[12:10] <sivang> fabbione: still here?
[12:12] <wasabi> I wonder what about Ubuntu and our system, udev, hal, etc, is unsuitable for the n770. I suspect the memory footprint is going to be higher.
[12:12] <wasabi> Probaby something that should be fixed anyways though.
[12:13] <sivang> anybody has an idea if when giving -o device_node to mkisofs it acts as cdrecord and "burns" the image directly to a CD?
[12:13] <sivang> it's described that way on the man page, but seems rather odd
[12:14] <HrdwrBoB> I highly doubt it
[12:14] <_ion> Unless mkisofs contains some magic for burning a CD instead of just cat'ing the image to a device, my guess would be "no".
[12:15] <wasabi> it might be possible if the driver for the drive supports burning.
[12:15] <wasabi> Which I dunno.
[12:15] <sivang> what why is it mentioend to you can give -o device_node where device_node = "correspond directly to the device name of the optical disc writer.."
[12:15] <mjg59> wasabi: Burning is slightly more complicated than that, I'm afraid
[12:15] <mjg59> Otherwise we could just use dd
[12:15] <sivang> yes, that what I Had in mind
[12:16] <sivang> mjg59: so I guess I do need to create backup images on the HD, and then use cdrecord to toss them to a CDR/CDRW
[12:16] <sivang> :)
[12:16] <Chipzz> hrrrm :)
[12:16] <_ion> sivang: http://johan.kiviniemi.name/tmp/backup-to-cdrw
[12:17] <Chipzz> must say, from all isv software I have seen yet, at least they understand what "dpkg" and "apt-get" are
[12:18] <Chipzz> not entirely, but at least a bit :)
[12:18] <sivang> Chipzz: is it a some kind of a web serveR?
[12:18] <tepsipakki> would it make sense to provide support for ISV:s to help make native deb-packages?
[12:18] <sivang> _ion: right, so cdrecord is the "driver" :)
[12:18] <Chipzz> sivang: plesk is a control panel
[12:18] <tepsipakki> I've just tried to package Matlab...
[12:18] <Chipzz> something like cpanel
[12:19] <Chipzz> tepsipakki: they actually do support debian sarge and ubuntu 5.04
[12:19] <sivang> Chipzz: why use an ISV software for that then?
[12:19] <Chipzz> but they have certain custom packages
[12:19] <mroth> mjg59: hrm, bug 35741 seems to be a duplicate of what I was about to file, but its filed against mesa for some reason
[12:19] <Ubugtu> Malone bug 35741 in mesa "Missing support for 945GM chipset" [Normal,Unconfirmed]  http://launchpad.net/bugs/35741
[12:19] <tepsipakki> Chipzz: yes, they do. Matlab builds on sarge, but they still ship 3CD:s full of stuff =)
[12:20] <Chipzz> sivang: because $dumb_user_wanting_a_webpage doesn't know shit about configuring apache and shouldn't know
[12:20] <Chipzz> sivang: this is for the company I work for, we offer webspace with php/mail/etc against bottom prices ;)
[12:21] <Chipzz> amongst several things
[12:21] <HrdwrBoB> Chipzz: I used to sysadmin for a webhost.. we wrote our own
[12:22] <Chipzz> HrdwrBoB: and how much time and money did you have for it? ;)
[12:23] <Chipzz> we actually have a front-end for plesk to do the actual creation of the domains, and then offer plesk to the users to give them a certain amount of control over their domains
[12:23] <Chipzz> currently installed on rh, but going to migrate to debian/sarge
[12:23] <HrdwrBoB> well you have to pay for plesk
[12:23] <Chipzz> we do have a license allready :)
[12:23] <HrdwrBoB> and I was working for them fulltime.. so the economics were on our side
[12:24] <mroth> mjg59: should I just mark that one as confirmed instead of opening a new one in kernel-source?
[12:24] <mjg59> mroth: What's it currently assigned to?
[12:25] <mroth> mjg59: no one.  status unconfirmed.  https://launchpad.net/distros/ubuntu/+source/mesa/+bug/35741
[12:25] <Ubugtu> Malone bug 35741 in mesa "Missing support for 945GM chipset" [Normal,Unconfirmed]  
[12:26] <mjg59> mroth: No, what package
[12:26] <mjg59> Or just ubuntu in general?
[12:27] <mroth> yeah, its under "mesa (Ubuntu)"
[12:28] <mjg59> Can you reassign it to linux-source-2.6.15?
[12:28] <mroth> mjg59: sure.  keep as unconfirmed, or change status?
[12:30] <mjg59> unconfirmed for now
[12:30] <mjg59> Have you attached lspci? If not, could you?
[12:31] <mroth> mjg59: I'll link to my lspci page for that machine on the LaptopTestingTeam wiki
[12:31] <mroth> since it has both normal, -v and -n
[12:33] <mroth> ok, bug 35741 should be in linux-source now, right Ubugtu?
[12:33] <Ubugtu> Malone bug 35741 in linux-source-2.6.15 "Missing support for 945GM chipset" [Normal,Unconfirmed]  http://launchpad.net/bugs/35741
[12:42] <sladen> mjg59: yup
[12:43] <mjg59> sladen: Do you still have an IBM restore and recovery partition?
[12:43] <sladen> mjg59: yes, but not the boot sector
[12:43] <mjg59> sladen: Can you mount it and stick up an ls -R somewhere?
[12:43] <sladen> mjg59: I can still boot the recovery partition (I think)
[12:45] <sabdfl> night all, sladen, mjg59 
[12:46] <mjg59> sladen: I'm just fixing up os-prober so it can autodetect the correct type of partition
[12:46] <_ion> Quite an interesting plugin for Gimp, "Resynthesizer". http://software.newsforge.com/article.pl?sid=06/03/15/1722200
[12:48] <_ion> http://www.logarithmic.net/pfh/resynthesizer
[12:54] <sladen> mjg59: http://www.paul.sladen.org/ubuntu/recovery/ibm-thinkpad-r52-rescue-recovery.ls-lR.txt.gz   and I've scp'ed up a lot of the small files like 'version.id' for you to `strings`
[12:58] <mjg59> Ta
[01:04] <infinity> mdz: http://us4.samba.org/samba/history/samba-3.0.21c.html <-- Samba changelog... Sorry for not providing it last night in my ping payload.
[01:05] <mdz> infinity:     * Update dhcp.conf files in Debian packaging <-- what's that about?
[01:05] <mdz> why does samba have a dhcp.conf in the packaging?
[01:06] <infinity> mdz: Upstream maintains a debian/ directory which is seperate from the one we (Debian/Ubuntu) maintain in SVN.  Can be (and is often) completely ignored.
[01:06] <mdz> infinity: looks fine, go ahead
[01:06] <infinity> mdz: But it's some trickery for allowing netbios-name-server from DHCP to trickle into samba's wins server setting.
[01:07] <hendry> i am getting an md5 mismatch on Universe. Anyone else?
[01:08] <infinity> mdz: Thanks.
[01:08] <infinity> Oh, hey, lookit that.  Leave it overnight, and eventually bzr will finish checking out all the seeds.
[01:13] <hendry> is there some reason ubuntu doesn't autoupdate like RH's up2date?
[01:14] <tseng> it has a red icon that pops up and tells you when to update
[01:14] <tseng> very similar to the up2date applet
[01:14] <LaserJock> but much better, IMHO
[01:16] <Burgwork> LaserJock, like it is actually up2date? :)
[01:17] <sivang> it's not up2ate for sure :)
[01:17] <sivang> it's update-manager no?
[01:17] <LaserJock> Burgwork: exactly
[01:18] <LaserJock> Burgwork: up2date had a lot to do with my advisor moving from Linux to OSX :(
[01:18] <infinity> mdz: Same question, mod_perl2 2.0.1 -> 2.0.2.  Very minor bugfix release, plus one added feature (exposes three new proxy constants) -- http://perl.apache.org/dist/mod_perl-2.0-current/Changes
[01:22] <mdz> infinity: OK
[01:49] <Kinnison> mdz: I'll get on those vim and aspell bugs tomorrow morning
[02:36] <hendry> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=358071 # i've exp this bug in Dapper too
[02:36] <Ubugtu> Debian bug 358071 in evolution "Subject: SIGABRT error after libc6 update" [Grave,Open]  
[02:36] <hendry> glibc seems b0rked
[02:37] <mjg59> You're sure it's not glib?
[02:37] <hendry> mjg59: ok, perhaps glib. is there an Ubuntu bug open for this too?
[02:38] <crimsun> no, it's glib
[02:38] <mjg59> hendry: It's already being dealt with
[02:38] <mjg59> glib's memory allocator changed and broke some false assumptions 
[02:39] <hendry> mjg59: i want to know when this is fixed. or how i can work around with it. it's already screwed up a days work for me
[02:40] <j^> http://stratusandtheswirl.blogspot.com/2006/03/for-those-who-care-about-gnomew-glib.html
[02:42] <mjg59> hendry: I recommend not depending on a development distribution
[02:43] <Amaranth> mjg59: found something weird in one of my xorg logs when i was messing with xgl
[02:43] <Amaranth> mjg59: for some reason it was trying to use the vga driver
[02:43] <Amaranth> i'll try to make it happen in a loggable way again
[02:44] <mjg59> Amaranth: Weird
[02:44] <Amaranth> yeah, the vga driver doesn't work on ppc
[02:44] <Amaranth> at least not on this one
[02:45] <Amaranth> bbiab
[02:46] <hendry> mjg59: you're right. :/
[03:33] <Lathiat> ooh, launchpad has as exy new little dropdown for changing status info
[03:50] <nictuku> where is the channel for discussing possible bugs? (ati driver)
[03:50] <nictuku> *what
[04:40] <Amaranth> mjg59: would an strace between aiglx glxinfo and xorg glxinfo help with anything? (no direct or indirect acceleration with aiglx)
[04:40] <Amaranth> err, diff of straces between the two
[05:00] <minghua> I need a core developer to review and sponsor an upload for scim
[05:00] <minghua> the debdiff is at http://www.mems.rice.edu/~minghua/ubuntu/scim_1.4.4-1ubuntu5_to_1.4.4-1ubuntu6.debdiff
[05:00] <minghua> it fixes bug #35852
[05:00] <Ubugtu> Malone bug 35852 in scim "scim xinput.d file causes login failure" [Normal,Confirmed]  http://launchpad.net/bugs/35852
[05:00] <minghua> #debian-l10n seems silent now (I posted half an hour ago), anyone willing to help here?
[05:54] <fabbione> minghua: looks sane...
[05:54] <minghua> fabbione: thanks, can you upload it for me?
