[00:14] asac, does it build ? my 1st try didn't went very far [00:16] fta: it built [00:16] and the unittests work [00:16] but its not the browser [00:16] just the infrastructure [00:16] the whole chrome/ and browser/ directory isnt build from what i saw [00:18] i probably didn't use the right sources then [00:19] are you packaging google chrome? [00:19] seems it's too early to do that [00:20] fta: the latest fixes landed in the last checkins [00:20] maybe run a svn up first [00:21] kgoetz: i think it will take quite some time untl we are close to the browser [00:21] the build dependencies dont even include any xlib ;) [00:21] so all the UI is most likey non-existing ;) [00:21] ah, hehe. this could be a long term project *g* [00:22] but maybe its all there and just excluded from build [00:22] * kgoetz wonders if this channel will become #ubuntu-webbrowsermanglers [00:22] but most likely behind the walls at google ;) [00:23] kgoetz: why not ;) ... approaching a broader audience gives more flexibility to bundle forces where they suite well ;) [00:23] asac: hehe, expanding your empire, one .deb at a time. [00:25] * kgoetz heads off to work. catch you in 90 minutes all :) [00:25] well i wil be in bed hopefully [00:27] sleep well then, catch you another day [09:48] Hi there. [09:48] I am updating a package that was depending on libxul previously and now depends on xulrunner-1.9 [09:49] The build went fine by the remplacement of libxul-dev with xulrunner-1.9-dev [09:49] but when I launch it, I get an error of a missing libxul.so [09:49] Of course I have xulrunner-1.9 installed [09:50] (which provides the.so in question) [09:50] but it is not yelling if I install xulrunner-1.9-dev [09:50] any idea what might be wrong ? [09:51] on a side note I have noticed with lintian that a rpath problem might be found there... [09:51] do you think it is the cause of my problem (i am not very handy with rpath) [10:06] huats: most likely it requires a bit of porting [10:11] asac: ok [10:12] huats: most likely your application uses pkg-config --cflags --libs libxul [10:12] asac: AFAIK the application has been ported to xulrunner-1.9... It is really the packaging here... [10:12] you need to use libxul-embedding-unstable [10:12] ok [10:12] huats: if its ported then you probably need to change a configure flag [10:12] ok I'll have a look then [10:13] asac: the thing is the problem is not at build time but at the execution time... which is why I am suprised [10:18] huats: i am not surprised [10:19] thats a wrong configure flag [10:19] ;) [10:19] :) [10:19] asac: ok [10:19] :) [10:19] I'll check then [10:19] asac: thanks [10:25] huats: remember that some apps claim to have been ported to 1.9, but they are not [10:25] asac: ok [10:25] :) [10:25] huats: the reason is that some project still dont understand the difference [10:25] asac: I'll try to understand that... :) [10:26] of compiling a plugin (which is basically what happens when you just build it like 1.8) [10:26] and compiling and starting a standalone application [10:26] asac: I think in that case it is the case since it works fine when I install xulrunner-1.9-dev [10:26] which is different now in 1.9 [10:26] lots of project claimed to support 1.9, but it didnt even work ;) [10:26] asac: I'll remember that :) [10:26] huats: it works fine? [10:27] asac: it can be executed yep [10:27] but not if I remove the xulrunner-1.9-dev (and thus simply have the xulrunner-1.9) [10:27] that is what suprise me [10:28] thus I think the app have been really ported to 1.9 :) [10:29] sounds wierd [10:29] that was my guess too :) [10:30] (and indeed with a simple ldd I can see that with the xulrunner-1.9-dev installed it founds libxul.so and libxpcom.so while not if I remove xulrunner-1.9) [10:31] I have tried to change the configure flag "--with-gecko=xulrunner" to "--with-gecko=xulrunner-1.9" but same pb [10:33] huats: look in configure.in/ac [10:34] do they use PKG_CHECK_MODULES( .... libxul) ? [10:34] nope [10:34] asac: the only reference to xul in configure.ac is [10:34] huats: so how do they get the build flags for xulrunner? [10:34] LIBXUL_INIT [10:34] LIBXUL_DEFINES [10:34] huats: ok. that means they have a .m4 file [10:35] which does that [10:35] search for it ;) [10:35] maybe libxul.m4 [10:35] asac: I was looking for it already :) [10:35] indeed [10:35] it is m4/libxul.m4 [10:35] huats: then look how that crappy thing does it ;) [10:36] most likely it doesnt use PKG_CHECK_MODULES; but attempts to do stuff manually ;) [10:36] but better check [10:36] standalone applications need: libxul-embedding ... or libxul-embedding-unstable if you need more symbols [10:37] ok [10:37] (pkg-config --config libxul-embedding) [10:37] err [10:37] --cflags ;) [10:37] and --libs [10:37] there is no CHECK_MODULES reference in it [10:37] simple CHECK_EXISTS [10:37] as usual ... this file most likely comes directly from stoneage [10:37] huats: ok. what CFLAGS do they define? [10:38] that is what I am looking [10:38] I need to read it [10:38] try to search for XPCOM_GLUE [10:38] if that string isnt there and they do it manually then its not properly ported to the standalone approach [10:38] since they mention both (libxul and libxul-embedding with a test case) [10:38] ok [10:38] huats: ok [10:38] looking [10:39] libxul-embedding is what is wanted [10:39] pkg-config --cflags libxul-embedding will give you -DXPCOM_GLUE [10:39] and other things [10:39] but that is the important part [10:40] no reference of XPCOM_GLUE in the .m4 file but there are such reference in the Makefile.{am,in} [10:42] Apparently I can give a parameter to LIBXUL_INIT to use libxul-embedding instead of libxul (I'll try that right now) [10:47] huats: yes. do that [10:47] withoug libxul-embedding it wont work [10:47] asac: ok [10:48] but the questions remain why it work with xulrunner-1.9-dev :) [10:49] not sure [10:49] don't worry [10:49] but i wouldnt waste time looking inot that [10:49] sure [10:49] unless we know they use libxul-embedding [10:49] sure :) [10:49] thanks a lot so much :) [11:20] asac: stupid question : in the control file is it better that I build-depend on xulrunner-dev or xulrunner-1.9-dev ? [11:21] so far I have used xulrunner-dev (since it is an empty package that depend on xulrunner-1.9-dev) [11:21] but I'd like to be sure... [11:21] and regarding the problem I am facing, I am working on that :) [12:06] huats: if its a merge from debian keep xulrunner-dev [12:06] otherwise use xulrunner-1.9-dev [12:06] ok [12:06] asac: thanks === Bambi_BOFH is now known as Kamping_Kaiser === vk5foss is now known as kgoetz [13:30] asac, yesterday, i finally had chrome fully built. indeed it's incomplete, just reftests for now. my previous attempt was with a direct svn checkout (~666M), about the same thing that we import in lp, it's far from complete. the sync using gclient brings 2.5G of junk, tons of ms binaries (exe, dlls) in the 3rd party dir but it's "buildable" [13:31] i think i'll wait for it to move a bit further before i touch those scons file to see make the browser part build [13:33] fta: we will create a chromium-project product [13:33] with all the bitsw we need [13:35] we need equiv of the --system-foo of mozilla, we already have most of the 3rd parties [13:35] not sure about webkit though, they use a tag, slightly patched, don't know what we have === `6og is now known as Kamping_Kaiser [13:38] seems there are also bin only files [13:38] ex gears [13:38] "The binaries in this directory were an official build for chrome" [13:43] fta: from what i can tell we need "base" "v8" and "webkit" and "build" and chrome [13:43] (out of my head) [13:43] "net" [13:43] asac: you are already trying to compile google chrome for Ubuntu? :) [13:44] Volans: it builds ... but is useless [13:44] no GUI? [13:44] Volans: there is no gui code for linux in the "Hammer" build apparenlt [13:45] they dont even have a single x-related library in build dependencies [13:45] fta: there are a bunch of bin-only files [13:45] i know [13:46] so they are far from an alpha version... [13:46] imo the google folks confused svn with "storage" ;) [13:46] maybe because they have this "we want all data syndrom" ;) [13:46] moreover I know that there are many problems with 64bit [13:46] they are far far away [13:46] there's even a full build of cygwin [13:46] 64bit isnt going to fly [13:46] and unless that works i wont even consider a alpha a real alpha [13:47] fta: i think we need non of those binary stuff on linux [13:47] we dont need third_party tools either i think [13:47] scons is in the ubuntu archive and worked for me [13:47] ill try if a build with the directories above works [13:47] or see what else is missing [13:47] we just need 1 or 2 of those stuff [13:48] fta: what do you think we need? [13:49] reconnect [13:50] 14:47 < fta> we just need 1 or 2 of those stuff [13:50] i don't remember which ones, i pointed out 1 or 2 yesterday. i need to recheck but it doesn't really matter for now, too early to spend time on that imho. [13:50] 14:48 < asac> fta: what do you think we need? [13:50] 14:48 < asac> if WebKit is the webkit code then maybe yes, [13:50] 14:48 < asac> but we probably want to try to use src/webkit bindings with our webkit ;) [13:50] 14:49 < asac> wtf is modp_b64 ;) [13:50] 14:49 < asac> sounds like a bomber :-D [13:50] fta: well. we could make a v8 package ;) [13:50] chromium-v8 [13:51] i feel reluctant to put all that in one big piece of orig ;) [13:51] agreed [13:52] if we split; we'll have to patch those scons files.. brrrr [13:52] well. i have to learn scons i guess ;) [13:53] but agree ... i feel like "why do i need to learn this again?" [13:53] doesnt look to difficult. but getting the best-practices right might not be trivial [13:57] reco [13:58] fta: when i run scons in v8 [13:58] it starts like http://paste.ubuntu.com/43623/ [13:59] (dont be confused ... its src/v8/src/third_party [13:59] not src/third_party ;) [13:59] lets see what comes out of building v8 [13:59] ./third_party [13:59] ./base/third_party [13:59] ./chrome/third_party [13:59] ./chrome/Hammer/third_party [13:59] ./chrome/Hammer/base/third_party [13:59] ./googleurl/third_party [13:59] ./v8/src/third_party [13:59] most likely someone has opened an ITP in debian [14:00] lets see [14:00] fta: ye [14:00] ss [14:00] just running scons in v8 appears to work ;) [14:00] it produces a libv8.a ;) [14:01] probably but running scons in chrome will want to rebuild v8, that's what i meant by patching if we split [14:01] ah [14:03] fta: debian bug #497701 [14:03] Debian bug 497701 in libv8 "RFP: libv8 -- Google V8 JavaScript Engine" [Wishlist,Open] http://bugs.debian.org/497701 === asac_ is now known as asac [14:04] Version: 20080903-1 yeck [14:05] fta: ok. want to take that ITP? [14:05] ;) [14:05] Binary: libv8-dev, libv8b1.. libv8b1 ?? [14:07] asac, not sure. i want to work on the full thing but it seems like a lot of work = lot of time and as i said, i'm still busy with work [14:07] but we'd better do something now if we think it's not going in the direction we want it to go [14:08] Build-Depends: debhelper (>= 5) .. not even python or scons ?? [14:08] fta_: that package of that guy? [14:09] yes [14:09] havent looked [14:09] i want branches for all this [14:09] does it produce a shared lib? [14:10] yes: http://paste.ubuntu.com/43630/ [14:10] a lib without include/-dev packages... [14:11] and with a non-existing shlibs version ;) [14:13] +++ libv8-20080903/debian/cron.d.ex [14:13] @@ -0,0 +1,4 @@ [14:13] +# [14:13] +# Regular cron jobs for the libv8 package [14:13] +# [14:13] +0 4 * * * root libv8_maintenance [14:13] ????????? [14:13] oh, .ex [14:13] most likely not a reasonable packaging ,) [14:14] jcastro: hurry. wake up [14:14] nope === fta_ is now known as fta [14:14] jcastro: we need to setup a bunch of more chromium projects ;) [14:14] jcastro: and you got stuck after https://edge.launchpad.net/chromium-project ;) [14:15] (which should become a super project) [14:15] https://edge.launchpad.net/chromium-v8 [14:15] https://edge.launchpad.net/chromium-net [14:15] yep yet [14:15] https://edge.launchpad.net/chromium-base [14:15] waiting for the super project to be approved [14:15] https://edge.launchpad.net/chromium-webkit [14:16] http://v8.googlecode.com/svn/trunk [14:16] http://src.chromium.org/svn/trunk/src/webkit [14:17] all a bit messy i have to admit ;) [14:17] jcastro: ok. lets wait then [14:18] asac, create a page on the wiki so we can write down ideas [14:20] what's happening to my motu application ? still stuck ? [14:21] refused ? [14:23] fta: no ... not all votes have been casted [14:24] fta: dholbach is on it ... kicking the other council members ;) [14:24] i see that one motu has been approved this