[00:00] <asac> all fine?
[00:01] <asac> bimberi: yeah. ap might not be happy with 1:
[00:01] <asac> you could also try 1:~
[00:01] <asac> Venus_Mars: the install location of components dir is not fixed
[00:02] <asac> at best wrap your component into an extension
[00:02] <Venus_Mars> I did it into extension and error console says "Failed to load XPCOM component: /home/nithin/.mozilla/firefox/gu81otyy.default/extensions/oggPusher@xiph.org/components/MyComponent.so"
[00:03] <asac> this device is really cool ;)
[00:03] <asac> just works I'd say
[00:03] <asac> Venus_Mars: i guess it wont load in xul components dir too
[00:04] <Venus_Mars> when I did a nspr logging this was the output http://pastebin.org/46632
[00:04] <asac> your component is probably wrong ;)
[00:04] <Venus_Mars> and the output of ldd -r MyComponent.so is http://pastebin.org/46634
[00:04] <asac> Venus_Mars: there is a missing symbol ;)
[00:05] <Venus_Mars> there must be some problem with compiling.. may be should change compiling flags in Makefile
[00:05] <asac> not sure. if you build with -fPIC and -shared it should probably work
[00:06] <Venus_Mars> I dint get you
[00:06] <Venus_Mars> the makefile is at http://www.it.iitb.ac.in/~nithind/src/Makefile
[00:06] <bimberi> asac: ok, trying 1:~
[00:07] <Venus_Mars> http://developer.mozilla.org/en/docs/XPCOM_Glue
[00:08] <asac> Venus_Mars: id say that its your code ;)
[00:08] <asac> Venus_Mars: yes. you get the right flags with libxul
[00:08] <asac> (XPCOM_Glue)
[00:09] <Venus_Mars> xpcom_glue shows a table which says -L/path/to/sdk/lib -lxpcomglue
[00:09] <Venus_Mars> as linker flags
[00:09] <asac> Venus_Mars: read more carefully ;)
[00:09] <asac> hint: there are two type of glues ;)
[00:10] <Venus_Mars> iam thinking to use the standalone glue. am I wrong?
[00:11] <asac> yes
[00:11] <asac> standalone is for standalone ... i dont think your component has a main method ;)
[00:14] <Venus_Mars> okie but the makefile is not using -lxpcomglue_s -lxpcom -lnspr4
[00:15] <asac> Venus_Mars: it should (pkg-config)
[00:15] <Venus_Mars> echo `pkg-config --cflags libxul`
[00:15] <Venus_Mars> -fshort-wchar -I/usr/include/xulrunner-1.9/stable -I/usr/include/nspr
[00:15] <asac> Venus_Mars: yes, but those are cflags
[00:15] <asac> what you look for here are "linker" flags
[00:16] <asac> Venus_Mars: now that i am looking, i see the problem ;) ... look closer :)
[00:16] <asac> hint: your linker flags are indeed bogus ;)
[00:16] <Venus_Mars> it should be  -L
[00:17] <Venus_Mars> ?
[00:17] <asac> what variable stand for linker flags?
[00:19] <Venus_Mars> --libs?
[00:20] <Venus_Mars> I guess I should replace --cflags with --libs
[00:20] <asac> how could you find out?
[00:20] <Venus_Mars> man pkg-config
[00:22] <asac> one way ;) not sure if that tells you everything :)
[00:23] <asac> man, give it a try ... go
[00:23] <Venus_Mars> am I missing anything more?
[00:24] <asac> cant tell
[00:24] <Venus_Mars> Well I tried, still its not able to load
[00:24] <Venus_Mars> :(
[00:24] <asac> are linker flags correct now?
[00:24] <asac> same issue or something else
[00:25] <Venus_Mars> http://pastebin.org/46639
[00:25] <Venus_Mars> the output of ldd -r MyComponent.so
[00:26] <asac> thats not really meaningful as you cannot even resolve libxul and libxpcom
[00:27] <Venus_Mars> yeah, so what can be done about libxul.so?
[00:27] <asac> looking for compiler warnings would be a good start
[00:27] <asac> Venus_Mars: thats not your problem
[00:27] <asac> you have to look elsewhere
[00:27] <asac> you just cant use ldd -r that way
[00:27] <Venus_Mars> I dint get compiler warnings
[00:28] <Venus_Mars> what should I do ?
[00:29] <asac> whats the error you are getting ;)
[00:29] <asac> did you install all files required?
[00:29] <Venus_Mars> symbol missing
[00:29] <Venus_Mars> No, firefox is failing to load the component
[00:30] <asac> so which files did you install where?
[00:31] <Venus_Mars> I am installing through extension
[00:31] <Venus_Mars> I have created  components dir in my extension rootdir and then copied MyComponent.so in  components
[00:31] <Venus_Mars> and the bundle it into .xpi
[00:32] <Venus_Mars> and then I open it with firefox
[00:33] <Venus_Mars> Am I meaningful?
