whiteasac: i was wondering about CVE-2008-4066 for icedove/iceape. Do we need something like https://bugzilla.mozilla.org/attachment.cgi?id=356697 backported? Just wondering, because i couldn't find the original bugnumber (448166) for CVE-2008-4066 (second part of MFSA 2008-43) in the series file01:19
asacwhite: https://bugzilla.mozilla.org/attachment.cgi?id=35669709:05
asacwhich bug number is that?09:05
asacthe patch looks familiar09:06
asacwhite: ok found the bug number in your line ;)09:24
asacwhite: yes, thats why stransky has asked for approval1.8.0.next.... e.g. it will be in next patchset unless we hit regressions09:25
asacwhite: also mozilla bug 316859 (regression) and the original escape patch (so three for this one in total)09:26
ubottuMozilla bug 316859 in XPCOM "undefined symbol in components/libhtmlpars.so" [Critical,Resolved: fixed] http://bugzilla.mozilla.org/show_bug.cgi?id=31685909:27
iaindaltonfta: your Ubuntu 8.10 PPA's firefox-3.1 requires libasound2 1.0.18 but Ubuntu 8.10 only has 1.0.17; is this intentional?  Should users of your PPA install 1.0.18 manually?09:44
=== iaindalton is now known as iain|away
asaciain|away: you sure you are using intrepid lines for PPA?10:04
asacif so, then fta probably added libasound2 to his archive ... which makes it hard to not pull it in10:05
fta2iain|away, asac_:  i didn't do anything special with libasound2, you may be using the wrong apt line12:06
asac_yeah .... i assume so too12:11
ubuntizatehi, can someone see me (hear me)?12:16
ubuntizatethere is someone there?12:16
asac_those folks that just quit12:25
asac_i answered again for nothing in query12:25
asac_oh PPA's are now signed ;)12:34
fta2yep, i µblogged about that yesterday12:36
asac_hmm ... probably too much traffic in my gwibber backlog ;) ... i didnt see that particular entry of yours12:37
fta2i noticed because the lp guys used me as an example to explain the feature to users. so yesterday, my xchat was blinking like a Christmas tree12:40
fta2asac_, http://identi.ca/fta12:41
asac_nice ;)12:48
asac_sigh sound is gone again13:01
asac_hmm ... now everything is busted ... typing this from console now13:03
asac_radeon ... missing symbol :/13:03
* asac_ runs update;upgrade in hope13:04
asac_pulse 0.9.14 coming13:04
asac_could this be what i want?13:05
asac_i doubt this will cure radeon driver though13:05
asac_hah ... X is back ;)13:09
hyperairhi. i'm having issues trying to get a package to link with xulrunner-1.9-dev. it links fine with libxul-dev, but with xulrunner-1.9-dev (linking using flags from mozilla-gtkmozembed-embedding), ld throws a hissy fit about undefined references and whatnot. what should i do to get it to link?13:57
asac_hyperair: yes.13:57
asac_hyperair: if its standalone app use the standalone glue13:57
asac_if its xpcom compnent or plugin use the dependent glue13:58
hyperairasac_: yes i'm using the standalone glue. -lxpcom_glue or something13:58
hyperairit uses gtkmozembed stuff13:58
asac_standalone glue == libxul-embedding(-unstable)13:58
hyperairand all the gtk_moz_embed_* functions are undefined13:58
asac_dependent glue == libxul(-unstable)13:58
hyperairasac_: tried with that, doesn't work eitehr13:58
hyperairlibxul-unstable won't link without -R13:58
hyperairlibxul-embedding-unstable won't link at all13:59
asac_hyperair: use standalone glue ... you also need some code magic13:59
hyperairasac_: what kind?13:59
asac_let me check14:00
asac_hyperair: https://wiki.ubuntu.com/XulrunnerGecko14:00
asac_its a bit outdated but should give you the basics14:00
asac_hyperair: and gives example patches on how various other apps had to be fixed14:00
asac_look at the yelp example for instance14:01
asac_if you are using unstable headers use 1.9.0.* as max version (not 1.9.*)14:01