morning, so it's not blocked for everyone :P [14:28] meeting [15:10] still frozen for alpha5 ... damn [15:53] I was forwarded here from ubuntu-devel with this question: [15:53] I have a question, I'm testing intrepid right now and when I open something I downloaded via the firefox download-dialog it does not respect my 'Open With' settings. Why isn't gnome-open or something similar used for this (I understand this opens it with the preferred operation)? [15:59] thvdburgt: where did you configure the open with ..? [16:03] Properties of the file. For instance. I use totem with .ogv files, but when I open it via FF's download-dialog it opens with banshee-1 . Looking at it now, does FF get the application from Preferences-Prefered Applicaions (Media-tab)? [16:06] highlight asac [16:15] Hi! [16:15] hi [16:16] thvdburgt: go to preferences -> applications [16:16] and change the app you wnat to use [16:16] XioNoX: what i made out of it is now on the branch [16:16] things work quite well [16:16] in order to use the plugin switcher without the full_path option you need the latest xulrunner [16:17] from ~mozillateam [16:17] PPA [16:17] the one in your PPA ? [16:17] ok [16:17] I have it [16:17] otherwise it will only work with full_path on [16:17] (well ... in hardy you need PPA anyway) [16:18] XioNoX: only feature left is that we add a menu entry "search and install ..." [16:18] which will spwawn the pluginfinder dialog [16:18] ? [16:18] We have already an menu entry [16:19] XioNoX: no ... i mean in the plugin alternatives window [16:19] you mean not only for the activ page ? [16:19] the popup that allows you to switch [16:19] should also get a menu entry to search and install ... ;) [16:19] (instead of switching) [16:20] ok [16:20] no news from mconnor ? [16:21] Ill do it different ;) [16:21] ? [16:21] ill open a new bug ... "allow preferred plugin for mime-type" ... to get the xulrunner patch with the prefs in [16:21] ok [16:21] nice [16:21] when that has landed Ill try to get the ability to switch plugins in the preferences -> applications panel [16:22] and when all that is done te only thing left is the plugin finder wizard [16:22] but given that alternatives are then supported, we have a better argument to beg for that [16:23] ok [16:23] good [16:26] so there are things that I can do ? [16:27] XioNoX: you can add that menu entry ;) [16:27] look at how the popups look like in applications pane [16:27] there is a separator and the then there are is an option "open with ..." [16:27] but I don't really understand what does this application menu [16:27] or "choose other..." [16:27] err menu entry [16:28] i opens the plugin finder wizard [16:28] with the mimetype [16:28] of the entry [16:28] ha [16:28] I understand [16:28] do you know which popup i am referring to? [16:28] http://people.ubuntu.com/~asac/screenshots/plugin.alt.png [16:28] the popup you see in that screen [16:28] so the user could chose directly what plugin file manage a mimetype ? [16:29] XioNoX: well. he can do that now with that dialog [16:29] what we want is that he can search for more plugins that serve that mime-type [16:29] so if you just have adobe flash [16:29] you cannot swtich atm [16:29] so we add a menu entry that installs a new one [16:29] so you can switch afterwards [16:30] XioNoX: antoher thing we need is to hardcode plugin -> description mapping [16:31] currently there is "Shockwave Flash 10.0..." for adobe [16:31] and "Shockwave Flash 9.xx r999" for gnash [16:31] i talked to gnash and swfdec guys [16:31] and they said they cannot add something better to description as detection scripts sometimes look at that [16:31] so we need to manually translate those plugins to something readable [16:32] the idea is to look what the filename is and if contains [16:32] libgnash ... its "Gnash, SWF Player" [16:32] ok [16:32] if its libswfdec... its "Swfdec, SWF Player" [16:32] XioNoX: so what we want is to fix the GetDesc function [16:32] to consider that [16:32] ok [16:33] ok ... i am going to sports now ;) ... cu later [16:33] (hopefully alpha5 freeze will be over tonight) [16:33] ok [16:33] cu [16:33] thanks [20:19] http://nion.modprobe.de/blog/archives/651-guid.html [21:21] asac, http://ubuntuzilla.wiki.sourceforge.net/