[00:33] <asac> install  the .xpt as well
[00:34] <Venus_Mars> Yeah I have included IMyComponent.xpt as well
[00:39] <Venus_Mars> well my plugin is accessible www.it.iitb.ac.in/~nithind/firefox-extension
[00:39] <Venus_Mars> I mean the extension
[00:41] <Venus_Mars> still no clue?
[01:36] <bimberi> asac: "Conflicts: j2re1.4 (<< 1:~)" works too.
[01:54] <Venus_Mars> asac: I have figured it out finally
[01:54] <Venus_Mars> there is one forum which suggest for changed in the linking order
[01:54] <Venus_Mars>  $(FILES) $(GECKO_SDK_PATH)/lib/libxpcomglue_s.a $(GECKO_LDFLAGS)
[01:54] <Venus_Mars> this linking order could make the undefined symbols vanish
[01:54] <Venus_Mars> Its a magic
[05:36] <jetsaredim> anyone know why on kubuntu html form fields don't have any borders in firefox 3?
[08:33] <Kamping_Kaiser> hi :)
[08:34] <Kamping_Kaiser> in the $PROJECT.mk file theres a MOZCLIENT_FILE which contains the project version - what is that path relative to?
[08:45] <Kamping_Kaiser> (i'm kgoetz from before, incase anyone wasnt aware :)). i'm still finding it all mildly confusing.
[10:00] <gnomefreak> asac fta_ and/or Jazzva the HTML Validator needs to be discussed as i just got done reading an email about it, please get in touch with me so we can decide if this is something we really want to add.
[10:01] <gnomefreak> well i guess nevermind i didnt see asac's post to it
[10:04] <gnomefreak> asac: when are we adding NM0.7 or are we not going to? As i understand it is stable or will be shortly
[10:12] <Jazzva> asac, I'll be back around 18h... off to the exam.
[10:12] <Jazzva> If all of you decide we don't need htmlvalidator, I'm fine with that :)
[10:13] <Jazzva> see ya later
[10:17] <gnomefreak> personally i dont like the amount of work that would be involved. I thought xulrunner was for plugins not so much extensions, im sure we can get it to work but packaging them separate i feel is better than adding it to build in a package
[10:27] <gnomefreak> [reed]: your last name isnt Friedman, please say no!
[11:08] <gnomefreak> whos here?
[11:51] <gnomefreak> i think we can trash htmlvalidator due to the way it needs to be packaged and it is missing any license and nowher eon any of its upstream pages says anything about a license. I added post on bug report for someone to email upstream <gave email address> so please feel free to email him
[12:23] <fta> Kamping_Kaiser: MOZCLIENT_FILE is relative to the VCS dir, except if you requested a MOZDIR
[12:23] <gnomefreak> rzr: you didnt get the email about mozillsquad due to not having email address on LP page. so i might make a note on the page to include a comment like that
[12:24] <gnomefreak> damn i havent seen that name in ages
[12:24] <gnomefreak> hes been using his other one
[12:25] <gnomefreak> does anyone happent o knwo the file to edit start up processes like bluetooth and trackerd ect...
[12:28] <Kamping_Kaiser> fta, thanks
[12:29] <fta> Kamping_Kaiser, could you paste your .conf ?
[12:30] <Kamping_Kaiser> fta, i havent made any changes yet, just trying to understand the firefox-3.0.conf before i do
[12:31] <Kamping_Kaiser> except when i try and find it in the mozilla-devscripts source package i cant see it :/ only in bzr
[12:32] <fta> yep, i've refactored mozclient in mozilla-devscripts and it's not released yet. it will be in 0.09
[12:32] <fta> asac, i want to release 0.09 ^^ wanna sponsor me ?
[12:34] <Kamping_Kaiser> ah, i see. is firefox-3.0.mk.in the correct file to edit if i'm editing the 0.7 source package?
[12:34] <gnomefreak> fta: can you add sunbird 0.9 to devscripts please. its about to release rc1 and i would like to get it in PPA for testing
[12:35] <fta> gnomefreak, i already have lightning-sunbird
[12:35] <gnomefreak> it will grab rcs?
[12:35] <fta> yes
[12:35] <gnomefreak> oh ok i thought it was just stable
[12:35] <fta> from MOZILLA_1_8_BRANCH
[12:35] <gnomefreak> ah ok
[12:35] <fta> doesn't matter, use the proper CVS TAG
[12:36] <gnomefreak> k
[12:36] <fta> that way, it's the same as fetching a tarball and repacking it
[12:37] <gnomefreak> yep i like devscripts for that reason including nobinonly ;)
[12:39] <Kamping_Kaiser> firefox-3.0.mk.in is quite a bit longer (and more complex) then firefox-3.0.conf :\
[12:40] <fta> you'd better use 0.09 from bzr otherwise you will have to redo everything once it's released
[12:41] <gnomefreak> fta: that was for Kamping_Kaiser right?