[05:55] <fabbione> minghua: in a bit.. i want more coffee first
[05:55] <minghua> fabbione: sure, no pressure :-)
[05:57] <fabbione> done
[05:58] <minghua> thanks!
[05:59] <fabbione> minghua: i will send you a bill to your bank.. welcome to pay in 2/3 rates.. 
[06:02] <minghua> fabbione: Hmm... pay to you or to Canonical? :-)
[06:04] <fabbione> minghua: me :)
[06:08] <infinity> minghua: I'll email you fabbione's Australian bank account details.
[06:08] <fabbione> ehheh
[06:09] <minghua> :-)
[06:09] <jkakar> infinity: According to the release notes there are "major ABI/API changes". :)
[06:10] <infinity> jkakar: Then that's almost certainly a "no", unless there are amazingly compelling reasons to upgrade, which you'd then need to discuss with pitti, seb128, and mdz (the first two would be responsible for implementing all the changes required, probably, and the latter would most likely veto it anyway without some really good arguments)
[06:11] <infinity> jkakar: DBUS upgrades are typically very (VERY) painful.
[06:11] <jkakar> infinity: Shucks.  Thanks, I'll see if I can come up with a compelling argument.  Right now there are bugs in the python bindings that make it basically useless...
[06:11] <jkakar> infinity: I suppose there's always monkey patching.
[06:12] <fabbione> jkakar: are the bugs only python related?
[06:12] <fabbione> or there is more than that?
[06:12] <infinity> jkakar: Can the bugs not be fixed in the python bingdings without upgrading the whole mess?
[06:12] <infinity> bindings, even.
[06:12] <jkakar> fabbione: I suspect only python-related... "successful" is spelt "sucessful" in three places in service.py:_method_lookup()
[06:13] <infinity> jkakar: Yeah, those are some pretty minor bugfixes, then.  Pretty please, prepare patches and file bugs (or poke me directly), and we'll get the fixes uploaded.
[06:13] <jkakar> Fixing the spelling in the python files directly fixes my problem.
[06:13] <jkakar> infinity: Awesome!  Thanks, will do.
[06:14] <jkakar> infinity: I'd like to wait to provide the patches until I've made it a bit further into what I'm doing... just in case there are other dumb things like that.  Is there any rush on this, or can I harrass you about sometime in the next week or two?
[06:15] <infinity> jkakar: Small and obviously correct bugfixes are appropriate anytime (before release).
[06:15] <jkakar> infinity: Cool, thanks.
[06:17] <fabbione> ogra: are you still around?
[07:06] <childe> Hi, if I don't line the default recycle bin icon, where should I report the bug?
[07:07] <childe> Or, to which team should I report the bug?
[07:07] <infinity> You're not alone, and I'm not sure that filing another bug will help.
[07:07] <childe> It looks too much like a pistol bullet shell.
[07:07] <infinity> Most people say "battery", but bullet works too. :)
[07:08] <childe> It's sweet to know that I'm not alone on this issue
[07:08] <whiprush> the ubuntu-art list probably.
[07:09] <childe> Thanks
[07:09] <Burgundavia> there is already a bug filed for it
[07:09] <childe> Burgundavia: Could you tell me the bug number?
[07:09] <childe> I want to see what other people say about this
[07:10] <infinity> Most of the complaining has been on IRC and on mailing lists and such.
[07:10] <infinity> But the artists are definitely aware of the dislike for the icon.
[07:10] <infinity> Not sure what they plan to do about it. :)
[07:10] <whiprush> infinity: are you on the server team?
[07:10] <infinity> whiprush: Aye.
[07:10] <childe> Maybe when Dapper release most of the icons will be changed to the orange color ones
[07:10] <Burgundavia> childe: sorry, not off hand. I am not worried, it will be dealt with
[07:11] <childe> Like the "Home Folder" icon.
[07:11] <childe> Burgundavia: It's OK :-)
[07:11] <whiprush> infinity: Kerberos support is much better in dapper than breezy, you guys should mention that in some release notes somewhere.
[07:12] <infinity> whiprush: Feel free to start some ubuntu-server release notes on the wiki for me.  I won't object.
[07:12] <whiprush> It's almost damn near trivial.
[07:12] <infinity> whiprush: If there's anything anyone knows about server-type people, it's that we hate creating documentation. ;)
[07:12] <whiprush> heh
[07:16] <childe> Synaptics feels much faster than before. I don't know why.
[07:18] <whiprush> infinity: has anyone shown up with any interest in LDAP server stuff? 
[07:19] <whiprush> It'd be neat to have someone working on the fedora directory stuff...
[07:32] <infinity> whiprush: Someone (like you?) really should try to make it a priority at the beginning of a release cycle, so someone else (like me?) might get assigned to work on it. :)
[07:35] <robitaille> childe: https://launchpad.net/distros/ubuntu/+source/ubuntu-artwork/+bug/35811
[07:35] <Ubugtu> Malone bug 35811 in ubuntu-artwork "Wastebacket and Battery icon look too similar" [Normal,Confirmed]  
[07:36] <childe> Thank you so much. Only I think this bug should be marked as critical, not normal.
[07:37] <childe> Brend: If you have tried xchat-gnome...
[07:37] <childe> Sorry, wrong channel
[07:45] <lamont> but at least it's ftbfs everywhere...
[07:58] <pitti> Good morning
[07:58] <G0SUB> pitti: hello!
[07:59] <pitti> Hi G0SUB 
[07:59] <_ion> Good time of day
[08:12] <pitti> fabbione: btw, will our ssl-cert changes go to Debian eventually? or is that deemed ubuntu specific?
[08:12] <fabbione> pitti: -> infinity 
[08:19] <infinity> pitti: No. it's not ubuntu-specific, and I need to upload ssl-cert to Debian before I can upload the new apache2.
[08:20] <infinity> pitti: I've just been somewhere between busy and lazy.
[08:20] <pitti> infinity: oh, it's not urgent for me, I was just curious :)
[08:20] <pitti> bzr merge does a wonderful job for me
[08:22] <pitti> Karma:  117884  <- WTH? Yesterday I had some 4000... the last LP update must have changed something...
[08:22] <Burgundavia> pitti: bug work now earns you more
[08:23] <infinity> A LOT more.
[08:23] <Burgundavia> pitti: as does spec work
[08:23] <pitti> whoa, did they just multiply everything by 100? :)
[08:24] <infinity> Oh, that's not that high... I wonder if anyone's over a million yet.
[08:24] <Burgundavia> he is now 4th. mpt has more, with about 1/2 million
[08:26] <pitti> so, that's not a cockroach any more, but closer to a guinea pig? :)
[09:06] <pitti> jdub: ping
[09:12] <Mez> hmm anyone here who wants to answer a question regarding library versioning?
[09:12] <Mez> basically - if I change backwards compatibility - I should increase the major...
[09:12] <Mez> and if not the minor
[09:12] <Mez> but whats the other part for
[09:15] <dholbach> good morning
[09:17] <Mez> morning dholbach
[09:17] <Mez> dholbach- know about best practice for library versioning?
[09:18] <dholbach> hey Mez
[09:18] <dholbach> hey seb128
[09:18] <pitti> hey dholbach, hey seb128 
[09:18] <dholbach> hey pitti
[09:18] <dholbach> Mez: what do you mean? versioning as in packaging?
[09:18] <seb128> hey dholbach pitti
[09:19] <Mez> dholbach: well with my app - we've renamed a function
[09:19] <Mez> so we need to bump the version of the so.
[09:19] <Mez> but I dont know whats the best way to bump it
[09:20] <Mez> see - most things talk about major and minor 
[09:20] <Mez> but there's 3 parts of the version
[09:20] <Mez> x.y.z
[09:20] <dholbach> Mez: http://www.openismus.com/documents/linux/using_libraries/using_libraries.shtml and http://www.openismus.com/documents/linux/building_libraries/building_libraries.shtml might be good reads
[09:21] <pitti> ogra: can you please retroactively add CVE-2006-1335 to the gnome-screensaver changelog for 2.14.0-0ubuntu1?
[09:23] <Treenaks> I'm having kerning problems with my fonts in firefox.. stuff like 'fix' looks distorted, and some other things as well
[09:23] <infinity> Mez: ABI changes change the minor (and the SOVER), API additions (that don't affect the ABI's compat) bump the minor, bugfixes bump the pachlevel.
[09:24] <infinity> Mez: In a nutshell, anyway.
[09:24] <Mez> infinity, well we changed from using keyReleaseEvent to keyPressEvent.
[09:24] <Mez> (changed the name of a function) b
[09:25] <Mez> so I have the choice to either add to the soname
[09:25] <infinity> Mez: Changed the name of the function without providing a wrapper for it?
[09:25] <seb128> Mez: http://www.gnu.org/software/libtool/manual.html#Libtool-versioning
[09:25] <infinity> Mez: If you didn't provide a wrapper (for ABI compat), then you're in SOVER bump land.
[09:25] <Mez> infinity, see - we didnt think of that until now ... but I'm planning on adding a wrapper ;)
[09:26] <Mez> though that might be not worth it
[09:26] <Mez> infinity, SOVER = major right ?
[09:26] <seb128> is anything using your lib?
[09:26] <Mez> seb128, only stuff within the application itself
[09:26] <Mez> nothing outside of it
[09:26] <seb128> you don't have to bother about the SOVER
[09:26] <infinity> Mez: Traditionally, SOVER is major, yes.  (but not necessarily)
[09:26] <seb128> you just update libtool versions and it makes the job for you
[09:26] <infinity> Mez: If it's a static library, then all bets are off.
[09:27] <Mez> seb128, each plugin is a library that links into the main library
[09:27] <Mez> infinity, .so and .la :D
[09:27] <seb128> plugins are not versionned usually
[09:27] <infinity> Mez: Well, s/static/private/ then for the same results.
[09:27] <seb128> they are not a shared lib
[09:27] <infinity> Mez: Shared objects are only truly "shared" if something other than your app will use them.
[09:28] <Mez> seb128, yet the things they link to in the main app ARE
[09:28] <Mez> infinity, well - they may :D
[09:28] <Mez> noone does yet - but we want to let them
[09:28] <seb128> if that's reserved to the app you don't care
[09:28] <Mez> seb128, i care when people get compile errors
[09:28] <lifeless> Mithrandir: killing gss stopped it until I rebooted
[09:28] <Mez>   /opt/kde3/lib64/libkatapult.so.0: undefined reference to `KatapultDisplay::keyReleaseEvent(QKeyEvent*)'
[09:29] <seb128> usually the app is in sync with its own source :p
[09:29] <infinity> seb128: Not if it's KDE...
