[01:02] jbwiv_: did you get your question answered yet? [01:38] Hmm...so I have two different ATI cards, both running with fglrx. One detects hotplugs; the other doesn't, but detects the new display if I run xrandr -q. I know fglrx is unsupported territory, but does anybody have thoughts on where I could look a priori to determine whether hotplug works or not? [01:38] The alternative is to background a shell script running "while true; do xrandr -q; sleep 5; done", which just seems kind of lame [01:40] (well, I guess I'll have to do that for Card #2, but doing it for Card #1 as well just seems unfortunate if it's avoidable) [01:43] broder, not sure myself; but you may have better luck in #ubuntu, which is the support channel. This is a development channel [02:03] broder: That's a fine question; I'd guess you'll find that it's based on card generation, but that's possibly the sort of thing that individual card manufacturers can screw up, too. === asac_ is now known as asac [02:21] good morning everyone! [02:37] good morning [02:48] chrisccoulson: Good morning, for it is still just morning for me. :) [02:49] TheMuso, hi! yeah, it's morning here for me too :) [02:54] definitely not morning here in the US === smspillaz|zzz is now known as smspillaz [02:57] rickspencer3, who would approve -devel e-mails to the list (i just replied to the unity as default e-mail)? [02:59] chrisccoulson, morning :( im about to contemplate going to bed [03:05] where is the wiki for travel info to pair up for UDS ? [03:07] bcurtiswx, did it not go through? [03:07] rickspencer3, i got a bounce e-mail from -devel [03:07] not sure exactly [03:07] It may appear to be a couple steps back, but I think in the end we [03:07] will find that Unity as the default desktop environment for 11.04 will [03:07] be a gigantic leap forward later on. [03:07] ? [03:08] the amount of attention it will receive if we make it default now vs later [03:09] bcurtiswx, oh, right [03:09] sorry, I was just trying to confirm for you that your mail got through [03:10] thanks for hte note bcurtiswx [03:10] rickspencer3, ah OK. thanks for allowing another voice [03:11] I think bcurtiswx is referring to ubuntu-devel@lists.ubuntu.com. [03:11] TheMuso, I am [03:11] but now i'm confused [03:11] Right, I am not sure who has mod powers on that list. [03:11] TheMuso, did his email not come through there? [03:11] it went through [03:11] maybe it only came throuhg on the desktop list, I got it twice [03:12] Haven't checked, ahng on. [03:12] jcastro, but I wrote to @u-desktop, and cc'ed u-devel [03:12] jcastro, thx [03:12] anyway, TheMuso, RAOF, robert_ancell, if you disagree, you can raise the issue with your no-good-boss jasoncwarner1 ;) [03:12] oh right, I see that [03:13] I have no issue with it. :) [03:13] I can't wait to be able to use it more. [03:14] * RAOF waits for his mail to finish filtering. [03:17] hi all [03:18] hi! [03:19] ayan, says that he just upgraded, it sounds like he's getting only the classic desktop, but he wants Unity [03:20] ayan, did you choose Ubuntu when you logged in? [03:20] ah ha! my unity does not support my graphics hardware. [03:20] just fyi, i'm using a thinkpad x201. [03:21] ayan, well, you could try installing hte Qt version [03:22] ayan: Which particular x201? They should all be pretty well supported, as they're using Intel graphics, but there's all sorts of variation possible. [03:24] RAOF: how do i distinuish it from other x201? would you like the model number? [03:24] robert_ancell: You're a GTK god, right? Banshee seems to be consuming ~10% CPU while playing music due to the position slider being updated at 5Hz, but it looks like it's triggering an expose event for the entire window. Is that something gtk normally does? [03:25] ayan: The question is mostly how new it is, and what CPU it has. Another factor is whether it uses eDP or LVDS (and there's not a trivial way to find out, really), but since you can apparently *see* stuff on the screen that's not likely to be a problem :) [03:25] ayan: This is probably better done in #ubuntu-x, too. [03:25] RAOF, um, don't know it that well. This is GTK2 right? I think it would have to, to draw the window underneath the slider. But the area should only be under the slider and