[12:41] <fta> gnomefreak, yes
[12:42] <Kamping_Kaiser> fta, hm, ok.
[12:42] <fta> Kamping_Kaiser, it should be similar, except the wrapped lines
[12:43] <Kamping_Kaiser> fta, so i need to customise m-d and instlal the package, then i can use the custom .conf from my new m-d to build a customised FF3 - is that correct?
[12:46] <fta> well, you only used touched the mozclient part of m-d so it should not me needed to build, just to create the tarball
[12:46] <fta> -used
[12:51] <Kamping_Kaiser> !pastebin
[12:53] <Kamping_Kaiser> re paste of my .conf: only one line in it so far - http://paste.ubuntu.com/23495
[12:53] <Kamping_Kaiser> some lucky person is 8 away froma  23456 get
[12:53] <Kamping_Kaiser> *11
[12:55] <fta> Kamping_Kaiser, so you only changed MOZCLIENT_APPNAME right ?
[12:55] <Kamping_Kaiser> fta, yes.
[12:55] <Kamping_Kaiser> i /just/ made the change (i havent worked on this at all today)
[13:01]  * Kamping_Kaiser changes copyright notice
[13:02] <gnomefreak> can i use mozdate and mozbranch?
[13:03] <fta> yes
[13:04] <fta> gnomefreak, what do you want to do ?
[13:04] <gnomefreak> for sunbird 0.9
[13:04] <fta> use a tag
[13:04] <fta> not a date
[13:05] <gnomefreak> ok and i still neeed branch right?
[13:05] <fta> yes
[13:05] <fta> in fact no, it's in the project file by default
[13:06] <fta> but there's no sunbird 0.9 tag yet
[13:06] <gnomefreak> so no way to grab it?
[13:06] <gnomefreak> i miss the old readme file
[13:06] <fta> http://paste.ubuntu.com/23501/
[13:07] <gnomefreak> so i guess im waiting
[13:07] <fta> was it announced ?
[13:07] <gnomefreak> i got an announcement that rc1 will be released in a day or 3
[13:08] <gnomefreak> i dont remember exact date. but string freeze is now
[13:08] <gnomefreak> one of the 100 mozilla mailing lists im on
[13:08] <gnomefreak> i want to say it was dev-app mailing list
[13:09] <gnomefreak> does it have to be announced if its on ftp.mozilla?
[13:10] <fta> oh, i saw the string freeze too but it doesn't mean the tag is there
[13:10] <fta> yes, i guess the sunbird team will blog about it
[13:10] <fta> is it just me or grep -i is broken in intrepid ?
[13:12] <fta> http://paste.ubuntu.com/23503/
[13:13] <Kamping_Kaiser> fta, i've created a new .conf file (in src/mozclient/), do i need to create a new .mk.in?
[13:13] <fta> yes, and edit the Makefile accordingly
[13:14] <fta> the .mk file is only useful if you want a get-orig-source in your debian/rules
[13:15] <gnomefreak> fta: works here gnomefreak@Development:~$ echo FOOBAR | grep -i FOO
[13:15] <gnomefreak> FOOBAR
[13:15] <gnomefreak> atleast that is working afaict
[13:15] <gnomefreak> oh nevermind it should be in lowercase
[13:15] <fta> no
[13:16] <gnomefreak> it should because -i == ignores case
[13:17] <fta> no, it should stay untouched
[13:17] <fta> must
[13:17] <gnomefreak> atleast according to --help
[13:17] <gnomefreak> -i, --ignore-case         ignore case distinctions
[13:17] <fta> yes
[13:17] <gnomefreak> define untouched
[13:18] <gnomefreak> your pastebin Hardy is how it should be right?