[09:29] <seb128> lol
[09:29] <seb128> right :)
[09:29] <Mez> seb128, well - people are building external plugins - that arent part of the app "proper" but are seperate packages
[09:30] <seb128> they should build plugins adapted to the app version
[09:30] <seb128> changing the soname will not change that fact
[09:30] <seb128> if the plugin uses a wrong API, soname changed or not it'll not build
[09:31] <fabbione> Makefile.am:24: BUILD_LINUXDOC does not appear in AM_CONDITIONAL
[09:31] <Mez> seb128, true ;) but we want people to be able to use old plugins with newer versions
[09:31] <fabbione> so does anybody remember what provides that to aclocal?
[09:31] <seb128> Mez: so don't rename a function :p
[09:31] <infinity> Mez: Then you need to make the library ABI compatible.
[09:31] <infinity> Mez: Nothing else you can do to "fix" that, really.
[09:32] <Mez> infinity, i know - which is why i'm adding the wrapper as we speak
[09:32] <siretart> hi
[09:34] <Mez> infinity, though now I've done that - I should increase minor - yes?
[09:34] <Mez> or should I increase major just because I made a release without that wrapper
[09:37] <seb128> increment the age and the revision numbers if you added a new function
[09:37] <seb128> hum
[09:37] <Mez> again though - there has been a release without the wrapper ...
[09:38] <seb128> you can declare that one was a bug and ignore it :)
[09:42] <Mez> seb128, lol
[09:42] <Mez> now I just need to make sure it all compiles nicely and stuff
[09:43] <Mez> seb128, the age AND revision numbers?
[09:44] <seb128> you have a new function?
[09:44] <seb128> increment current,revision,age I think
[09:45] <seb128> cf http://www.gnu.org/software/libtool/manual.html#Updating-version-info
[09:45] <seb128> grumpf no
[09:46] <seb128> increment current and age and set revision to 0
[09:50] <pitti> (daily-live as well)
[09:51] <seb128> pitti: if you are borred I can give you some patches to review or bugs or stuff to do :p
[09:51] <pitti> seb128: oh, yes please, give me ALL your gnome bugs!
[09:51] <seb128> oh, really? :)
[09:52] <pitti> hi carlos 
[09:52] <carlos> pitti: hi
[09:52] <carlos> pitti: don't worry, I had to fix a couple of bugs on imports and I already have some export fixes to do
[09:52] <Mez> pitti: o_O are you sure of that ? I'm sure there'sa lot
[09:53] <carlos> pitti: take the time you need
[09:53] <carlos> pitti: I'm generating daily exports now
[09:53] <carlos> on the usual URL
[09:53] <pitti> Mez: ... and what shall I do in the afternoon? :-P
[09:53] <pitti> carlos: cool
[09:54] <Mez> pitti: go to the pub?
[09:54] <pitti> Mez: now, that's a brilliant idea
[09:55] <Mez> pitti: of course - I always have them
[10:00] <jdub> pitti: pong
[10:01] <pitti> jdub: what's the plan wrt. the EOL announcement for warty?
[10:01] <pitti> jdub: I read sth about that on a ML recently, it should be done soon
[10:02] <lifeless> Mithrandir: are you still lookin at that gnome-screensaver bug ?
[10:02] <Mithrandir> lifeless: not as much looking as knowing that it exists.
[10:02] <Mithrandir> lifeless: gss is ogra-land
[10:02] <lifeless> Mithrandir: ok. well it still does :)
[10:04] <dholbach> Diziet: hello! Could we please get a firefox-dbg package? What do you think about it?
[10:06] <Kagou> hi
[10:06] <carlos> is there any backport of cyrus22 for breezy?
[10:08] <dholbach> hey sabdfl
[10:08] <Kagou> seems that the daily iso build process have failed today 
[10:10] <seb128> Kagou: we have noticed but thank you :)
[10:10] <sabdfl> hiya dholbach!
[10:10] <Kagou> seb128, :) always here to help you
[10:10] <Kagou> ;)
[10:10] <seb128> hi sabdfl
[10:11] <sabdfl> so, the drake is shaping up nicely
[10:13] <Treenaks> dholbach: new bug reports doesn't mean new bugs :)
[10:13] <dholbach> Treenaks: i'm perfectly well aware of the difference :-)
[10:14] <dholbach> Treenaks: 1500-2000 desktop bugs mails per week teach you quite well
[10:14] <Treenaks> :)
[10:14] <siretart> infinity: do you still intend to upload a patched l-r-m package (madwifi) or have you already uploaded it?
[10:14] <Kamion> pitti,Kagou,etc.: fixing the CD breakage now
[10:14] <sebest--> dholbach: hello, so around 250 new bug repots per day?!
[10:15] <Lathiat> mails, not necesarily new reports?
[10:15] <dholbach> seb128: is sebest--'s estimation accurate?
[10:15] <seb128> dholbach: somebody told yesterday than launchpad gets a bug every 10 min or something like that
[10:15] <Lathiat> launchpad guys might be able to give us a better idea of new bug nos?
[10:15] <Lathiat> would be interesting
[10:15] <Lathiat> so 144/day ?
[10:15] <Kagou> ok Kamion 
[10:15] <Lathiat> 1000 a week??
[10:16] <infinity> siretart: Still intend to, before I head to bed.
[10:16] <seb128> Lathiat: good math :p
[10:16] <sebest--> would be nice to have a graph to see the trend
[10:16] <Burgundavia> seb128: that was my very rough estimate. I suspect it might be less
[10:17] <siretart> ok
[10:18] <sebest--> the number of open bugs is growing/shrinking or stable?
[10:18] <Lathiat> seb128: thanks, need to put high school to work somewhere ;p
[10:18] <Lathiat> sebest--: i'd bet growing
[10:18] <Burgundavia> I would got with Lathiat on that
[10:19] <fabbione> Kamion: i need a UVF exception for util-macros. It's a pkg in universe that is required to bootstrap some drivers from CVS. We forgot to update it from breezy and it is quite important for me now to pull fixes from CVS.
[10:19] <Kamion> fabbione: mail please with changelog
[10:19] <fabbione> Kamion: we are going from 0.99 to 1.0.1 and the pkg contain only .m4 definitions
[10:20] <sebest--> but maybe a lot of these bug report are not valid any more, but people forget to close them when it's solved for them
[10:34] <jdub> pitti: it's on silbs announcement agenda - perhaps ask her about it?
[10:34] <fabbione> Kamion: you have (an extra) mail from me ;)
[10:40] <Mithrandir> do we support SATA dvd/cd devices?
[10:41] <ajmitch> Mithrandir: afaik, yes, since quite a few laptops have drives which appear as SATA
[10:41] <Mithrandir> 'k, thanks
[10:45] <sladen> Mithrandir: it's weird, even things like the Intelimac have SATA HDD's and PATA optical drives
[10:47] <Treenaks> sladen: yeah, why would you waste a perfectly good SATA port on slow drives? :)
[10:47] <Mithrandir> Treenaks: because you want better air flow and/or longer cables?
[10:47] <ploum> Hello 
[10:47] <Treenaks> Mithrandir: naaah!
[10:47] <Treenaks> ploum: Goeiemorgen :)
[10:47] <Mithrandir> Treenaks: also, most of my machines use one or maybe two of the eight or so SATA ports.
[10:48] <Treenaks> Mithrandir: None of my machines has SATA ports anymore..
[10:48] <ogra> pitti, will do, thanks for the heads up 
[10:48] <ploum> sabdfl: isn't http://daniel.holba.ch/ubuntu/ic/#network-wireless-encrypted and http://librarian.launchpad.net/1818438/nm_sshot.png the same icon ?
[10:49] <ploum> (I replied in the comment *above* your question I think ;-) )
[10:49] <ploum> bug #36017
[10:49] <Ubugtu> Malone bug 36017 in ubuntu-artwork "Network-Manager WEP icon not really useful" [Normal,Unconfirmed]  http://launchpad.net/bugs/36017
[10:50] <fabbione> ati users the next driver upgrade will bring TONS of fixes
[10:50] <seb128> ploum: what version of n-m do you use?
[10:50] <ploum> seb128: the 0.6 test 
[10:50] <seb128> k
[10:51] <ploum> https://wiki.ubuntu.com/DapperNetworkManager
[10:51] <sabdfl> ploum: ah, yes it is. Interesting, because NM has its own vpn icons, but it uses tango for that
[10:52] <Mithrandir> seb128: did you find out anything more about the xkb-c thing you tested yesterday or can I upload?
[10:52] <ploum> Indeed, I've seen that on dholbach page (great work this page, really fine !)
[10:52] <ploum> So it's a tango bug..
[10:52] <ploum> I will report it upstream
[10:52] <seb128> Mithrandir: the patch I pointed on the bug fixes some issue according to svu, if you want to ship it with your upload feel free
[10:53] <Mithrandir> seb128: sure, I can do that, then
[10:53] <seb128> Mithrandir: remaining issue is an xkbcomp or xorg issue according to him, he's working on it
[10:53] <Mithrandir> seb128: what do you want me to put in the changelog?
[10:53] <seb128> Mithrandir: I had that
[10:53] <seb128>    * symbols/level3, types/pc:
[10:53] <seb128>      - CVS patch to fix the issue with "compose:ralt" (Ubuntu: #35845)
[10:54] <sabdfl> ploum: i don't know if its a bug or a feature
[10:54] <seb128> Mithrandir: maybe changing it by "some issue"
[10:54] <Mithrandir> seb128: yup, thanks
[10:54] <sabdfl> i will promote those icons anyway, so we get good ones done.
[10:54] <sabdfl> ploum: good catch!
[10:54] <seb128> Mithrandir: or something pointing there is something to fix :)
[10:54] <seb128> Mithrandir: np, thank you
[10:54] <ploum> sabdfl: thanks. I will report it upstream and they will close it if it is a feature..
[10:56] <Kamion> doko: openoffice.org-l10n accepted, sorry for the delay
[10:58] <Mithrandir> seb128: it's lovely that svu actually uses malone, though
[10:58] <seb128> yeah, that's nice when upstreams do that
[10:59] <seb128> there is an xchat-gnome upstream subscribed to the package, he reply on bugs, forwarded them, etc ... really handy :)
[11:00] <Kamion> pitti: postgresql-client-common accepted
[11:00] <pitti> Kamion: thank you
[11:01] <pitti> yay, fresh live isos - another espresso testing time today :)
[11:06] <jdub> BenC: ping
[11:06] <jdub> BenC: ping (sata_promise stuff)
[11:15] <doko> Kamion: thanks. no problem, if it gets onto Flight6 :)
[11:17] <pvanhoof> hey guys
[11:17] <pvanhoof> yesterdays kernel upgrade seems to have broken my boot
[11:17] <pvanhoof> you want me to help you investigate this, or are you guys already aware ?