hyperairyeah i'm using the unstable headers14:02
hyperairi mean the package requires headers from unstable/14:02
asac_thats what i ment yes.14:02
asac_1.9.0.* is your maxversion then14:02
asac_e.g. for GREVersionRange14:03
asac_(or if yoiu link against 1.9.1 use 1.9.1b1 and 1.9.1.* as bounds14:03
hyperairi see14:03
asac_important is the glue initialization code as well as +#include <gtkmozembed_glue.cpp>14:04
asac_the rest is more or less the same14:04
hyperairwhere does gtkmozembed_glue.cpp come from?14:04
hyperairit's also in /usr/include?14:04
hyperairi see14:05
asac_thats also in unstable i think .... as all gtkmozembed is unstable ;)14:05
hyperairi'll give it a go14:05
hyperairmeh. i've spent over 24 hours working on this package. the last thing i'd do is enjoy, lol14:06
asac_heh ... success is in reach ;)14:06
asac_and finally you found where to ask :-P14:06
asac_thats a win14:06
hyperairi sure hope so14:06
hyperairheh yeah14:06
hyperairsome guys at #ubuntu-motu guided me14:07
hyperairi found some blog posts by you, so if all else failed, i was going to start emailing and/or pinging you on IRC =p14:07
asac_i know ... its just that you left so i couldnt answer there14:07
hyperairah i did?14:07
asac_you left -motu channel ... yes :)14:08
hyperairhmm when was that14:08
asac_hyperair: not sure ... some time between you asking and me opening my screen this morning ;)14:08
asac_could be 1 or 6 hours :)14:08
hyperairi was having some connectivity issues. everytime i enter a certain lecture, my wireless goes haywire. almost as if somebody's aireplaying me off the net14:08
hyperairmust be during that time14:08
asac_sure... no problem. i didnt complain ;)14:09
hyperairi figured i was throwing up a lot of spurious disconnect/reconnect/nick collision messages, so i quit14:09
asac_heh. ok14:09
asac_i dont see QUITS/JOINS14:09
hyperairmust be when i went to sleep then14:13
hyperairbut that's over 12 hours ago14:13
hyperairmaybe not14:13
hyperairor maybe i was shifting location14:14
hyperairanyway, what's the difference between xpcomglue and xpcomglue_s?14:14
asac_hyperair: the xpcomglue includes the standalone pieces14:24
hyperairasac_: what standalone pieces?14:25
asac_hyperair: http://mxr.mozilla.org/mozilla-central/source/xpcom/glue/standalone/14:26
hyperairwhat does gtkmozembed startup code look like minus the glue?14:28
hyperairi can't seem to figure out where to put the XPCOMGlue startup stuff14:28
asac_hyperair: do that before you do anything else with xpcom14:29
asac_dont understand your question i guess14:30
asac_hyperair: depends on how its done14:30
asac_look for gtk_moz_embed_push_startup ();14:30
hyperairi can't find it14:30
asac_or gtk_moz_embed_set_comp_path (MOZILLA_HOME);14:30
hyperairwhich is strange14:30
hyperairalso not there14:30
asac_hyperair: maybe your app uses stuff like XRE_* ?14:31
asac_XRE_InitEmbedding or something like that14:31
asac_but that isnt good ... if you use gtk_moz_embed symbols you really should push/pop startup stuff ;)14:32
asac_found it?14:33
hyperairokay, a list of functions starting with gtk_moz_embed which are used... can_go_back, can_go_forward, get_js_status, get_link_message, get_location, go_back, go_forawrd, load_url, new, reload, set_chrome_mask, stop_load14:33
hyperairall prefixed with gtk_moz_embed_14:34
hyperairare any of them relevant?14:34
asac_hyperair: and XRE_ ?14:34
asac_NS_Ini ?14:34
asac_(as a substring)14:35
hyperairNS_InitXPCOM2 (getter_AddRefs (sm), directory, nsnull);14:35
hyperairwould that be it?14:35
hyperairno wait14:35
hyperairthat's in gecko.m414:35
hyperairhow do gnome-panel applets get started?14:36
hyperairi mean this looks like an ELF executable14:36
asac_hyperair: gnome-panel applets get started by gnome-panel ;)14:36