nothing else should need drawing [03:26] RAOF, does banshee have a lot of custom drawing in the main window expose method? [03:26] Banshee is *all about* custom drawing. [03:26] yeah, I thought that might be the case [03:26] simple-player by default in 11.10? [03:27] it shows the song title [03:27] I'll check my hypothesis then go expose hunting. [03:27] RAOF, thinking about it, it really must draw the background - if you had a pixmap window background then it would have to draw it, then the slider elements over it [03:27] rickspencer3: Muinshee works too, presumably because it's got a much simpler main window :). [03:27] * robert_ancell -> lunch [03:32] Ah. I see we've managed to hook up fsck.btrfs to the boot sequence now. Hurray ☹. [03:36] RAOF: hrm interesting. I thought that would have been one of the first things that needed doing for btrfs support. :) [03:37] Well, given fsck.btrfs doesn't actually *fix* any errors yet I can see why it might have been lowish priority :) [03:38] Ah, right. [03:38] Probably because the FS is still maturing. [03:42] RAOF: did dbo and/or smspillaz talk to you guys about documenting video hw requirements for unity? [03:43] jcastro: Not as far as I'm aware. [03:44] don't worry they were going to do it, I just told them to run it by you guys first. [03:44] * jcastro will follow up with them [04:44] Wow. Why is text rendering so expensive? [04:45] RAOF: 3d ? [04:45] No; A (pretty simple) pango layout. [04:46] But rendering “1:02 of 3:33” at 5Hz results in ~5% CPU utilisation. [04:47] win [04:48] Yeah. [04:49] try xterm :P [04:49] * lifeless doesn't have any pango based text visible on screen most of the time [04:50] Yay bitmap fonts. [04:50] hmm, I thought it did them all these days [06:08] Ok. I've got *most* of the Banshee CPU use killed. [06:08] robert_ancell: I don't suppose you'd know whether receiving two essentially simultaneous expose-events for a widget is normal? [06:08] RAOF, I would have thought that shouldn't happen [06:09] So would I. [06:09] is anything calling the expose handler directly? [06:09] Not that I can tell. [06:09] otherwise everything should just call gtk_widget_queue_draw and only one redraw occurs [06:10] does anything call queue_draw inside the expose event? [06:10] Sequence is: UpdateLabel() calls QueueResize(); OnSizeRequested gets called exactly once; OnExposeEvent gets called twice. [06:11] The only thing which could possibly do that is the call to PaintLayout, but if that were the case I'd expect to get an unending stream of redraw events, as it's called each redraw. [06:12] RAOF, hmm, I just made a little PyGTK example. There are four exposes on startup, and two when queue_resize is called [06:12] but they have different areas [06:13] (Specifically, that's gtk_paint_layout()) [06:13] http://paste.ubuntu.com/591096/ [06:13] Hm. I get identical areas. [06:16] Although that might be because the actual size isn't changing. [06:18] Hm. And if I don't redraw the second time I don't get any text. [06:22] And if I don't redraw the *first* time and do the second time the text is there. Presumably the first redraw is clearing the background, or somesuch. [07:02] RAOF, hey, do you know why we didn't go with mono 2.10 in natty? [07:02] Yup. [07:02] Because it wasn't packaged until about 5 days ago. [07:02] And 2.10 had a bunch of regressions. [07:02] fixed now? [07:03] Yeah. [07:03] Mono's really maintained in Debian, so the freeze had a certain chilling effect, too. [07:04] But 2.10.1 is in experimental now, so we'll have it in Oneiric early. [07:07] RAOF: poke [07:07] RAOF: was the nvidia driver updated recently ? [07:07] Um... [07:07] Not by me, obviously. [07:07] (I'm looking at bug 752445) [07:07] Launchpad bug 752445 in compiz (Ubuntu) "Intermittent white window contents when maximizing/switching windows" [Undecided,New] https://launchpad.net/bugs/752445 [07:08] RAOF: ah, well in the pacakge ? [07:08] Most recent change was 18 Mar. [07:08] I'm thinking that we must have gotten a broken driver and was wondering if we could revert it if possible [07:09] since window contents should definitely *not* be white just because the card is running low on texture memory (it only happens with large windows which is why I suppose this