[11:18] <Kamion> mjg59: have time to test a stripped-down i386+mac CD in a minute?
[11:18] <pvanhoof> I managed to boot by booting a live cd and changing the default to 2 in menu.lst
[11:18] <pvanhoof> which is of course the previous kernel (if you count the recovery kernel)
[11:18] <ogra> Kamion, no cdimages last night ? 
[11:18] <pvanhoof> it also looks like grub's ESC menu is broken too
[11:19] <Kamion> ogra: scroll up
[11:19] <pvanhoof> anyway, if you have questions about what might have happened ... let me know, I'll assist you and will launch your funny commands and paste the output
[11:20] <ogra> Kamion, thanks
[11:20] <mjg59> Kamion: Should do
[11:22] <pvanhoof> anyway, I've been trying to boot my dapper since this morning, and this is in messages:
[11:22] <pvanhoof> Mar 22 17:31:28 localhost exiting on signal 15
[11:22] <pvanhoof> Mar 23 11:15:04 localhost syslogd 1.4.1#17ubuntu4: restart.
[11:22] <pvanhoof> so basically .. syslogd didn't yet start
[11:23] <pvanhoof> it hung at the second line in the boot-splash screen (Mounting the LV /dev/Ubuntu/root)
[11:23] <pvanhoof> the restart line is this session, my first that actually succeeded (after revering the kernel upgrade)
[11:23] <Kamion> mjg59: http://people.ubuntu.com/~cjwatson/tmp/dapper-install-i386+mac.iso
[11:23] <Kamion> mjg59: I've taken out dists and pool so it won't actually do anything useful, but I'd like to know if it boots
[11:23] <pvanhoof> title           Ubuntu, kernel 2.6.15-19-686
[11:24] <mjg59> Kamion: Ok
[11:24] <pvanhoof> is the latest in my menu.lst (the one which I didn't boot for this session)
[11:24] <mjg59> May take a little while to download
[11:24] <Kamion> hence taking out dists/ and pool/ :)
[11:36] <doko> Kamion: the eclipse-pydev sources in NEW are reviewed from my side, don't contain any non buildable/non-free jars anymore.
[11:41] <Kamion> doko: thanks, accepted
[11:49] <Seveas> infinity, ping
[12:08] <mjg59> Kamion: Nope, not recognised as bootable
[12:08] <mjg59> Kamion: Did I need to burn it with any special options?
[12:08] <Kamion> ogra: you have updated images now
[12:08] <Kamion> mjg59: hmph, bugger
[12:08] <Kamion> no, shouldn't have needed any special options
[12:08] <Kamion> maybe it just doesn't like the hybridness or something
[12:08] <Kamion> or maybe I screwed up with -hfs-bless-file
[12:08] <ogra> Kamion, thanks 
[12:08] <mjg59> I should check whether the install DVDs are hybrid
[12:09] <Kamion> could be a byte-swapping issue
[12:09] <ogra> they'll be oversized anyway i guess :)
[12:09] <Kamion> ogra: install CDs are, live CDs seem OK
[12:10] <ogra> yep, which is astoiushing ... :)
[12:10] <ogra> *astonishing
[12:10] <mjg59> Kamion: Given a known bootable DVD, is there anything useful I can give you?
[12:11] <mjg59> Hm. The install DVDs are pure HFS.
[12:11] <mjg59> Pure hfsplus, that is.
[12:11] <mjg59> I wonder if that matters...
[12:11] <mjg59> (the hfsplusness)
[12:12] <Kamion> yeah, I noticed that
[12:13] <Kamion> mjg59: I've got a CD image here from mactel-linux.org - I'll start comparing notes
[12:13] <mjg59> Kamion: Ta
[12:14] <Kamion> it certainly could be that it only takes HFS+
[12:14] <Kamion> which would be painfully awkward
[12:14] <Kamion> I'd have to go for the second ElTorito image approach if that's the case, I think
[12:15] <mjg59> The source to mfs.hfsplus is available, I believe, but under APSL...
[12:15] <mjg59> (mkfs)
[12:15] <Kamion> we have hpformat
[12:15] <Kamion> er, no we don't
[12:16] <j^> but udf dvds should work too
[12:16] <mjg59> Yeah :(
[12:16] <mjg59> j^: No
[12:16] <mjg59> The firmware appears to understand HFS and FAT
[12:16] <Kamion> how can we have hpfsck but not hpformat? bah
[12:16] <mjg59> Kamion: Hang on, I'll go and check whether I can actually read the CD in the firmware
[12:19] <mjg59> Kamion: Hm. Seemingly not.
[12:21] <Kamion> I'll go look up the HFS+ specs and see if I can reconcile them
[12:21] <Kamion> (with bless, and with mkisofs, and with the mactel-linux CD image)
[12:22] <Kamion> ah, http://developer.apple.com/technotes/tn/tn1150.html lists a startup file as an HFS+ extension
[12:23] <Kamion> but it's talking about something different from what bless seems to do ...
[12:24] <mjg59> Kamion: Ah. I can ls the install DVD from the firmware, but not this CD
[12:25] <mjg59> Which implies a fair degree of unhappiness
[12:25] <mjg59> (Wonder if it needs a partition table?)
[12:26] <Kamion> it should have one, I think
[12:27] <Kamion> dapper-install-i386+mac.raw1     Apple_partition_map Apple                    2 @ 1       (  1.0k)  Partition map
[12:27] <Kamion> dapper-install-i386+mac.raw2               Apple_HFS Ubuntu 6.06 i386     30732 @ 16      ( 15.0M)  HFS
[12:27] <Kamion> although mac-fdisk goes on to witter a great deal
[12:27] <Kamion> so it could be that stuff is a bit corrupted
[12:27] <mjg59> Hang on, I'll dump the one from the install disc
[12:28] <Kamion> Desktop/ubuntu-base-20060306.cdr1     Apple_partition_map Apple                   63 @ 1       ( 31.5k)  Partition map
[12:28] <Kamion> Desktop/ubuntu-base-20060306.cdr2               Apple_HFS disk image          819120 @ 64      (400.0M)  HFS
[12:28] <Kamion> Desktop/ubuntu-base-20060306.cdr3              Apple_Free                         16 @ 819184  (  8.0k)  Free space
[12:28] <Kamion> (mactel-linux)
[12:29] <infinity> Seveas: Pong?
[12:30] <mjg59> Kamion: http://ubuntu.pastebin.ca/46694
[12:31] <Kamion> ok, looks normal enough
[12:31] <mjg59> Hm. Both the working ones have bigger partition maps.
[12:31] <mjg59> Wonder why.
[12:32] <Kamion> 31.5k probably won't fit in the room that hybridisation uses in ISO9660
[12:32] <Mithrandir> fabbione: do you have an xorg pending?
[12:32] <Kamion> mjg59: do you have an Ubuntu powerpc ISO lying around?
[12:32] <Mithrandir> xorg upload, even
[12:32] <Kamion> mjg59: if so, it would be interesting to know whether you could ls that
[12:33] <mjg59> Kamion: Hm, good point.
[12:33] <mjg59> Hang on
[12:33] <fabbione> Mithrandir: i think so.. i need to check
[12:33] <fabbione> gimme 2 minutes please
[12:33] <mjg59> (I have a nasty feeling that the answer is "no")
[12:34] <mjg59> Hmm. Yes, in fact the answer does appear to be "no"
[12:34] <mjg59> I can download one, but it'll take a few hours
[12:34] <mjg59> Kamion: Of course, at this point you're entirely welcome to the imac for a while :)
[12:35] <Kamion> I have a sick child to take care of today, but I might come over tomorrow if that's possible
[12:39] <Mithrandir> fabbione: arewethereyet or do you want a patch?
[12:39] <fabbione> Mithrandir: i have no pending changes.. my last upload was ubuntu22 yesterday
[12:39] <Mithrandir> fabbione: cool, thanks.
[12:42] <mjg59> Kamion: Sure
[12:42] <heno> Mithrandir: any news on the seeding of the accessibility stuff? (did you get my email with package names?)
[12:42] <Mithrandir> heno: I got the mail, but I have neglected to take any action on it.
[12:42] <Mithrandir> heno: iirc, Kamion was wondering about festival vs flite
[12:43] <heno> Mithrandir: so flite is slightly smal;er, but festival is much better technically and sound-quality wise
[12:43] <heno> and supports multiple languages, fwiw
[12:44] <mjg59> Kamion: Ok, the OS X disk utility recognises it as partitioned media, and claims that it's HFS extended (journaled)
[12:44] <mjg59> Which doesn't sound right
[12:44] <Mithrandir> heno: -> Kamion; can you two battle it out? :-)
[12:45] <Kamion> the problem is that both of them are too damned huge
[12:45] <Kamion> libflite1 is 8MB
[12:46] <heno> Kamion: are we really in a sqeeze for space still?
[12:47] <Kamion> we can probably fit this lot in, but you're not the only one wanting the space
[12:47] <heno> Kamion: No, I can imagine
[12:47] <Kamion> we might have to make it amd64/i386 only (they're much better off for space on the live CD)
[12:48] <heno> I guess many more languages will have support this time
[12:48] <Kamion> Mithrandir: incidentally, if you can look at why the powerpc live CD is 60MB bigger than amd64, that would be good ...
[12:48] <Kamion> heno: s/more/fewer/
[12:48] <TheMuso> Kamion: Thats probably ok, as most people using the a11y tools are on thoe systems anyway.
[12:48] <Kamion> stuff has got bigger
[12:48] <heno> Kamion: that sounds fine
[12:48] <Mithrandir> Kamion: sure, I can take a look
[12:48] <Kamion> if we can figure out why powerpc live is so huge, it could go in there
[12:49] <Kamion> Mithrandir: the squashfs itself is much bigger
[12:50] <Mithrandir> Kamion: that's.. odd.
[12:50] <Kamion> Mithrandir: oh, never mind ...
[12:50] <Kamion> it's because we have many more languages seeded on powerpc ;-)
[12:50] <Kamion> might as well stay that way, we know we have room for expansion by taking those out
[12:50] <Kamion> in fact I might add those languages to the other architectures now
[12:51] <Mithrandir> Kamion: heh, ok. :-)
[12:52] <Kamion> Mithrandir: can you make a copy of the dapper seeds, add the stuff that heno wants to them, and use germinate on that copy to get a size report? we can then make a decision
[12:52] <Mithrandir> Kamion: I'll look at that, sure.