hyperairbut i don't see a main() anywhere14:37
hyperairuh gee, i think i know that much. more like what functions are there?14:37
asac_its a bonobo activation thing i guess14:37
asac_hyperair: i think the NS_InitXPCOM thing is a good point to start14:37
asac_and see in which function thats called14:38
hyperairasac_: it's test code inside m4/gecko.m414:38
hyperairhmm it seems everything begins from init_libs14:39
hyperairvoid init_libs (int argc, char *argv[])14:39
hyperairi guess i'll just toss the xpcom init stuff there and hope it works14:40
asac_hyperair: so you try to fix the testcode now?14:41
asac_hyperair: look at epiphany or yelp gecko.m414:41
hyperairtest code?14:41
asac_that might be a good start14:41
hyperairno the test code doesn't really do much does it14:41
asac_5:38 < hyperair> asac_: it's test code inside m4/gecko.m414:41
hyperairi mean it doesn't fail14:41
asac_probably doesnt run anthing then14:41
asac_(pure luck) ;)14:42
hyperairso if it clears ./configure, i don;'t really have to bother right =p14:42
hyperairi patched a few things14:42
hyperairit was complaining about needing gtk214:42
hyperairbecause some default toolkit macro thing was set to cairo-gtk214:42
asac_hyperair: well. thts just installing a -dev package ;)14:42
hyperairso i replaced strcmp with strstr14:42
asac_hyperair: i would suggest to take a look at latest ephy or yelp gecko.m414:43
asac_maybe you can even just copy that over14:43
asac_might need a few adaptions in main configure.in i guess though ... but well14:43
hyperairi'm packaging this thing, so i'd rather not end up with a humongous patch14:43
asac_hyperair: patch properly, and send upstream14:44
asac_i mean sending them a "more" modern gecko.m4 shouldnt be that bad ;)14:44
hyperairheh yeah i guess14:44
asac_i saw many gecko.m4 and they are all a mess ;)14:44
hyperairthere hasn't been a release in years14:44
hyperairi wonder if i should even bother continuing to package this =\14:44
asac_yeah. that means that the gecko.m4 probably comes from 200014:44
asac_or before14:44
asac_how do you know;)?14:45
hyperairhttp://www.kaskaras.net/vazaar/index.php <-- it says 200714:45
hyperairno wait14:45
asac_gecko.m4 gets copied around14:45
asac_nobody knows where it originates form i think ;)14:45
asac_maybe it was once in the mozilla tree ;)14:46
hyperairwhy don't they ship it?14:46
hyperairthen aclocal can just add it14:46
asac_nowadays there is pkg-config ;)14:46
hyperairgecko.m4 still uses PKG_CHECK_MODULES14:46
asac_yes. only reason why it exists is to provide backward compatibility i think14:47
hyperairso it seems to be something built on top of pkg-config14:47
hyperairah damn14:47
hyperairso i should just convince upstream to toss out his gecko.m4?14:47
asac_hyperair: i think evolution was : gecko.m4 -> manually; then later someone added stuff using pkg-config ... and so on14:47
asac_and folks copied that around on the net14:47
hyperairevolution uses xulrunner?14:47
hyperairi wasn't aware14:48
asac_no ... evolution like "how things became like they are at present" ;)14:48
asac_darwin ;)14:48
hyperairthen it should be put into aclocal-archive or something14:48
hyperairi mean if it's so common and being copied around14:49
asac_its definitly a pain that its floating around14:49
hyperairbut what's more frustrating is why mozilla can't just use proper SONAMEs damnit14:49
hyperairthat would make linking so much easier14:49
hyperairrather than using the stupid glue14:49
asac_hyperair: previously nobody really used the glue ;)14:49
asac_which caused the main issue14:50
hyperairfeels like an extremely cheap hack gone wrong14:50
asac_now we enforce it. which is better ;)14:50
asac_but painful14:50
asac_(like -rpath is not working anymore)14:50