[13:18] <fta> but ignore case during the test, but the output is not transformed, just filtered
[13:18] <fta> yes
[13:18] <gnomefreak> well my intrepid works that way
[13:18] <gnomefreak> so i would have to say it works
[13:18] <gnomefreak> fta@cube:~ $ echo FOOBAR | grep -i FOO
[13:18] <gnomefreak> FOOBAR
[13:19] <gnomefreak> that is how it should output it
[13:19] <fta> yes
[13:19] <gnomefreak> intrepid for me
[13:19] <gnomefreak> gnomefreak@Development:~$ echo FOOBAR | grep -i FOO
[13:19] <gnomefreak> FOOBAR
[13:19] <gnomefreak> Host 'Development', running Linux 2.6.26-2-generic - Cpu0: Intel 1681 MHz; Up: 3:48; Users: 4; Load: 0.11; Free: [Mem: 26/248 Mio] [Swap: 605/729 Mio] [/: 23913/38173 Mio]; Vpenis: 36.2 cm;
[13:20] <gnomefreak> fta: you have this version of grep? 2.5.3~dfsg-5ubuntu
[13:20] <gnomefreak> 2.5.3~dfsg-5ubuntu1
[13:20] <fta> yes
[13:20] <gnomefreak> sorry didnt paste the 1
[13:21] <gnomefreak> not sure why yours isnt working
[13:21] <gnomefreak> im guessing you did todays 6 packages but grep wasnt one of them IIRC
[13:21] <gnomefreak> The following NEW packages will be installed: discover discover-data libdiscover2
[13:21] <gnomefreak> The following packages will be upgraded: debhelper debianutils gdebi gdebi-core hotkey-setup xserver-xorg-video-ati xserver-xorg-video-radeon
[13:21] <gnomefreak> those packages?
[13:22] <fta> it's been broken for days
[13:22] <gnomefreak> hmmmm
[13:22] <gnomefreak> it works without -i for you
[13:22] <fta> yes
[13:22] <gnomefreak> do the rest of the flags work?
[13:23] <gnomefreak> or atleast some of them (theres a crap load of flags to use)
[13:23] <fta> yes
[13:23] <fta> fta@ix:~ $ echo FOOBAR | grep -v FOO
[13:23] <fta> fta@ix:~ $ echo FOOBAR | grep -c FOO
[13:23] <fta> 1
[13:23] <gnomefreak> fta: what happens if you use ' around FOO
[13:23] <fta> no difference
[13:23] <fta> i've already tested tons of stuff
[13:24] <fta> i guess it's a locale issue
[13:24] <fta> but i can't see what's wrong
[13:25] <gnomefreak> it shouold affect other flags unless you built your own/hacked it?
[13:26] <fta> i didn't
[13:26] <gnomefreak> i dont see why you would hack or build grep because honestly it already is full of crap
[13:27] <gnomefreak> it works if you use real files?
[13:27] <gnomefreak> oh thanks for reminding me to look stuff up
[13:28]  * gnomefreak didnt know what an ibex is until just now and it looks like ill never see one :(
[13:30] <fta> 66-match_icase.patch
[13:31] <fta> This fixes
[13:31] <fta>     echo Y | LC_ALL=en_US.UTF-8 grep -i '[y]'
[13:31] <fta> The expected output is:
[13:31] <fta>     Y
[13:31] <fta> Without this patch, it works on non UTF-8 environment, but fails on UTF-8
[13:31] <fta> environment.
[13:31] <fta> fta@ix:~$ echo Y | LC_ALL=en_US.UTF-8 grep -i '[y]'
[13:31] <fta> fta@ix:~$
[13:34] <fta> gnomefreak, does it work for you ?
[13:34] <fta> intrepid
[13:34] <gnomefreak> no outpu
[13:34] <gnomefreak> t
[13:35] <fta> are you using utf8 as locale ?
[13:35] <gnomefreak> gnomefreak@Development:~$ echo Y | LC_ALL=en_US.UTF-8 grep -i '[y]'
[13:35] <gnomefreak> gnomefreak@Development:~$
[13:35] <fta> $ echo $LANG
[13:35] <fta> en_US.UTF-8
[13:35] <gnomefreak> lol
[13:35] <gnomefreak> C
[13:36] <gnomefreak> not sure where the hell that came from
[13:36] <gnomefreak> gnomefreak@Development:~$ echo $LANG
[13:36] <gnomefreak> C
[13:37] <fta> ok, so it was expected
[13:37] <fta> fta@ix:~$ echo Y | LC_ALL=C grep -i '[y]'
[13:37] <fta> Y
[13:39] <gnomefreak> i get Y
[13:39] <fta> i guessed ;)
[13:42] <fta> debian bug 387704
[13:43] <gnomefreak> oh i think im C because i have it in .bashrc to get rid of locale errors/warnings in my chroots (i found no other way to apply it only to chroots
[13:44] <fta> i correctly setup locales in chroots too
[13:49] <fta> bug 243717
[15:09] <asac> hi
[15:10] <fta> hi
[15:11] <asac> fta: so 0.09?
[15:11] <fta> asac, it's stable enough for my taste, and already big
[15:12] <fta> i'm ready to close it if you want to sponsor me
[15:12] <Volans> Hi asac :)
[15:12] <asac> fta: typo: "taggued mozilla/client.mk"
[15:13] <fta> where?