[01:06] <Seveas> infinity, according to the changelog you applied _ion/Pygi's madwifi patch to l-r-m -- it seems not to work though since NM won't recognize my atheros card as WPA capable although the madwifi driver in the l-r-m package from _ion/Pygi does
[01:06] <infinity> Seveas: Erm, which package from them are you referring to?
[01:06] <infinity> Seveas: URL?
[01:06] <Seveas> their l-r-m package -- which is no longer available since the patch is supposed to be in the official l-r-m
[01:07] <infinity> Seveas: Well, the -18- packages they were distributing were built by me...
[01:07] <infinity> Seveas: With the same patch...
[01:07] <Seveas> hmm, odd
[01:07] <infinity> Seveas: Like, identical.  The only difference between those packages and these are control file changes.
[01:07] <Seveas> that means I'll have to apt-get source and poke around 
[01:08] <infinity> (Well, and being built against the -19- kernels)
[01:08] <Seveas> thanks for the information - I'll get back to you if their really is something wrong
[01:08] <infinity> Please do.
[01:08] <infinity> I wanted that patch in the mainline LRM specifically so it could get more widespread testing.
[01:09] <infinity> (It seems pretty obviously harmless, even if it doesn't work, it can't really hurt anything...)
[01:09] <Seveas> indeed
[01:11] <infinity> Seveas: Do you know if the NM 0.6.0 kids are seriously looking at backporting background scanning to madwifi-old?
[01:11] <infinity> (I suspect it would require too many architectural changes to the code, but maybe it's simpler than that..)
[01:11] <Seveas> infinity, they are seriously thinking about it but I have serious doubt about the doability of it...
[01:12] <infinity> Yeah, me too.  It'll still be pretty nice.
[01:12] <Seveas> indeed
[01:12] <infinity> Other than support for new devices (which is going to require madwifi-ng in some hacked-up fashion, *sigh*), background scanning is the only complaint about madwifi-old.
[01:14] <Seveas> madwifi-ng is on serious crack
[01:15] <Seveas> I have serious doubts about it being in a good enough shape for dapper soon
[01:15] <Mithrandir> heno: you just want the packages onto the live cd, right?
[01:15] <infinity> It's starting to look like a much better codebase, technically, but it's got a long way to go to shake the bugs out that come from a major code reorg.
[01:15] <infinity> Seveas: It'll never be in good enough shape for dapper, but for some new devices, we have no choice.  We ship a half-broken driver, or no driver.  \o/
[01:15] <Kamion> mjg59: can you put an 'od -Ax -tx1 -N65536' of your working install DVD somewhere for me?
[01:16] <Seveas> infinity, go us :/
[01:16] <heno> Mithrandir: yes, but it would be good if they would then also get installed
[01:16] <sladen> fabbione: I'm going to do an upload of mesa for the i945GM if you're blocking
[01:16] <Kamion> Mithrandir: it's a lot easier to have them in the desktop seed
[01:16] <sladen> fabbione: ...*not* blocking
[01:17] <heno> Mithrandir: otherwise we cannot really claim an accessible install procedure
[01:17] <DVSoftware> hello
[01:17] <mjg59> Kamion: http://www.codon.org.uk/~mjg59/tmp/macosx_dump
[01:17] <DVSoftware> hello
[01:17] <Kamion> although it does kind of bloat up the installation rather a lot for people who don't need it, so I don't know
[01:17] <infinity> Riddell: Around?
[01:18] <DVSoftware> i got one question, and i hope someone is willing to help me
[01:18] <Kamion> perhaps it should go only in the live seed and we should make espresso keep it installed only if you specify the relevant a11y options
[01:18] <DVSoftware> it's regarding ubuntu's hardware detection method on livecd
[01:18] <TheMuso> Kamion: Is that trivial?
[01:18] <Kamion> TheMuso: relatively, yes
[01:19] <Kamion> it would then go in the ship seed too so that the regular install CD can install it
[01:19] <heno> Kamion: in that case that seems ideal
[01:19] <DVSoftware> actally, i'm developing my custom livedvd for learning and demonstration purposes filled with all kind of software
[01:19] <heno> cool!
[01:19] <Kamion> DVSoftware: the hardware detection on the live CD is pretty much just the same as that in the regular system
[01:19] <Kamion> i.e. udev
[01:19] <DVSoftware> and i need some help with hardware autodetection
[01:19] <DVSoftware> mostly graphics card
[01:20] <infinity> Riddell: kdenetwork is FTBFS on all arches (despite having at one point built on a few of them) with the following:
[01:20] <heno> Kamion: Could it be installed directly during the install procedure (in Experft mode, say)?
[01:20] <infinity> dh_install -pktalkd  
[01:20] <infinity> cp: cannot stat `./debian/tmp/etc/kde3/ktalkdrc': No such file or directory
[01:20] <infinity> dh_install: command returned error code 256
[01:20] <Kamion> heno: on the install CD, sure
[01:20] <infinity> Riddell: ^^^^
[01:20] <infinity> Riddell: Pls fix, kthxbye.
[01:20] <sladen> DVSoftware: if you find anything that doesn't work, please file bugs with the PCI IDs and drivers that didn't work
[01:20] <Kamion> heno: I see no reason for it to be expert-mode-specific; we could key off the a11y boot options in gfxboot
[01:20] <DVSoftware> sladen: no, everything is alright with ubuntu
[01:21] <DVSoftware> that's the reason i come here
[01:21] <DVSoftware> :)
[01:21] <heno> Kamion: even better :)
[01:21] <Kamion> we basically just do dpkg-reconfigure -fnoninteractive xserver-xorg in the live filesystem
[01:21] <Kamion> DVSoftware: ^--
[01:21] <DVSoftware> hmmmm
[01:21] <DVSoftware> what that command do exactly?
[01:21] <Kamion> 'apt-get source casper' for more details
[01:21] <Kamion> DVSoftware: tells X to reconfigure itself for the current system
[01:21] <DVSoftware> ehm.....
[01:22] <Kamion> using its debconf maintainer scripts
[01:22] <DVSoftware> but how it tells it to X... when i try to use X's autodetection it fails
[01:22] <DVSoftware> problem is that my livedvd is not debian based :(
[01:22] <Kamion> it's not X's native autodetection, it's the Debian/Ubuntu maintainer scripts
[01:23] <Kamion> DVSoftware: oh, we can't help you then
[01:23] <DVSoftware> hmmmm
[01:23] <sladen> DVSoftware: X's install scripts use discover1 to select the driver and then build a xorg.conf based on that
[01:23] <Kamion> the graphics card detection on our live CDs is intrinsically based on using the Debian/Ubuntu maintainer scripts, which are a couple of thousand lines of code
[01:23] <Riddell> infinity: ok, thanks
[01:24] <DVSoftware> can you point me to that script so i can (hopefuly) customize it for my needs?
[01:24] <Kamion> sladen: but if it's not Debian-based he won't have our discover1/discover1-data patches
[01:24] <fabbione> sladen: is that to add the support for it?
[01:24] <Kamion> DVSoftware: apt-get source xserver-xorg, see debian/xserver-xorg.*
[01:24] <sladen> DVSoftware: apt-get source xserver-xorg
[01:24] <Kamion> er, 'apt-get source xorg' on breezy
[01:25] <Kamion> but it's really too much to ask to help you reimplement it from scratch, I think
[01:25] <pitti> fabbione: did you see my msg wrt. git-core's main inclusion report from the other day?
[01:25] <sladen> fabbione: yes, it's about a 10 line delta, aliasing i945GM to the i945G code-path in most places
[01:25] <fabbione> sladen: ok.. go ahead
[01:25] <Kamion> it's not a standalone script and relies on a lot of our infrastructure
[01:25] <fabbione> pitti: no i think i missed it?
[01:25] <DVSoftware> Kamion: i'm not going to reimplement it from scratch :)
[01:25] <TheMuso> Mithrandir: I should have a majorly trimmed down accessibility startup script for casper some time later tomorrow my time, as I will be finalizing the move of most package-specific settings into the packages themselves.
[01:25] <Kamion> DVSoftware: trying to customise it on a non-Debian-based system would be nearly equivalent
[01:26] <Mithrandir> TheMuso: excellent.
[01:26] <Kamion> you can try if you like, but expect it to be a good deal of hard work
[01:26] <pitti> fabbione: https://wiki.ubuntu.com/MainInclusionReportGitCore -> it needs a security update before I'll approve it
[01:26] <sivang> hi all
[01:26] <sivang> do I need something special to be able to build kernel modules from source other then linux-source ?
[01:26] <fabbione> pitti: ok.. works for me
[01:27] <pitti> fabbione: http://lwn.net/Articles/169623/, btw
[01:27] <sivang> (trying to build cisco vpn client which needs to build CiscoVPN kernel module)
[01:27] <fabbione> sivang: you only need the headers
[01:28] <Aegir> sivang, Make sure you have gcc-3.4 and the kernel headers, that's about it, really. Do an export GCC=gcc-3.4 before compiling, and your set. Atleast that's the method working for me.
[01:29] <fabbione> pitti: ahahha "please practice safe merge ;-)."
[01:29] <sivang> fabbione: k, thanks. weird thing, the cisco installer tries to run 'modules' target from the include/linux tree, and can't find this target and fails.
[01:29] <fabbione> pitti: ok.. is the new source already in debian?
[01:29] <Kamion> mjg59: interesting that they've stopped doing the HFS wrapper thing; according to http://developer.apple.com/technotes/tn/tn1150.html#HFSWrapper "Apple software currently initializes all HFS Plus volumes with an HFS wrapper"
[01:29] <sivang> Aegir: thanks, why 3.4 specifically ?
[01:30] <Kamion> mjg59: so I suspect that they've just ripped plain HFS support out of the firmware - sigh
[01:30] <pitti> fabbione: save merge> :)
[01:30] <infinity> sivang: Because in breezy, that's what we built the kernel with.  If you're running dapper, his advice is stale, we use gcc-4.0 :)
[01:30] <pitti> fabbione: I think so, we can just sync if mdz approves the new version
[01:30] <pitti> fabbione: 1.1.4->1.1.5 is safe, no idea about 1.1.3->1.1.4
[01:30] <Kamion> mjg59: which means either we need to add format support to hfsplus (APSL isn't free enough for main, right? wrong?), or we need to add an -hfsplus option to mkisofs, or both
[01:31] <fabbione> pitti: usually git is safe.. tons of eyes on it
[01:31] <sivang> infinity: cool, I'm using dapper then :)
[01:31] <pitti> fabbione: as long as our kernel maintainers use 1.1.5, I had no objection anyway, since it's just for them :)
[01:31] <sivang> infinity: linux-kernel-headers is a meta package against the haders suitable for the current running kernel for me?