hyperairthere are still apps which link on libxul-dev14:50
asac_its more cross-platform14:50
hyperairparticularly all the java gecko stuff14:50
asac_so i agree to some degree with them ;)14:50
hyperairi don't see gtk and gtkmm folks messing around with glues14:51
hyperairi mean i don't know how they pull it off on windows, but at least they use proper SONAMEs on POSIX systems14:51
asac_no clue. i think windows solution is to ship your own copy or do some registry rumbling ;)14:53
hyperairfugly shit14:54
asac_one thing that one has to accept is that xulrunner is not designed to be a lib, but a runtime ;)14:54
asac_+ and sdk14:54
asac_if one looks from that perspective things become saner imo14:54
asac_the fact that there is gtkmozembed is a different story14:55
asac_and its degree of messyness is understood ... which is why there is work on a new, stable abi for embedding14:55
hyperairoh there is?14:55
asac_hyperair: thats supposed to be shipped on top of xulrunner providing the stability required to solve all the issues in the world ;)14:56
hyperairmm cool14:56
asac_probably will take some time14:56
asac_but definitly result of years of complains combined with webkit mania14:57
asac_they even use libtool ;)14:58
asac_so hopefully gtkmozembed will die soonish ;)14:59
hyperairthen a lot of programs will have to switch15:02
hyperairand those that are abandoned... will have to be dropped out i suppose15:02
asac_depends. if anyone is interested it will be ported i guess15:05
asac_most apps shouldnt be that hard to port imo15:05
hyperairhmm finished patching. let's see if it builds15:28
asac_dont give up ;)15:30
asac_you are really close (i have the feeling)15:30
hyperairokay, i'm having a whole series of compilation issues15:32
hyperairmostly from gtkmozembed_glue.cpp15:32
hyperairam i supposed to #include it or just compile it together?15:32
hyperairasac_: ^15:33
asac_hyperair: just include it15:33
hyperairasac_: only in one file or in multiple files?15:34
hyperairi added it on top of all instances of #include <gtkmozembed.h>15:34
asac_hyperair: thats ok15:35
asac_hyperair: whats the problem?15:36
hyperairIn file included from blah blah blah /usr/include/xulrunner-bla/unstable/nscore.h:117:1: Warning: NS_HIDDEN redefined15:36
hyperairwait those are warnings15:36
hyperairerror expected = , ; asm or __attribute__ before nsISupports15:37
hyperairimplicit declaration of GRE_GetGREPathwhatever.15:38
asac_hyperair: missing an include i guess15:38
hyperairasac_: are there any needed includes before including gtkmozembed_glue.cpp?15:43
=== asac_ is now known as asac
asacinclude nsXPCOMGlue.h15:46
asacbut thats included through glue.cpp implicitly iirc15:47
hyperairaaaaaah i think i know what's wrong now15:51
hyperairthis is a C file15:51
asacyeah ;)15:51
hyperairnot a C++ file15:51
asacthats a good pit-fall ;)15:51
asachyperair: the idea is to rename it15:51
hyperairso can i still #include it?15:51
hyperairuh rename what?15:52
hyperairi mean i can't use C++ code inside a C file right15:52
asacthe c file to cpp ;)15:52
* hyperair gapes15:52
hyperairyou can't be serious15:52
asachyperair: no. but you can use C++ for that individual file usually15:52
hyperairrenaming a file in a patch.... makes the patch twice the size of the original patch15:53
hyperairof the original file*15:53
asacyou can also add a tiny .cpp file that only does the bootup15:53
hyperairgood idea15:53
asace.g. with function startup_standalone_glue15:53
asacor something15:53
hyperairwhat about the other .c files?15:53
asachyperair: you only need the _glue.cpp once15:53
asacand only once15:53
hyperairin only one of the files?15:54
hyperairi'll have to put extern "C" right?15:57
asac#ifdef __cplusplus16:01
asacextern "C" {16:01
asac#endif /* __cplusplus */16:01