[15:13] <armin76> bumbed
[15:13] <asac> oh wait let me pull the latest :)
[15:13] <asac> fta: you still appear to like "uncommitting" :) ... diverged
[15:14] <fta> i sometimes do that in the next second when i realize something is wrong, but not later
[15:15] <asac> fta: well. but i had the uncommitted change ;)
[15:15] <asac> so either i pulled in that second or you waited longer ;)
[15:15] <asac> the changelog is incredible long
[15:16] <asac> not really useful
[15:16] <asac> imo
[15:16] <asac> i think we could make one entry for all the new VARAIBLES
[15:16] <asac> another for all added VCS backend
[15:16] <asac> another for all added projects
[15:17] <fta> as i said, it's big, i wanted this in long ago
[15:17] <asac> yeah. still the content could be compressed to something comprehensible ;)
[15:17] <asac> anyway. can do it that way
[15:18] <asac> fix the typo inthe changelog if you want ;)
[15:18] <asac> oh even two times taggued ;)
[15:18] <fta> fixed
[15:19] <asac> fta: have you tested it?
[15:19] <fta> sure
[15:19] <asac> fta: i think what we want for 0.10 is kind of make check ;)
[15:20] <asac> e.g. that produces several origs, etc.
[15:20] <fta> ?
[15:20] <fta> explain
[15:20] <asac> test-cases
[15:20] <asac> that we can run with make check
[15:20] <fta> oh
[15:20] <armin76> !info libnss3 intrepid
[15:20] <fta> -0d
[15:20] <armin76> !info libnss intrepid
[15:20] <armin76> bah
[15:20] <asac> i dont feel in the position to test each and every combination of project/backend anymore :)
[15:20] <asac> which means its time to do automated tests
[15:21] <asac> !info libnss3-0d
[15:21] <asac> !info libnss3-1d
[15:21] <armin76> fta: bumb!
[15:22] <armin76> when are you guys going to fix the ppc segfault? :P
[15:23] <asac> armin76: we rely on contributions for that
[15:23] <armin76> just make it compile with -O1
[15:24] <armin76> looks like its a bug on glibc or something, because i've been told that guile and glibc-2.8 do the same on ppc
[15:32] <asac> fta: you think you can come up with a basic script that gets latest xpi for a given AMOID from addons.mozilla.org before your leave?
[15:32] <asac> that would help us a lot to setup new upstream branches manually i guess
[15:32] <fta> possible
[15:32] <asac> and can certainly be reused in the final auto syncher
[15:33] <asac> fta: so whats your idea about how to code this? parse HTML?
[15:33] <Volans> asac, fta can I help in some way for that?
[15:33] <asac> Volans: yeah. what we need is a simple script that gets latest .xpi for a given AMOID from addons.mozilla.org
[15:33] <asac> i think one way would be to parse the website
[15:34] <asac> but i think there is a smarter way
[15:34] <Volans> yes for sure... I will try
[15:34] <asac> as firefox queries addons.mozilla.org as well, there probably exists an xml interface
[15:36] <fta> http://paste.ubuntu.com/23524/ better ?
[15:36] <Volans> asac: I have found that if you do:
[15:36] <asac> http://lxr.mozilla.org/seamonkey/source/toolkit/mozapps/extensions/src/nsAddonRepository.js
[15:36] <Volans>  wget https://addons.mozilla.org/it/firefox/downloads/file/AMOID/
[15:36] <asac> thats the code that queries addons
[15:36] <Volans> it download the last public version
[15:36] <Volans> after some redirect
[15:37] <asac> fta: definitly more compressed. though mostly just the file references missing i guess
[15:38] <fta> that doubles the size
[15:38] <fta> people could still read bzr logs or code.lp
[15:38] <asac> yeah. thats ok
[15:38] <asac> i mean i woudl have been fine with the other changelog too. but this is better
[15:39] <asac> fta: i guess you have Vcz-Bzr in control?
[15:39] <fta> yes
[15:39] <asac> maybe add it as reference to copyright too (if not already done)
[15:39] <asac> and README ;)
[15:39] <asac> Volans: why "it" :)
[15:39] <asac> ?
[15:40] <asac> can i replace that with nothing? or "en"?
[15:40] <Volans> sorry... I was visiting with browser and auto localizations o mozilla site ;)
[15:40] <asac> Volans: so what is the genuine URL to use for that?
[15:40] <Volans> let me try
[15:41] <fta> asac, closed & pushed
[15:41] <asac> gbood
[15:42] <asac> doing some basic tests and will upload if those work ;)
[15:42] <fta> i want to change that: http://mxr.mozilla.org/mozilla/source/browser/app/Makefile.in#429
[15:43] <asac> that specific line or that area?
[15:43] <fta> not sure if we want to set SKIP_COPY_XULRUNNER or patch to make a symlink or something else
[15:43] <fta> for xulapps
[15:43] <fta> there's no way to pass SKIP_COPY_XULRUNNER from configure, or at least i don't know of any
[15:44] <asac> fta: what do you want to prevent?