[01:32] <infinity> sivang: No, linux-kernel-headers are for userspace.
[01:32] <Aegir> infinity, It's gcc-4 now? Oh goody, I've been using 3.4 for kernels for a while now... Woops...
[01:32] <infinity> sivang: You want linux-headers-`uname -r`
[01:32] <sivang> infinity: k, thanks.
[01:32] <Aegir> Sorry for the bogus advice then folks
[01:33] <Kamion> hppa's still on 3.4, I think - but you probably don't care
[01:33] <infinity> Yeah, it is.
[01:33] <sivang> hmm, it can't find a kernel headers suitable for me kernel
[01:34] <sivang> (2.6.15-17-386)
[01:36] <Seveas> infinity, hmm, in linux/wireless.h: #define WIRELESS_EXT 17 
[01:36] <Seveas> that may very well explain the lack of WPA support in NM (needs 18/19)
[01:37] <Seveas> odd that it worked before...
[01:37] <sivang> infinity: any idea why we seem not to have the headers of 2.6.15-17-386 
[01:37] <sivang> ?
[01:37] <Mithrandir> -17 is obsolete
[01:37] <infinity> sivang: Because you're severl kernel versions behind, and that's gone from the archive?
[01:37] <Mithrandir> -19 is the newest
[01:38] <infinity> Seveas: Hrm, but madwifi (both old and ng) don't support WEXT 18/19, afaik...
[01:38] <infinity> Seveas: And I certainly never changed that define in the previous builds.
[01:38] <infinity> Seveas: Does rebuilding LRM with that one define changes make the world different for you?
[01:38] <Seveas> infinity, no, but NM checks that #define before even attempting WPA iirc
[01:39] <infinity> s/changes/changed/
[01:39] <Seveas> which makes it all really weird
[01:39] <Seveas> anyway, I'll have to get my facts straight here
[01:41] <infinity> Seveas: I expect that if that IS true, we're going to have to patch NM to skip that check on ath_pci, cause I refuse to fake the WEXT version in the driver.
[01:41] <infinity> Seveas: That way lies madness will all sorts of other tools.
[01:41] <Seveas> indeed
[01:42] <Seveas> but I'm trying to dig some more into this horrible mess
[01:49] <Mithrandir> seb128: 35845 ; what's you initial keyboard setup?  It works fine for me.
[01:52] <doko> Kamion: please remove the swt-gtk source, now built from the eclipse package
[01:55] <azeem> did the changelog links on packages.u.o stopped to get updated at some point?  Most of the ones I tried are 404 and only older ones are available
[01:57] <infinity> azeem: They haven't worked for ages.  After clicking on one, do s/packages/changelogs/ in the URL bar.
[01:57] <azeem> infinity: oh, thanks
[01:59] <infinity> azeem: changelogs.ubuntu.com is a service we control, packages.ubuntu.com is provided by djpig, so we have no control over its brokenness.
[01:59] <azeem> I thought so
[02:00] <infinity> azeem: If you could ask him so s/packages/changelogs/ in his HTML generation code, though.. :)
[02:00] <azeem> infinity: I hope I remember next time I'll see him
[02:00] <sivang> Mithrandir, infinity : yes, thanks, I'm upgrading now
[02:02] <ogra> infinity, azeem, they only have changelogs for the stable release ... (always had) 
[02:02] <ogra> changelogs.ubuntu.com is what you look for ;)
[02:03] <mvo> does anyone mind if I do a ubuntu-meta upload now? Kamion, is that ok with you?
[02:04] <infinity> mvo: I have no objections.  I have a bunch of seed changes/tweaks to commit, but they're all in supported, so won't affect -meta anyway.
[02:04] <JaneW> jdub: :)
[02:04] <ogra> mvo, any changes that require a edubuntu-meta update ?
[02:05] <mvo> ogra: some fonts from me ... not sure if you take those too (they take ~1,5mb space)
[02:05] <ogra> mvo, then not :)
[02:06] <ogra> oh, btw
[02:06] <mvo> ogra: lovely thai and lao gylphs
[02:06] <ogra> seb128, do you mind if i do a gdm upload with the added edubuntu-artwork dependency ? 
[02:06] <j^> hm, the volume keys on my X30 now also trigger a software volume change of Master. they are hardware volume keys.
[02:06] <ogra> mvo, suuure ... but amd64 and ppc are ~9MB oversized :)
[02:07] <ogra> no chance for 1.5MB :/
[02:07] <Mithrandir> Kamion: would you mind if I did ran cron.daily on little?
[02:08] <sladen> j^: yes.  Now, forgetting that you happen to know that ;-)  Does it do what you expect?
[02:08] <j^> no
[02:08] <j^> mute does not work
[02:09] <sladen> j^: pressing hardware-mute and then unmuting from the software mixer does not sync it back.  correct, that is the one corner case
[02:09] <j^> i never change Master volume, only PCM, changing Master volume i get distortion
[02:10] <j^> sladen right the mute key is in a corner case since its on the right side :)
[02:10] <sladen> j^: disortion while changing it?  disortition at/near 100% on most laptops is normal.  about 90% tends to work out best
[02:10] <sladen> j^: :)
[02:11] <j^> sladen anything above ~80% is bad, fixing Master at ~75% and changing PCM from 0-100 sound is ok
[02:12] <Mithrandir> Kamion: also, installing the stuff heno asked about on the live cd seems to need 46.9MB (unpacked), according to apt.
[02:12] <sladen> j^: one thing can you check for me;  when you press volume-up, is the jump the same amount as volume-down?
[02:12] <desrt> j^; i don't suppose you're talking about powerbooks?
[02:12] <Mithrandir> seb128: 13:49 < Mithrandir> seb128: 35845 ; what's you initial keyboard setup?  It works fine for me.
[02:13] <sladen> j^: here I'm now seeing that volume moves it about 1/3rd as much as volume-down (bigger steps)
[02:13] <Mithrandir> s/you/your/, obviously.
[02:13] <Mithrandir> sladen: bug 35080 ; do you have any suggestions on how to fix it too?
[02:13] <Ubugtu> Malone bug 35080 in xkeyboard-config "Numlock key doesn't work on IBM t42" [Major,Confirmed]  http://launchpad.net/bugs/35080
[02:14] <j^> sladen down is bigger, up is 1/3rd
[02:14] <j^> desrt no about thinkpads
[02:14] <infinity> Do we still need/want gst-plugins0.8 in main?
[02:14] <infinity> seb128: ^^^
[02:14] <desrt> sound gets really bad on the PB too above like 80% vol on either master or pcm
[02:15] <j^> sladen the mute button works, but the icon that is shown is not mute, other than that seams ok
[02:15] <infinity> seb128: Oh, nevermind, it's not in main.  soyuz bug (which Celso has fixed, but not in production yet)
[02:15] <infinity> seb128: Ignore me. :)
[02:16] <sladen> j^: it sould the same icon, but without the o/~ musical notes coming out of it
[02:17] <sladen> Mithrandir: disable Shift-NumLock => Pointer_EnableKeys by default in all key-maps?
[02:17] <j^> sladen a right, its confusing that it still shows the volume bar at the bottom
[02:17] <Mithrandir> sladen: how will people enable Pointer_EnableKeys then?
[02:17] <j^> sladen something like a striked out speaker would be better.
[02:17] <sladen> heno: does Accessibility Support have a way to enable Pointer Mouse Emulation on the Numerica Pad?
[02:18] <heno> sladen: that's standard in gnome
[02:18] <sladen> j^: can you file a wish-list against  gnome-settings-daemon
[02:18] <Mithrandir> heno: from the keyboard?  How?
[02:19] <sladen> heno: so would dropping the  shift-numlock  to turn on pointer-emulation be bad/good/ugly
[02:19] <heno> Mithrandir: Cleverly hidden Under System -> Pref -> Keyboard -> Accessibilty -> Mouse
[02:19] <Mithrandir> heno: and you're going to enable that how if you can't use a mouse? :-)
[02:20] <Mithrandir> heno: or is that a silly question to ask?
[02:20] <heno> Mithrandir: no, I think the whole way these things are managed should be reworked
[02:20] <heno> It should be collected in one place
[02:21] <heno> I've been thinking of making a SoC project of it
[02:21] <Mithrandir> is SoC going to exist this year too?
[02:21] <sladen> Mithrandir: thing is.  If we do an xmodmap fix just for ThinkPads, mouse-emulation would be disabled on an external keyboard anyway
[02:22] <heno> You would be able to enable it with the keyboard if key navigation workeed everywhere
[02:22] <Mithrandir> sladen: yes, which sucks.  I would at least think some people use it.
[02:22] <sladen> Mithrandir: or some Evilness at the input layer so that on_numlock, send shift-up, numlock down, numlock up, shift-down
[02:23] <heno> On the live CD we should be good because we enable sticky keys with F4-option 4
[02:23] <heno> otherwise you cannot get to the application or system menu
[02:23] <heno> (Alt-F1)
[02:24] <mhz> TheMuso: ping
[02:24] <sladen> Mithrandir: to me, the shift-numlock => mapping by default is wrong as it's an X policy decision;  nobody's keyboard has little blue writing that says 'mouse emulation'
[02:25] <heno> sladen: because it's not a very common use case
[02:25] <heno> but important if you need it
[02:25] <Mithrandir> sladen: no Norwegian keyboards have the m marked in a way that suggests that  should come when you press lvl3 shift + m either.  Should X stop doing that?
[02:25] <heno> sladen: so does this conflict with input methods?
[02:27] <Kamion> doko: swt-gtk removed
[02:27] <sladen> Mithrandir: bah your ee keyboard
[02:27] <Kamion> mvo: seed changes> go ahead
[02:28] <Kamion> mvo: (er, ubuntu-meta upload, I mean)
[02:28] <Kamion> Mithrandir: feel free to run cron.daily
[02:28] <Mithrandir> Kamion: already done. :-P
[02:28] <Kamion> Mithrandir: 46.9MB> how about .deb size? that's more important
[02:28] <heno> sladen: normally these features would be enabled by pressing NumLock five times in a row for mouse keys
[02:28] <Mithrandir> Kamion: 10.8
[02:28] <sladen> heno: if people expect it to work, I guess it would
[02:28] <Kamion> 10.8 is just about doable
[02:28] <heno> and Shift 5 times for StickyKeys
[02:28] <Kamion> what is the figure with flite?