happens) [07:09] I haven't been able to confirm this myself unforuntately [07:10] Bah. [07:10] yeah, exactly [07:10] Hurray for the black window bug! [07:11] RAOF: I'm wondering if this is related to the memory usage smackdown [07:11] Now back, but, like the iPad, available in white! [07:11] * TheMuso can check natty-changes for the last upload if thats of any help. [07:11] RAOF: I haven't been able to get any compiz logs of this, but I would imagine that glXTexturePixmapEXT is returning false because the server could never allocate the pixmap [07:12] Yeah 18 march with a new upstream release. [07:12] Trouble is, the new upstream release properly supports the new ABI... [07:12] brilliant [07:12] this is why prorprietary drivers suck [07:12] [07:13] smspillaz: Time to force indirect rendering on nvidia again? :) [07:13] RAOF: unity doesn't work with indirect rendering [07:13] RAOF: it needs the available GL version to be >= 2.0 [07:13] otherwise it willdie [07:14] nvidia's GLX doesn't have protocol for newer GL? [07:14] * RAOF has never tried, obviously. [07:14] RAOF: I'm pretty sure the server GLX protocol only goes up to 1.6 ? [07:14] smspillaz: The *open source* GLX protocol goes up to 1.4, but the nvidia drivers replace that. [07:14] RAOF: I guess what we could do for a workaround is try to detect where texture bindinding failed and attempt to re-bind a few seconds later once the server's got its stuff together [07:15] RAOF: Hm, I don't really have unity available to try on me right now, so I can't confirm if --indirect-rendering will work or not [07:15] Now that's the sort of high-quality workaround we like to see :) [07:15] *sigh* [07:16] RAOF: unfortunately, we can't rely on nvidia to fix this [07:16] and I see this going from "high" to nuclear if we don't do something about it [07:16] It'll only affect crappy nvidia cards… :) [07:16] unfortunately we have to support crappy nvidia cards [07:16] and crappy drivers too [07:17] You're right that nvidia won't have a fix by release time, though. [07:17] indeed [07:17] RAOF: indirect is an option if we can make it work, however we got forced to indirect by fglrx and look at that disaster [07:17] indirect is also really slow [07:17] it will make unity slow [07:18] Isn't unity already slow on nvidia? I seem to remember reports of such. [07:18] its quite snappy on both nvidia and nouveau [07:18] I've tested on nouveau, yeah. Just not nvidia. [07:19] why does that seem strangely familiar [07:21] :) [07:27] * micahg is running nvidia ATM [07:29] RAOF: Have you ever looked into how one could force sbuild to continue should there be a checksum mismatch with archive package metadata? [07:30] Granted I haven't either, but you may know of something that would take me a while to find out otherwise. [07:30] Since you have explored sbuild in greater depth than I. [07:30] ... [07:30] that seems undesirable [07:30] What do you mean? In the case of a package download having an incorrect checksum? [07:30] nag mvo to make apt try harder [07:32] Yeah I know its undesirable, but sbuild never checked/updated archive metadata in the past... [07:32] Oh, you mean the pre-build apt update? [07:33] set $apt_update = 0; in ~/.sbuildrc [07:33] ooh thanks [07:36] good morning [07:38] Hey didrocks. [07:38] hey TheMuso :) [07:39] Hey didrocks [07:39] evening RAOF [07:40] Good morning [07:40] Hey pitti. [07:40] Heya pitti [07:41] hey RAOF, hello TheMuso! [07:42] morning pitti [07:42] bonjour didrocks, ca va? [07:43] * bryceh_ waves [07:43] pitti: ça va bien, merci! :-) you? [07:43] hey bryceh_ [07:43] heya didrocks [07:43] pretty well, thanks! [07:45] Ah crap, taking out the games from powerpc only knocked off 4MB. :S [07:45] I guess next step is to try and knock off all dev related packages... [07:53] morning didrocks [07:53] hey smspillaz [08:22] fta2: poke [08:28] TheMuso: do we actually have that many? you mean gcc/binutils/linux-headers etc? [08:29] pitti: yeah, mostly gcc/binutils, since kernel headers were cut in the maverick cycle. [08:29] Morning all [08:29] I haven't looked hard, because I was hoping the games alone would be enough, but obviously not. [08:29] hey Sweetshark [08:30] an extra kernel weighs a lot :-( [08:34] pitti: Yeah but I am tlaking about headers only