[15:44] <fta> ship a full copy of xulrunner in each xulapp
[15:45] <asac> ok the SKIP_COPY_XULRUNNER
[15:45] <asac> so how is it set atm?
[15:45] <asac> or is it just a "not-used" feature?
[15:45] <fta> this is prism with my new build system: http://paste.ubuntu.com/23529/
[15:47] <Volans> asac: finally  wget  https://addons.mozilla.org/en-US/downloads/file/AMOID/ works, but wget made 3 redirect before download the final file on order to autodetect the destination program (FF, TB, SM) and the name of the latest file to be downloaded
[15:47] <asac> Volans: hm
[15:47] <asac> m
[15:47] <Volans> if you use HTTP instead of HTTPS he made another redirect... I don't know why he continues do https and http request
[15:48] <fta> asac, SKIP_COPY_XULRUNNER is nowhere else, just in this test
[15:48] <asac> fta: ok, so its a left-over
[15:48] <asac> fta: have you tested that setting it actually works?
[15:48] <Volans> what trouble you asac, too much redirect?
[15:49] <fta> no, according to blame, it has been added in one shot with mozilla bug 390361 and the whole block never changed since
[15:49] <fta> they want the xulapp to be independent of xul at runtime
[15:50] <fta> i don't think we want that
[15:50] <asac> yes. i understand that pov
[15:50] <asac> fta: depends
[15:50] <asac> fta: what kind of dependence do we need?
[15:51] <Volans> the final link asac is something like this:  http://releases.mozilla.org/pub/mozilla.org/addons/AMOID/FILENAME-APP.xpi
[15:51] <Volans> whit APP = 'fx' for firefox and similar... but you have to find the correct FILENAME! without dosn't work
[15:51] <fta> in our ff3 package, I patched it to set SKIP_COPY_XULRUNNER long ago
[15:51] <fta> that's installer_shouldnt_copy_xulrunner.patch
[15:52] <fta> we will need xulrunner-1.9 as dep, like ff3
[15:53] <asac> fta: ok. but we still copy the stub right?
[15:53] <fta> asac, ^^, or we can let each xulapp decides
[15:53] <fta> yes
[15:54] <asac> i think firefox is somewhat special
[15:55] <Volans> asac: wait I have found it! :D
[15:55] <Volans> for example: http://releases.mozilla.org/pub/mozilla.org/addons/3863/
[15:55] <Volans> open it in browser
[15:55] <Volans> is a watchable folder!
[15:55] <asac> note that I am with you with the point that we want to skip xulrunner copies
[15:56] <asac> Volans: good
[15:56] <asac> thats nice
[15:56] <Volans> and I can assure you that there are only public versions... I have one waiting for approval and it is not in list
[15:57] <Volans> then watching http://releases.mozilla.org/pub/mozilla.org/addons/AMOID/ can be the best choice if you want to avoid automatic redirect from addons.mozilla.org site
[15:59] <asac> Volans: ok you can get the version, by: xpath  -q -e '/RDF/Description/em:version/text()' install.rdf
[15:59] <asac> and if that fails its an attribute of Description:
[15:59] <fta> asac, worse case, i can add a symlink in an "else" block.. it will then be in ff3 too but that shouldn't hurt
[16:00] <asac> /RDF/Description/@em:version
[16:00] <fta> hm, no. the symlink will not survive xul updates
[16:01] <Volans> asac: you want to made this check after download the xpi file that we suppose to be the latest?
[16:03] <asac> Volans: no. i want two scripts: 1st. download-latest <AMOID> <TARGET_DIR>
[16:03] <asac> 2nd. get-xpi-info VERSION <XPI_FILE>
[16:03] <Volans> bash script?
[16:03] <asac> get-xpi-info NAME <XPI_FILE>
[16:03] <asac> Volans: whatever ;)
[16:04] <Volans> fta: you are already working on that or I can start from scratch?
[16:04] <fta> i'm already working on that
[16:05] <asac> fta: so can you give us download-latest?
[16:05] <asac> ;)
[16:05] <Volans> I can help you in some way?
[16:05] <fta> not at that point, it's still half in my brain
[16:06] <Volans> and in you files? :)
[16:06] <fta> but if you think i'm too slow, go ahead, i'll drop the ball
[16:06] <asac> fta: its just about the two scripts :)
[16:07] <fta> get-xpi-info is currently something that is in my old shell script
[16:07] <fta> but i wanted a class in the main framework
[16:07] <fta> instead of a shell script
[16:07] <asac> fta: ok. we dont need get-xpi-info now
[16:08] <asac> what i want is basically just download-latest
[16:08] <Volans> fta:  in what language you are writing the framework?