[02:29] <Mithrandir> Kamion: (what apt-get install wants to download, which is sensible)
[02:29] <sladen> heno: the five times thing I'd expect from having hit it so many times on windows
[02:29] <heno> that should notmally not get in other people's way
[02:29] <Kamion> Mithrandir: that's why I suggested germinate, which will let you get a definite figure
[02:29] <Mithrandir> Kamion: well, germinate didn't want to play with me, so I reverted to the ways I know.
[02:30] <Mithrandir> uh, flite is bigger?
[02:30] <Kamion> really? huh, go figure
[02:30] <Mithrandir> something's pulling in festival too
[02:31] <Mithrandir> festival vs flite is 7.1 vs 8.3
[02:37] <DVSoftware> i got discover1 running, now i need to look at the script
[02:40] <sladen> Mithrandir/heno: oooh, if you go into System->Settings->Keyboard->Accessibility->Mouse  and then press  Shift-Scrlock, it nicely toggles the mouse-keys
[02:40] <Mithrandir> sladen: which you can't do if you can't use a mouse.
[02:41] <heno> Mithrandir: you can if sticky keys are truned on already
[02:42] <heno> I would still like the Applicaton menu be be available with the Windows/Apple key though
[02:42] <heno> You can navigate to those menus with the keyboard
[02:42] <heno> you just cannot start the menu opening
[02:43] <heno> if at leadt one of those features isn't already on
[02:44] <heno> On Windows it's not a problem because the Windows key will take you to the Control panel where you can switch on the rest
[02:44] <heno> all with one finger
[02:44] <sladen> Mithrandir: the other option is to remap numlock to Something Completely Different in hotkey-setup on ThinkPads
[02:45] <seb128> Mithrandir: re
[02:45] <Mithrandir> sladen: I guess I'll punt it to upstream.
[02:45] <sladen> heno: you could put in for Ctrl-Esc too, that's the windows key-stroke that I would expect to work (it might even work on OS X too)
[02:45] <heno> sladen: you should try this https://wiki.ubuntu.com/HandsFreeEmailing if you haven't already
[02:45] <Mithrandir> seb128: hiya
[02:46] <seb128> Mithrandir: $ setxkbmap -rules xorg -layout us -model pc105 -option '' -option 'compose:ralt' -option 'grp:alts_toggle' -print | xkbcomp - :0.0
[02:46] <seb128> Mithrandir: does that work for you?
[02:46] <sladen> Mithrandir: there isn't a solution as the bug is in the thinkpad hardware---they never expected shift-numlock to ever be used, so it wasn't something that was considered
[02:46] <heno> sladen: how does that help if you cannot press ctrl and esc at once?
[02:46] <Mithrandir> seb128: hm, no.
[02:47] <seb128> Mithrandir: that's the bug :p
[02:47] <sladen> heno: it helps in other cases (no mouse, but I happen to expect it to work)
[02:47] <seb128> Mithrandir: but if you -print it, use xkbcomp -xkb and use the xkb it works fine
[02:48] <Mithrandir> seb128: so it's not an xkeyboard-config bug, as you noted in the bug.  Ack.
[02:48] <seb128> Mithrandir: the xkeyboard-config part is fixed by the patch from the bug
[02:49] <seb128> Mithrandir: there is xkbcomp or xorg issue still
[02:49] <Mithrandir> heno: are you allowed to use your foot instead?
[02:49] <Mithrandir> (for the hands-free emailing)
[02:51] <pitti> Mithrandir: hmm, espresso still crashes after keyboard layout selection on ppc on today's live CD
[02:51] <Mithrandir> pitti: is espresso-kbd-chooser or espresso-keyboard-selector on the cd?
[02:51] <pitti> Mithrandir: did the live image not yet catch the latest kbd-chooser version? or is the bug not yet fixed?
[02:51] <heno> Mithrandir: sure, cheating is alowed :)
[02:52] <heno> I think using just your feet will give a realistic impression of the issues
[02:52] <pitti> Mithrandir: kbd-chooser
[02:52] <heno> Mithrandir: let's say just one foot :)
[02:52] <Mithrandir> pitti: hmm, can you put the backtrace somewhere?
[02:52] <heno> The point is to try and think a bit about it
[02:52] <Mithrandir> heno: I believe using multiple feet on the mouse might be counter-productive.
[02:52] <pitti> Mithrandir: you mean s/selector/setup/?
[02:53] <Mithrandir> pitti: indeed.  and espresso-kbd-chooser is correct.
[02:53] <pitti> Mithrandir: sure, I can file a bug with ESPRESSO_DEBUG=1, will that suffice?
[02:53] <Mithrandir> pitti: if you have the time, I would prefer to have you help me track it down.
[02:54] <pitti> Mithrandir: I have to go in 20 minutes, but I'm fine to take the time for that
[02:54] <Mithrandir> pitti: if you have the backtrace from /v/l/i/espresso, that'd be a good start.
[02:54] <heno> Mithrandir: true, but multiple feet on the keyboard can be useful
[02:54] <pitti> Mithrandir: it's a python exception: DebconfError: (10, "kbd-chooser/do_select doesn't exist")
[02:55] <pitti> Mithrandir: yep, second
[02:55] <Mithrandir> heno: true.
[02:55] <siretart> doko: re bug 12471 which you just closed
[02:55] <Ubugtu> Malone bug 12471 in subversion "missing javahl library" [Normal,Fix released]  http://launchpad.net/bugs/12471
[02:55] <Mithrandir> pitti: hmm, ok, I'll take a look at that.
[02:55] <siretart> doko: the bug is that this binary package libsvn-javahl is built in debian, but not in ubuntu
[02:55] <siretart> doko: the buildlog says this: dpkg-genchanges: warning: package libsvn-javahl in control file but not in files list
[02:56] <pitti> Mithrandir: http://people.ubuntu.com/~pitti/tmp/espresso
[02:56] <pitti> Mithrandir: let me know if you need a more verbose log
[02:56] <pitti> wow
[02:57] <Mithrandir> pitti: uh, that's really, really weird, given that it's shipped in the deb
[02:57] <Mithrandir> pitti: does grep do_select /var/lib/dpkg/info/espresso-kbd-chooser.templates
[02:57] <Mithrandir> give you something sensible?
[02:58] <pitti> Mithrandir: I guess so; Template:, Type: and Description: look sensible
[02:59] <doko> siretart: ok, I'll look
[02:59] <Mithrandir> pitti: does it exist in /var/cache/debconf/templates.dat and /var/cache/debconf/config.dat ?
[02:59] <pitti> Mithrandir: not in /var/cache/debconf/templates.dat
[02:59] <pitti> Mithrandir: neither in config.dat
[03:00] <Seveas> doko, dennis@mirage:~$ echo $PATH
[03:00] <Seveas>  /bin:/usr/bin
[03:00] <Seveas> aargghh 
[03:01] <Mithrandir> Seveas: what does grep PATH /etc/environment say?
[03:01] <Mithrandir> pitti: can you try apt-get --reinstall install espresso-kbd-chooser and tell me if it says "Extracting templates from packages"?
[03:02] <pitti> Mithrandir: 'kbd-chooser/do_select' vs. 'debian-installer/kbd-chooser/title' is correct?
[03:03] <Mithrandir> pitti: as in, whether it should exist in both namespaces?  Yes, that's right
[03:03] <pitti> Mithrandir: no sign of extracting templates with reinstall
[03:04] <Mithrandir> oh, it only prints that if you install 30 debs or more.
[03:04] <pitti> mvo: is there some debug option? ^
[03:04] <Mithrandir> pitti: can you strace -f -e execve it and see if it calls apt-extracttemplates ?
[03:04] <Mithrandir> pitti: it's debconf, not apt.
[03:04] <mvo> pitti: for what?
[03:05] <pitti> mvo: whether debconf extracts templates from adeb
[03:05] <Mithrandir> pitti: it's all controlled from debconf using apt hooks.
[03:05] <sladen> heno: press control-five times, then press escape?  :)
[03:06] <mvo> pitti: sorry, I don't know.
[03:06] <pitti> Mithrandir: yep, execve /usr/bin/apt-extracttemplates is called and succeeds
[03:06] <pitti> Mithrandir: it's called with 35 vars though, so I can't see all of them
[03:07] <Mithrandir> pitti: -s 512 ?
[03:08] <seb128> is "gettext" on the CD?
[03:08] <pitti> Mithrandir: no, that doesn't help
[03:08] <seb128> do we have an easy way to know what is on the CD or not? :)
[03:08] <pitti> FWIW, pacakge 'gettext' is not in the live session - does it have to?
[03:08] <ogra> seb128, grep the manifest ? 
[03:09] <seb128> pitti: pygtk upstream asking
[03:09] <seb128> ogra: what is "the manifest"?
[03:09] <pitti> ah, sorry, I thought it was related to Tollef's question
[03:09] <ogra> seb128, err, it was called mainfest earlier i think, look at the .list files in the cdimage folders
[03:09] <Seveas> Mithrandir, nothing
[03:09] <pitti> Mithrandir: I'll try with -v
[03:09] <seb128> ogra: k, thank you
[03:10] <Mithrandir> pitti: it seems like I need to make espresso-kbd-chooser depend on debconf to have it extract the templates.
[03:10] <Mithrandir> Seveas: uh, is this a clean dapper install, or?
[03:10] <Seveas> not exactly
[03:10] <j^> would be nice if Ubugtu would know whats on the CD
[03:10] <infinity> So, uhh, doing "bzr push" over sftp is a stupid idea, isn't it?
[03:11] <Mithrandir> infinity: no, just a slow one.
[03:11] <pitti> Mithrandir: I have the full execve() call here now, do you need it?
[03:11] <mvo> infinity: well, it seems to be a bit slow
[03:11] <Mithrandir> pitti: not really, no, I found the bug.
[03:11] <pitti> Mithrandir: cool, you rock :)
[03:11] <infinity> mvo: Yeah, I'm doing it for all 4 seed branches right now.  I think I may pass out before it finishes.
[03:11] <pitti> Mithrandir: any workaround for me to try here?
[03:11] <ogra> infinity, depends ... if you want to do your dishes between two pushes, its very helpful 
[03:12] <infinity> Yeah, well, I had this silly idea that if I was only pushing a single commit, it might be fast.
[03:12] <Seveas> Mithrandir, in my fresh dapper install a few weeks ago it wasn't in that file either. I added a path myself, which I removed yesterday. After that my path was normal except for not including /sbin:/usr/sbin
[03:12] <infinity> Ha ha ha ha.