here. I can't really drop one or the other unfortunately. [08:34] I can't drop one kernel or the other. [08:45] bug 512395 is kind of pointless as the dekstop files are currently generated directly in the packaging, right? [08:46] Launchpad bug 512395 in openoffice.org (Ubuntu) "Openoffice.org's .desktop files do not contain translation domain info" [Wishlist,Triaged] https://launchpad.net/bugs/512395 [08:46] morning [08:46] hey rodrigo_, Sweetshark [08:47] hi didrocks, Sweetshark :) [09:08] Riddell: good morning [09:08] Riddell: is the package split in bug 597254 on your radar? [09:08] Launchpad bug 597254 in srtp (Ubuntu Natty) "[MIR] srtp" [High,Incomplete] https://launchpad.net/bugs/597254 [09:10] rodrigo_: hey, how are you? [09:10] rodrigo_: can you please give me a quick status about bug 649809? is the problem there understood, so it's mainly a manpower issue, or still needs to be debugged? [09:10] Launchpad bug 649809 in gnome-settings-daemon (Ubuntu Natty) "the session settings manager can try starting before the login screen one exits" [High,In progress] https://launchpad.net/bugs/649809 [09:13] smspillaz, hi [09:13] pitti, still trying to find what's failing [09:13] pitti, hi, is the retracer dead (once again)? [09:13] pitti, I thought it was the u1's g-s-d plugin, since on vanilla version it works for everyone [09:14] fta2: describe in detail this "screen corruption" issue you're having and exact instructions to reproduce [09:14] fta2: yes, someone updated udev again, which keeps breaking the dist-upgrade; fixing [09:14] and whether or not you are on nvidia [09:14] rodrigo_: ah, thanks; so in the "debugging" phase then? [09:14] smspillaz, which bug? [09:14] pitti, yes [09:15] rodrigo_: do you think that this is a regression and affects so many people that it should be release critical? [09:15] fta2: the one you pinged me about yesterday (can't find it in my scrollback) [09:15] pitti, it seems to affect less people than before the fix, so not sure [09:15] pitti, but at least high priority, yes [09:16] rodrigo_: thanks for the heads-up [09:16] smspillaz, hmm, i didn't ping you since you told me to stop with the xterm bug [09:16] fta2: right, I know about the xterm bug. didrocks said you had an issue with screen corruption [09:16] smspillaz, but indeed, i filed a regression bug against compiz yesterday [09:16] not xterm [09:16] smspillaz, with a video [09:16] chromium [09:16] hey desktopers [09:17] fta2: link me to it [09:17] bonjour seb128 [09:17] didrocks, hi, I have a new unity crasher for you :) [09:17] bug 754098 [09:17] Error: Launchpad bug 754098 could not be found [09:17] pitti, hmm, just saw the last comments, some people using maverick, so I guess we should backport the fix to maverick also [09:17] pitti, guten tag! ;-) [09:17] hey seb128 [09:17] hey rodrigo_ [09:17] rodrigo_: oh, you think that they said "still happens for me" because of that? [09:17] good point [09:18] pitti, no, one of them did, not sure about the others, so just asked them in the bug [09:18] rodrigo_: perhaps you can follow up asking if anyone still gets that under natty? [09:18] smspillaz, https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/753369 [09:18] Ubuntu bug 753369 in compiz (Ubuntu) "artifacts in chromium" [Undecided,New] [09:18] rodrigo_: great, thanks! crossing fingers :) [09:18] fta2: exact instructions to reproduce [09:18] there's no video attached [09:18] pitti, yes, do I prepare a maverick update anyway? [09:18] fta2: checking [09:19] rodrigo_: if it's a safe patch and will help people, sure [09:19] smspillaz, no, 753369 for you, 754098 was for didrocks ;) [09:20] smspillaz, the video says it all, any page with a selector will do [09:20] there's no video attached to that [09:20] pitti, well, let's see what people answer first [09:20] fta2: is is private or whatever? [09:20] oh found it [09:20] seems like https://bugs.launchpad.net/bugs/754098 is invalid [09:20] Error: Launchpad(https://launchpad.net) bug 754098 not found [09:20] nevermindthat [09:20] didrocks, it's probably not retraced, I'm checking on those [09:21] didrocks, is it? it's waiting for the retracer [09:21] fta2: can't reproduce it [09:21] fta2: if it's a damage bug at this stage it will be too difficult to look into