[16:08] <fta> Volans, OO perl
[16:08] <fta> like mozclient
[16:08] <fta> maybe even cleaner
[16:09]  * Volans never coded in perl....
[16:09] <fta> with external plugins in any langage
[16:09] <asac> fta: did you review the config branch?
[16:09] <fta> yes, i had questions but you were offline
[16:09] <asac> fta: still have them in your mind?
[16:10] <fta> hm, nope, i'll have another look later today, after i'm back from sports
[16:10] <asac> fta: so what interface would a download-latest plugin have? isnt download-latest-amo <BRANCHID> <TARGET_DIR) ok?
[16:11] <asac> or would it rather be generic download <CONFIG_DIR>
[16:11] <asac> maybe with <VERSION> == LATEST if ommitted
[16:12] <Volans> asac: where this framework should run? (server based, your pc, etc...)
[16:12] <asac> Volans: we have a semi-spec for that
[16:13] <asac> Volans: it should not matter where its run in the end
[16:13] <asac> its just that it need permissions to proper branches
[16:13] <asac> https://wiki.ubuntu.com/MozillaTeam/Firefox3Extensions/LargeScaleMaintenance
[16:14]  * Volans looks
[16:14] <fta> Volans, by framework, i mean a set of classes used by all the binaries, to parse conf files, do amo parsing, do xpi parsing, do branch management, etc
[16:14] <asac> Volans: the current focus is on the "Auto Updater"
[16:15] <asac> which consists of two parts: "auto upstream branch synchs"
[16:15] <asac> and "auto branch merging"
[16:15] <Volans> ok
[16:15] <asac> we could talk about both as "tasks"
[16:15] <asac> for the syncher ;)
[16:15] <asac> aka BOT
[16:16] <asac> so we have a configuration format proposal here:
[16:16] <asac> https://code.edge.launchpad.net/~mozillateam/firefox-extensions/BOT.TASKS
[16:16] <asac> thats the configuration tree we would give the bot
[16:16] <asac> (just a first pitch)
[16:17] <asac> the idea is that you can run the bot with a task config: bot upstream/amo/firefox-3.0.HEAD
[16:17] <asac> err firebug.HEAD :)
[16:18] <asac> and the bot would know about all the magic to sync the upstream branch
[16:18] <asac> by loading plugins and so on
[16:19] <asac> the configuration is a tree configuration. the bot would first load upstream/config, then load upstream/amo/config and then would load upstream/amo/firebug.HEAD config to get the complete configuration.
[16:20] <asac> just look at the files in the BOT.TASKS branch
[16:21] <Volans> I understand... only a question: why not to do a separate bot only for sync with AMO xpis that run on cron and alert you and all the people needed when a new upstream version is out?
[16:22] <asac> Volans: we dont want to alert everyone that new upstream version is out, we want to alter the right people that there is a new merge ready to be signed off / or that needs manual interaction
[16:22] <Volans> then all the BOT will run on a cron basis...
[16:22] <Volans> s/all/the whole/
[16:24] <asac> Volans: yes.
[16:24] <asac> the bot would basically run over the config tree and process each config file (which represents a task)
[16:25] <Volans> you can made a choice that I think is no very restrictive... insert the basename of the xpi in the config file
[16:25] <asac> Volans: why?
[16:25] <asac> we just need AMOID
[16:25] <asac> Volans: look at the config file . there already is a name for the upstream branch.
[16:26] <Volans> when I upload a new version of my extension on AMO site... the site will change the name automatically based on extension name and version adding also a APPLICATION suffix at the end, maybe can help for finding the latest version
[16:29] <asac> Volans: if its required we can add it. just hoped there is something that doesnt require us to intepret the version
[16:30] <fta> leaving for 2h~3h. please write down your ideas & concerns
[16:30] <fta> c u
[16:31] <Volans> asac: problem... I have found a AMOID not present in oregon state archive...
[16:31] <Volans> damnit
[16:31]  * Volans doing random tests to verify if it works in every case
[16:33] <Volans> sorry.... my fault I have used fileid instead of AMOID... works! :D
[16:33] <asac> Volans: i have written a protoype bot that reads the config hierarchy
[16:33] <asac> https://code.edge.launchpad.net/~mozillateam/firefox-extensions/BOT
[16:34] <asac> runtask is a shell script that parses the config as specified
[16:34] <asac> and runs the task plugin found for the config
[16:34] <asac> e.g. in the example its "UPSTREAM"
[16:35] <Volans> yeah I see
[16:35] <asac> so you run runtask upstream/amo/flashblock.ubuntu.HEAD
[16:35] <asac> to test ;)
[16:37] <asac> should probably be AMO_UPSTREAM ;)
[16:38] <asac> or UPSTREAM will exists and uses GET_LATEST_AMO + UPDATE_BZR_TREE
[16:38] <asac> :)
[16:38] <asac> most likely thats the right idea ;)
[16:39] <Volans> ok, I'm reading the sh script... you write bash with different style
[16:39] <asac> most likely. never really tried to look for "clean" scripting ;)
[16:40] <Volans> no is clear... only tests... I use [] in if and other tests... you the "test" statement
[16:40] <asac> yeah
[16:41] <asac> Volans: i am always confused by what is available in which sh-dialect
[16:41] <asac> so i use what appears to be "most" primitive ;)
[16:41] <asac> without actually knowing anything ;)
[16:44] <Volans> correct... I use always bash not sh... but just for "habitude!