asac#ifdef __cplusplus16:02
asac#endif /* __cplusplus */16:02
asturubuntuHi, I need help16:04
asturubuntuI'm a member of Ubuntu Asturian Translators, and I need help about Firefox16:05
asacasturubuntu: are you the one whose email i answered a minute ago?16:09
asacMarcos Alvarez ...?16:09
asturubuntuno.... but I know Marcos Alvarez.16:10
asturubuntuhe's a member of Ubuntu Asturian Translators too.16:10
asacasturubuntu: ah. i updated the bug and marked it as pet-bug (if its about the same issue)16:10
asacbug 30931216:10
ubottuLaunchpad bug 309312 in langpack-o-matic "make po2xpi aware of per-release whitelists (Was: Please update asturian translation)" [Undecided,In progress] https://launchpad.net/bugs/30931216:10
asacso will get fixed soon i guess16:11
asturubuntuThat's me https://launchpad.net/~malditoastur16:11
asturubuntuand this is Marcos https://launchpad.net/~marcos.alvarez.costales16:11
asacasturubuntu: did you test your translations yet?16:12
asturubuntuthanks for the link16:13
asacasturubuntu: you used it already?16:14
asturubuntuWe have tested it16:14
asacwith those instructions? nice16:14
asturubuntuWait a moment (I need to think in english, arf, arf.... I want to say many things but you are more quickly than me... :-$)16:15
asturubuntuActually, we have Firefox in Asturian language. To do that, we have to place some files into /usr/lib/firefox-3.0.5/chrome16:16
asacasturubuntu: go ahead ... write down what you want to ask and so on ;)16:16
asaci have to make a coffee anyway ;)16:16
asaci will answer your questions if you number them when back ;)16:16
asturubuntuarf, arf... arf.. :P16:17
asturubuntulong time ago, we did the translations in Launchpad. (Firefox and Xulrunner templates)16:18
asturubuntuyou can see it here: https://translations.launchpad.net/ubuntu/hardy/+lang/ast16:19
asturubuntuWhen Intrepid was released, we thought that we can use Firefox in our language. But it was no possible.16:20
asturubuntuThen, Marcos Alvarez, tried doing https://wiki.ubuntu.com/MozillaTeam/Testing_your_translation , we have get a   .jar! files.16:22
asturubuntuYou can find it here: http://code.google.com/p/softasturubuntutranslations/  If you go to this site, and download the compressed file named      intrepid_tornes_ast_Firefox_3.0.x_v.0.2.zip  , you can see the files that we must put into /usr/lib/firefox-3.0.5/chrome  and  /usr/lib/xulrunner-  to get Firefox in Asturian language.16:25
asturubuntuWe are happy, but we have to do this routine everytime that Firefox updates16:26
asacasturubuntu: right16:26
asturubuntuThe question is:16:27
asacyou also have to run the second script ... which is a .xpi16:27
asacthat can be installed by all users then16:27
asacasturubuntu: see point 6. of the instructions16:27
asacthat produces a .xpi16:27
asturubuntuI don't know how to do that. (I need to talk with Marcos)... wait a moment16:29
asacno problem. i have a meeting now ... include my nick if your question and i will glance over to answer16:31
asturubuntuI'm talking with marcos right know (he's in his job)16:33
marcos_astHi asturubuntu16:33
marcos_astHi asac16:33
asturubuntuand now is here! :D16:34
asturubuntuhi marcos!16:34
marcos_astI writed asac 1/2 hour ago16:34
asturubuntuI don't understand very well what asac is telling me...16:34
asturubuntuabout the 6th point in https://wiki.ubuntu.com/MozillaTeam/Testing_your_translation16:35
marcos_asti will go to red16:35
asacmarcos_ast: asturubuntu explained to me that you cop ythe jar! files to chrome/ ... with step 6 you should be able to produce proper .xpi's out of them16:36
asacwhich would allow you to ship stuff from websites or anywhere16:36
marcos_astwell, 2 persons work in this, I and Mikel16:37
marcos_astI learn create the xpi from US language16:37