time wise [09:21] smspillaz, with chromium? [09:21] yeah, the update failed again [09:21] fta2: yes [09:22] smspillaz: I confirm it [09:22] fta2: nvidia? [09:22] didrocks, yes [09:22] didrocks: I can't reproduce it [09:22] smspillaz: nvidia with proprieratery driver maybe? [09:22] and damage issues are *very* *very* *very* difficult to debug [09:22] didrocks: I have the properietary driver [09:22] pitti, hum, did you work on the retracers today? [09:22] it's very difficult because it requires tight timing and I don't have time to be spending hours on it [09:22] seb128: as we speak [09:23] usually race conditions like that are the fault of the driver, there's simply nothing we can do about it [09:23] seb128: i386 fixed/restarted, amd64 fixing right now [09:23] pitti, doh ok, that's why upgrade,dist-upgrade lists 0 package to update [09:23] seb128: see fta2's question from 5 mins ago [09:23] smspillaz, well, it's a regression with the last update, should be easy to track down [09:23] pitti, yeah and see my response [09:24] fta2: right, but I can't reproduce it on *either* the update before or the newest update [09:24] seb128: ah, sorry about that then [09:24] and the newest update doesn't even touch the damage handling code [09:24] pitti, no worry ;-) [09:24] smspillaz, (last as in when the bug was filed) [09:24] right, I know [09:24] smspillaz, which chromium? i see it with ch12 (trunk) [09:25] if its a matter of which chromium version you are using then it is *definitely* not a compiz bug [09:25] seb128, fta2: both fixed and restarted [09:25] fta2: 10.0.648.204 (79063) FYI [09:25] fwiw, i see similar artifacts with opera (nvidia blob) [09:25] smspillaz, no, i don't know, as i said in the bug, i checked older versions of ch that used to work fine before the last compiz and they are now broken [09:26] hence the regression in compiz [09:26] pitti, thanks, I've cleaned the logs so I'm reading to run my grep magic on those to see what issues were raised and which ones have duplicates today [09:26] fta2: right, ok, well the damage code in compiz was *not* touched at all recently [09:27] so I can't see any good reason for why that is [09:27] didrocks, the retraced nux crash was http://paste.ubuntu.com/590979/ [09:27] didrocks, i had a bamfdaemon crash yesterday too [09:27] fta2: (damage == the bit that handles which parts of the screen need redrawing) [09:27] fta2: can you ping me once both are retraced? I'll add them to the list for a fix release [09:28] pitti, btw I will like send an apport patch your way today to put some extra infos in the retracer logs, like add the ubuntu version and the timestamp to the "Report is a duplicate of" lines [09:28] not good. [09:28] seb128: oh, thanks [09:28] pitti, it would make easier for me to make daily stats about natty crashes and to report the most popular ones [09:29] I clicked around a bit in the system settings->appearance and now my windows controls minimize/maximize/close are on the RIGHT [09:29] didrocks, i didn't file bug for the bamfdaemon crasher, not sure how it happened [09:30] apparently an assert.. "Failed to register GObject with DBusConnection" [09:31] but i can file one if you want [09:33] Does anyone have an idea how I reset that setting? Where is that stored? ~/.compiz? [09:34] Sweetshark: I'm pretty sure it's in gconf [09:35] Sweetshark: /apps/metacity/general/button_layout « you can change there [09:35] vish: also for unity/compiz? [09:35] yup [09:35] * Sweetshark tries [09:37] Sweetshark: if it has not been altered there, it usually is cause the gnome-settings-daemon has bad memory ;p , you'd have to close and reopen the appearance settings and re-select the theme [09:37] pitti: I dropped our build-depend on srtp and linphone so it should be possible to move to universe now [09:37] or that way around, thanks [09:37] RAOF, vish: thank you very much. [09:38] smspillaz: issue in chromium after bookmarking a page: http://img.xrmb2.net/images/985729.png in opera after closing a bookmark menu: http://img.xrmb2.net/images/955227.png - does not happen without compiz, does only happen with nvidia blob [09:39] * Sweetshark is tempted to start a "gconf is like the windows registry"-flamewar. After all it is friday. [09:39] htorque: I'll try