[16:44] <Volans> "
[16:44] <asac> so runalltasks would basically just be: "find -type f /path/to/BOT.TASKS/ | grep -v config$"
[16:44] <asac> err flip arguments of find obviously :)
[16:45] <Volans> or better making a config file with all the BOT.TASK files and parse it line per line
[16:45] <asac> find -type f /path/to/BOT.TASKS/ | grep -v config$ | xargs runtask
[16:45] <Venus_Mars> hi Volans
[16:45] <asac> flipped again
[16:45] <Volans> hi Venus_Mars
[16:46] <Volans> asac: in order to test it I have to download also a task tree from the previous branch?
[16:48] <Volans> (~mozillateam/firefox-extensions/BOT.TASKS)
[16:48] <asac> Volans: no that happens automatically iirc
[16:48] <asac> yes: bzr branch $ENV_TASKS_BRANCH $tasksconfigdir
[17:06] <Volans> asac: sorry, probably I don't have understand something... I run:  ./runtask upstream/amo/config
[17:07] <asac> Venus_Mars: btw, i found a better example for you :)
[17:07] <Volans> or  ./runtask upstream/amo/flashblock.ubuntu.HEAD with different results
[17:07] <asac> http://developer.mozilla.org/en/docs/Creating_XPCOM_Components:Using_XPCOM_Utilities_to_Make_Things_Easier
[17:07] <asac> just get the webblock.cpp
[17:07] <asac> and build it like:
[17:08] <asac> gcc -o /usr/lib/xulrunner-1.9/components/libnssample.so -shared -fPIC nsSample.cpp `pkg-config --cflags --libs libxul`
[17:08] <asac> (with sudo ;))
[17:08] <asac> well nsSample.cpp is webblock.cpp ;)
[17:08] <asac> weblock2.cpp ;)
[17:08] <asac> hehe
[17:09] <Venus_Mars> Yeah I got this link yesterday but was almost exhausted...
[17:09] <Venus_Mars> Yeah today I may try with more experiments
[17:09] <asac> Venus_Mars: just try that ;)
[17:09] <asac> then you can use:
[17:09] <asac> sh /usr/lib/xulrunner-1.9/run-mozilla.sh /usr/bin/xpcshell-1.9
[17:09] <asac> js> Components.classes['@company.com/sample'].createInstance()
[17:09] <asac> [xpconnect wrapped nsISupports]
[17:09] <asac> ;)
[17:10] <Venus_Mars> yeah sure....
[17:11] <asac> after that, stop reinventing the wheel and code you xpcom components in a proper build system where you dont have to fight things you dont want to fight
[17:11] <Venus_Mars> heh
[17:11] <Venus_Mars> yeah
[19:05] <Venus_Mars> asac: want some examples where an extension can you upload files
[19:19] <Volans> Venus_Mars: the site where it can upload or what?
[19:20] <Venus_Mars> upload files to a site
[19:21] <Volans> I don't have undestand you question Venus_Mars.... what kind of examples you want?
[19:21] <Venus_Mars> a javascript which ll upload a local file to a remote system
[19:22] <Volans> you don't have already found such similar extensions?
[19:22] <Volans> (iirc in the past talk we had)
[19:23] <Venus_Mars> well, their code involves lot of classes...simply untraceable
[19:40]  * Volans dinner time... come back later
[19:56] <Jazzva> does anyone know if gnomefreak set the membership for ubuntu-mozillasquad to auto-expire one month after joining the team?
[19:59] <[reed]> do members of MozillaTeam _have to_ get e-mail for everybody that joins ubuntu-mozillasquad?
[19:59] <[reed]> :(
[20:10] <armin76> haha
[20:14] <armin76> asac: i wonder if the sparc segfault could be related to some -O flag...-O3?
[20:14] <armin76> s/segfault/bus error
[20:20] <armin76> nope
[20:21] <armin76> i've rebuilt it using mozilla's default and it doesn't give bus error here
[20:28] <Jazzva> [reed], Mozilla Team is the owner of ubuntu-mozillasquad so far... So, I suppose the answer is yes. Maybe it can be turned off somewhere, or something. :)
[22:29]  * Volans back.... :)