:52] didrocks, well anyway I think nautilus could use with some fixing and valgrinding out of your issue, but I will do that later [11:52]