marcos_astDo you need the step 6, asac?16:37
marcos_astdo you like that I check it?16:38
marcos_ast(Now i'm working, but in 4 hours, I can try it)16:38
hyperairasac: IT BUILT OH MY GOD IT BUILT16:39
* hyperair dances around the room16:39
hyperairoh fsck it doesn't run16:40
asachyperair: now check that it doesnt crash16:40
* hyperair deflates16:40
hyperairit crashes16:40
hyperairT_T what could be wrong16:41
asacmarcos_ast: step 6 is important as it does  everything automatically for you16:41
asaci dont need it to be tested16:41
asacjust wanted to point out that its easy for you to distribute the .xpi to asturians until its in officially16:41
marcos_astthen you're working in the update16:42
marcos_astI can see later16:42
marcos_astdo you need something? :)16:42
marcos_astfor it?16:42
ubottuLaunchpad bug 309312 in langpack-o-matic "make po2xpi aware of per-release whitelists (Was: Please update asturian translation)" [Undecided,In progress]16:42
marcos_astwe had a problem with the xpi in the past16:59
asachyperair: how does it crash?16:59
marcos_astbecause it's a bug in po2xpi16:59
marcos_astthat not create fine locale/ast-ES/global/netError.dtd16:59
marcos_astbut us xpi is fine16:59
asacit doesnt?17:00
asacmarcos_ast: i think it does that17:00
asacall other translations have that too afaik17:00
hyperairasac: it's a little hard to debug because there's no main() function17:00
hyperairdon't ask me how it even got compiled into an ELF17:00
hyperairasac: i'm trying to write a small program that loads the said function through dlopen17:01
hyperairbut it's not working very well17:01
marcos_astasac: I will check in the night17:01
marcos_astare you here?17:01
marcos_astwill you here?17:01
marcos_astwill you be here?17:01
marcos_ast;) ok17:02
asacmarcos_ast: i might not be here17:03
asacbut i will read and answer if you are here17:03
asacif you go offline in between just ping me when you are back so i remember to send the answer17:04
marcos_astok ;)17:06
whiteasac: ok thanks, i'll add a note that it comes into the next patchset17:20
whiteasac: i guess we'll need the fix in the next icedove as well? :)17:53
whiteasac: what about CVE-2008-5023 aka MFSA 2008-57? bugnumber 424733, attachement: https://bugzilla.mozilla.org/attachment.cgi?id=34367018:07
whiteasac: also I believe CVE-2008-5510 aka MFSA 2008-67, bugnumber 228856, attachement: https://bugzilla.mozilla.org/attachment.cgi?id=345028 needs to go into the next patchround, right?18:13
whiteasac: and can you confirm that https://bugzilla.mozilla.org/attachment.cgi?id=256470 is the complete patch for bugnumber 367428?18:19
hyperairasac: ping18:24
hyperairhttp://pastebin.com/f5c27793e <-- for some reason, the segfault happens right at the end of the function, after gtk_moz_embed_set_path (xpcomLocation);18:25
whiteasac: i am also unsure about CVE-2008-5019 aka MFSA 2008-53 regarding xulrunner18:25
whiteasac: we are approaching the end :)18:25
whiteasac: do we need a fix for CVE-2008-5504 aka MFSA 2008-62, bugnumber 453526 for xulrunner?18:27
whiteasac: i'll put this in an email, might be easier18:30
whiteasac: mail sent :)18:36
[reed]white: what distro are you representing? Debian?18:40
hyperairasac: okay, it seems i can't call gtk_moz_embed_set_path at all without segfaulting18:41
white[reed]: debian, yes18:43
white[reed]: in this case debian stable (etch)18:44
hyperairasac: solved the issue. it builds and runs fine now =D19:25
hyperairasac: thanks for the help, and could you revu this for me please?  http://revu.ubuntuwire.com/details.py?package=vazaar :D19:28
* Nafallo haz a question19:50
Nafalloin the menu it says "Firefox Web Browser", but "Mozilla Thunderbird Mail/News".19:51
Nafalloasac, fta: ^^19:51

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!