reinstalling. I cannot reproduce any of these issues [09:41] didrocks, just curious, why can't you see private bugs? I can [09:42] pitti: re: the icon for nautilus on launcher, i removed the icon which has a pointer on the folder and the icon of the cabinet is from gnome-icon theme(but seems Mark doesnt like tht either and I'm not sure if anyone is working on a new version, maybe sladen knows or didrocks ); the home folder icon i've pushed to ubuntu-mono and that needs to be released [09:42] fta2: you need to be susbribed to see it [09:42] vish: sladen is the design packager [09:42] didrocks, i'm bug control apparently, but maybe you should be too [09:43] didrocks: cool! but just letting you know as well, i dont know what's going on there :) [09:43] smspillaz: maybe easier to reproduce: make a gnome-terminal small enough so a menu would stick out of the window, then move the window. here's what i get: http://img.xrmb2.net/images/112558.png [09:43] chrisccoulson: just discussed mono for bug 740815; this will resolve mono, gluezilla, and xulrunner-1.9.2 tasks; so I think the remaining one is LibO [09:44] Launchpad bug 740815 in xulrunner-1.9.2 (Ubuntu) "[FFe] Updates to enable us to drop xulrunner from main" [High,In progress] https://launchpad.net/bugs/740815 === kklimonda is now known as Guest46494 [09:44] pitti - excellent, thanks! [09:45] pitti - would you mind sponsoring the xdg-utils diff on bug 670128 when you get a few moments please? :) [09:45] Launchpad bug 670128 in xdg-utils (Ubuntu Natty) "gnome-open uses firefox while it's not the preferred browser" [Medium,Triaged] https://launchpad.net/bugs/670128 [09:45] vish: I don't know either, let's wait sladen dealing with that [09:45] smspillaz: sorry, that wasn't a gnome-terminal menu, but the window manager's context menu [09:45] vish: ah, so the effect of the bug is now different (having an "ugly" instead of a "pointer" icon), but still needs to be fixed then [09:46] chrisccoulson: sure, queueing [09:46] thanks [09:46] chrisccoulson: thanks for getting this fixed! one more bites the dust \o/ [09:47] pitti - yeah, i just need to fix it in chromium too now (that ships a copy of xdg-utils) [09:47] i'll send the mimeapps.list changes upstream as well [09:47] i had to fix that bit myself [09:48] chrisccoulson: ah, that's different to bug 727372, though? [09:48] Launchpad bug 727372 in firefox (Ubuntu Natty) "FF 4 requires both .desktop and gconf entries for url handlers" [High,In progress] https://launchpad.net/bugs/727372 [09:48] pitti - yeah. i'll land the patch for that in our packaging branch today [09:48] hey chrisccoulson [09:48] how are you? [09:48] chrisccoulson: ah, thanks; they sounded similar [09:48] hi seb128, i'm good thanks, how are you? [09:49] I'm fine thanks === Guest46494 is now known as kklimonda [09:55] Sweetshark: do you know what LibO needs from xulrunner? we need to drop the build dependency [10:05] hi Sweetshark. I've added a proposal on bug 512395, would you mind checking out if it makes sense? [10:05] Launchpad bug 512395 in Ubuntu Translations "Openoffice.org's .desktop files do not contain translation domain info" [High,Triaged] https://launchpad.net/bugs/512395 [10:14] I need to run out for a couple of hours to care for my grandma a bit; I expect to be back around 1400 UTC [10:16] pitti: IIRC mostly for adressbook integration for databases and mailmerge. But there is one when building with --disable-mozilla, "macro security" breaks. [10:19] breaks meaning at least: Tools->Options->Security->Macro Security wont open a dialog and wont allow changes (thus giving any selfrespecting admin a heartattack) === chaotic_ is now known as chaotic === chaotic_ is now known as chaotic [10:31] dpm: makes sense from what little i know about the translation infrastructure. [10:32] Sweetshark, cool. Would you be up for having a session at UDS on LO translations going forward? [10:39] dpm: The only thing I have done wrt l10n is rewriting the build system for it. I dont know much about the actual workflow of translators currently. What I remember is that the use pootle, then translate to LOs internal ressource format, which is then used in the build. [10:40] I would have to get it contact with the people doing that stuff first. [10:42] Sweetshark, ok, let me send you an e-mail and let's have a chat about it when you've got a minute [10:43] Sweetshark, so, without --disable-mozilla, do you know which parts of mozilla that openoffice is using? [10:46] dpm: great, lets do that. [10:50] chrisccoulson: hard to define, there is an option to use a precompiled mozilla and not build you own, but that is a _complete_ mozilla build. [10:52] chrisccoulson: here is one hint: http://opengrok.libreoffice.org/xref/libs-extern/libxmlsec/makefile.mk#40 [10:54] Sweetshark, oh, so is it actually just using nss? [10:55] chrisccoulson: no [10:56] chrisccoulson: thats one place, it also uses it for the addressbook around here http://opengrok.libreoffice.org/xref/libs-core/connectivity/source/drivers/mozab/makefile.mk#34 [10:56] and that stuff also get used in mailmerge IIRC. [11:00] chrisccoulson: what is the root problem with the xulrunner build dep? [11:04] chrisccoulson: according to this: http://openoffice.org/bugzilla/show_bug.cgi?id=17187#c24 [11:04] Error: Could not parse XML returned by OpenOffice.org: HTTP Error 404: Not Found (http://openoffice.org/issues/xml.cgi?id=17187) [11:05] --disable-mozilla means no addressbook, no xmlsec (because of nss), no LDAP [11:09] Sweetshark, xulrunner is basically unsupportable now :( [11:10] also, xulrunner doesn't support LDAP (you need to use the thunderbird SDK for that). are you sure openoffice doesn't switch to another ldap provider for that? [11:13] chrisccoulson: OOo used their own old flaky mozilla precompiles from http://tools.openoffice.org/moz_prebuild/680/ [11:13] Error: Could not parse XML returned by OpenOffice.org: HTTP Error 404: Not Found (http://openoffice.org/issues/xml.cgi?id=680) [11:14] * Sweetshark is gonna ask on #libreoffice [11:14] thanks [11:14] Sweetshark, basically, this is the problem for us: http://mozilla.github.com/process-releases/draft/development_overview/ [11:15] if we keep it in main, we're committed to either maintaing old branches with security fixes, or rolling out new major versions across all stable releases every few weeks (and fixing all of the associated breakage) [11:15] and i don't want to end up spending all of my time doing that ;) [11:15] chrisccoulson: come one! [11:15] lol [11:15] * didrocks hugs chrisccoulson :-) [11:15] i care about firefox, i don't care about xulrunner ;) [11:15] * chrisccoulson hugs didrocks [11:18] chrisccoulson: well ... [11:20] rodrigo_, could you check on bug #754152 [11:20] Launchpad bug 754152 in evolution (Ubuntu) "Evolution unable to open CouchDB address book" [Undecided,New] https://launchpad.net/bugs/754152 [11:21] seb128, yes, I guess it's a duplicate, but checking... [11:21] i always get frightened when i see people talking about couchdb bugs [11:21] i think that they're going to blame me for breaking it again ;) [11:22] loosing xmlsec in LO is not an option IMHO, loosing the Thunderbird integration maybe. For LDAP Id need to check if that still needs xulrunner. [11:23] i'm pretty sure the ldap stuff doesn't require xulrunner, as xulrunner has no LDAP library in it at all [11:24] all the LDAP stuff is in thunderbird, and i don't think openoffice is depending on that is it? [11:25] that would mean to split up the dependency on nss and the one thunderbird/xulrunner etc. -- but I suspect that to be nontrivial otherwise it should have been long done. [11:26] chrisccoulson: http://opengrok.libreoffice.org/xref/libs-core/connectivity/source/drivers/mozab/makefile.mk#61 [11:27] dunno about that libs, do you? [11:27] ah, wait ... [11:28] we are building with SYSTEM_MOZILLA, I guess which disables the mozilla addressbook anyway, because that only work against those old rotten prebuild mozilla binaries. [11:29] so it might be that we really actually only use nss (because the other stuff is broken anyway). [11:29] yeah, i hope we really only use nss :) [11:30] in that case it might be just a bit of configure.in tweakage. [11:30] there are certainly no binary depends between openoffice and xulrunner to keep xulrunner installed on the system [11:30] so, if it really is using it, we'd need to fix that anyway [11:31] and xulrunner is also not on the CD now either [11:31] no guarantees though, LO/OOo is a treacherous beast. [11:31]