[03:12] <Mithrandir> pitti: debconf-loadtemplate debian-installer /var/lib/dpkg/info/espresso-kbd-chooser.templates
[03:12] <Mithrandir> Seveas: it _should_ be in /etc/environment, though.
[03:13] <pitti> Mithrandir: debconf-loadtemplate doesn't exist here
[03:13] <Mithrandir> infinity: if you're pushing to something where only you have write access, use rsync
[03:13] <Mithrandir> pitti: install debconf-utils
[03:13] <infinity> Mithrandir: I'm not, I'm pushing to the main seed branches.  Which was dumb, in retrospect.
[03:13] <infinity> Mithrandir: For LP code, I push with rsync, then do all my merging on chinstrap, which seems much saner.
[03:14] <pitti> Mithrandir: cool, thanks
[03:14] <Seveas> Mithrandir, ok, riddle me this: in a VC my path is different than the one I get in gnome...
[03:14] <siretart> _ion: ping
[03:14] <Mithrandir> Seveas: compiled-in defaults.  Shouldn't be if you do the supported thing and just set PATH in /etc/environment
[03:14] <Seveas> ok
[03:14] <Mithrandir> infinity: yeah, sounds saner.
[03:15] <Seveas> i'm 100% sure though that it wasn't there right after install (flight 4), but tha could be solved already
[03:15] <pitti> Mithrandir: heh, now I see the keyboard window for 2 seconds, then it automatically advances to the next screen :)
[03:16] <Mithrandir> pitti: heh, ok.  I think I should try this on a live install myself.
[03:16] <Mithrandir> pitti: this isn't ppc specific.
[03:16] <Mithrandir> (I hope)
[03:16] <pitti> it so much doesn't sound arch specific...
[03:16] <pitti> but it just worked on amd64 for me
[03:16] <Mithrandir> hmm
[03:17] <Mithrandir> I'll see if I can reproduce it.
[03:17] <sladen> seb128: if you've been following above, what do you think of a   -option "mousekeys:shiftnumlock"   in all cases that it's not a thinkpad?
[03:17] <Treenaks> Whoa! ubuntu-announce isn't moderated?!
[03:24] <dholbach> Treenaks: it is
[03:24] <seb128> sladen: I don't know enough of xkb stuff to have an option on that, dunno what the option does in fact :p
[03:25] <Treenaks> dholbach: then why did I get that mail from Simon Jardine?
[03:25] <Treenaks> dholbach: which obviously isn't ubuntu-announce material
[03:25] <dholbach> Treenaks: there was one spam message which accidentally slipped through jdub's hands
[03:25] <dholbach> Treenaks: and he excused himself for that already
[03:25] <Treenaks> oh
[03:26] <sladen> seb128: the option doesn't exist at the moment.  I would add one
[03:31] <Mithrandir> sladen: I don't think we want to have that.
[03:32] <Kamion> Mithrandir: I bet you just forgot to add a postinst to espresso-kbd-chooser
[03:32] <Mithrandir> Kamion: indeed, it doesn't have a postinst.
[03:32] <Kamion> Mithrandir: you need a postinst that just does . /usr/share/debconf/confmodule, or else debconf won't load up the templates
[03:33] <Kamion> it's related to the hacky way that debconf hooks into package installation
[03:33] <Kamion> you do need to depend on debconf as well, certainly
[03:34] <Kamion> and yeah, it's not arch-specific, I can reproduce it here
[03:34] <Kamion> Mithrandir: want me to fix it now?
[03:35] <Mithrandir> Kamion: if you'd like to, please.  I'm in the middle of testing that korean actually works so kinda occupied.
[03:37] <Kamion> will do
[03:37] <Kamion> Mithrandir: can I remove some of the big pile of debugging code?
[03:38] <Mithrandir> Kamion: please do.
[03:38] <Kamion> or do you still want it in there?
[03:38] <maswan> BenC, mjg59: coordinate. :)
[03:38] <Mithrandir> Kamion: no, I tend to just add it when I need it, then forget to remove it afterwards. :-/
[03:39] <dholbach> Diziet: Hello... what do you think about building a firefox-dbg binary package?
[03:45] <slomo> fabbione: thanks for updating the ati driver :) fixes all the problems i had including the ppc 24bit display problems
[03:45] <dholbach> slomo: seems like we can set a lot of bugs needinfo :)
[03:45] <Treenaks> it might well save me from having to poke Radeon registers :)
[03:45] <slomo> dholbach: probably :)
[03:46] <Treenaks> hunger: did you update today?
[03:46] <dholbach> slomo: or to fix-released-please-reopen-if-things-are-still-b0rked
[03:46] <hunger> Treenaks: Yeap, even restarted X after reading the changelog of the ati driver:-)
[03:47] <glatzor> Chipzz: mvo: hi
[03:47] <dholbach> Diziet: i had the idea, when i had to rebuild firefox again, because yelp explodes on amd64 (http://daniel.holba.ch/ubuntu/yelp.bt)
[03:48] <hunger> Treenaks: I suspect it is the kernel starving some task or another... Desktop switching freezes for secounds, but the IRC window is still updated.
[03:49] <glatzor> mvo: Chipzz: I think that because many people mistake the "add" dialog for the "edit" dialog, doesn't make the "add" dialogs to a good "edit" dialog. The separation between both dialogs should be ratherer made clearer.
[03:50] <glatzor> shoudl be rather made clearer
[03:50] <ogra> dholbach, just remove the ff dependency :P
[03:51] <ogra> the i could switch edubuntu to epiphany if you remove that dep as well ... :/#
[03:51] <dholbach> ogra: narf!
[03:51] <glatzor> Chipzz: mvo: I don't know how deep ui changes should be made at this time. If we want to fix this issue, I would vote for the design, that I have suggested on the ubuntu-desktop list some time ago.
[03:52] <ogra> whats the reason everything depends on firefox now ... its very annoying ...
[03:52] <sladen> fscking slowkeys pop-under brokeness
[03:52] <glatzor> Chipzz: mvo: I could port the corresponding changes to the dapper branch at the weekend.
[03:54] <sladen> heno: slow-keys requires a mouse to raise and activate the dialogue, do you know of that bug already?
[03:55] <infinity> ogra: Because everything uses the firefox renderer?
[03:55] <dholbach> ogra: they embed firefox
[03:55] <infinity> ogra: epiphany is kinda useless without firefox embedded.
[03:55] <ogra> infinity, yes, but it was better when we had libnspr ... 
[03:55] <seb128> we still have libsnpr
[03:55] <dholbach> libnspr is only the tiniest part
[03:55] <ogra> i didnt need to kepp firefox and 100MB of translations around 
[03:55] <Chipzz> hrrrm, lemme catch up
[03:56] <ogra> seb128, yes, but we dont build epy against it ...
[03:56] <heno> sladen: how does that require a mouse more than anything else (I just did it without)
[03:56] <ogra> we discussed switching edubuntu to epy, but thats a no go if i still need ff on the CD ... then the switch is pretty pointless
[03:56] <seb128> ogra: 
[03:56] <seb128> $ ldd /usr/bin/epiphany | grep nspr
[03:56] <seb128>         libnspr4.so => /usr/lib/libnspr4.so (0xb7ea7000)
[03:56] <seb128> we do
[03:56] <Chipzz> glatzor: I think changing the UI too much would not be a good thing at this point
[03:57] <Chipzz> glatzor: my patch just reuses the existing UI, and is small and contained
[03:57] <seb128> ogra: but that's not enough to get randering, epiphany uses gtkmozembed
[03:57] <ogra> hmm, then i dont understand the principle of nspr
[03:57] <heno> One problem with gnome though is that you cannot tab backwards (shift-tab)
[03:57] <Chipzz> glatzor: http://chipzz.studentenweb.org/um.patch
[03:57] <ogra> i thought that was the gecko engine
[03:57] <dholbach> ok guys, I'm off
[03:57] <Kamion> Kinnison: hmm. I think the "Step N of M" could do with being moved down a bit so that it's at the same level as the cancel/back/forward buttons
[03:57] <ogra> dholbach, enjoy the weather :)
[03:57] <seb128> ogra: no, gtkmozembed is the gecko stuff
[03:57] <heno> I guess that's a bug. what should I file it under
[03:57] <dholbach> merci
[03:57] <ogra> seb128, ah ...
[03:58] <Kamion> (have only just actually seen it)
[03:58] <seb128> ogra: apt-cache show libnspr4
[03:58] <Chipzz> glatzor: it is by no means finished though
[03:58] <seb128> ogra: it doesn't do any rendering, just communication stuff
[03:58] <ogra> seb128, so if we have everything in external libs, what for do we need ff in there ? 
[03:58] <ogra> ah, k
[03:58] <seb128> ogra: because firefox does the rendering
[03:58] <ogra> yep, understood now ...
[03:58] <glatzor> Chipzz:  I am amware of the prototype nature of your path.
[03:58] <seb128> ogra: we will probably switch to xulrunned next cycle though
[03:58] <ogra> cool
[03:59] <seb128> xulrunner
[03:59] <glatzor> Chipzz: But your path also is a change in the UI. So I ask myself why not to fix the whole situation
[03:59] <ogra> i'd really love to be edubuntu the guinea pig for 100% epiphany
[03:59] <ogra> and drop the huuge firefox ...
[03:59] <Chipzz> glatzor: becaus my path, as it reuses existing code, is much more tested
[03:59] <Chipzz> glatzor: and it's not *really* a UI change
[04:00] <Chipzz> glatzor: big UI changes would probably not be accepted at this point anyway
[04:00] <glatzor> I would suggest to merge the "add" and "add custom" dialog, skip the the components from the "add" dialog at all.
[04:00] <infinity> ogra: xulrunner still won't be TINY (though smaller than firefox)... Keep in mind that firefox's original goal was to be a "less bloated mozilla" :)
[04:00] <jdub> and it is indeed a less bloated mozilla
[04:00] <infinity> Much so, yes.
[04:00] <danimo> BenC: ping?
[04:00] <ogra> infinity, i dont care... i care about the translations that eat my space, not about the 8MB ff binary :)
[04:00] <jdub> it just depends on how you define the completely meaningless word "bloat"
[04:01] <glatzor> Chipzz: I would not call this a big change. And of course your path includes definitely a change of the UI
[04:01] <Chipzz> glatzor: the idea, as I discussed it with mvo, is to make it easier to add components to existing channels, without being to invasive, for dapper, and then do a major UI overhaul for dapper+1
[04:01] <BenC> danimo: pong
[04:01] <jsgotangco> yelp depends on ff though right?
[04:01] <seb128> hate hate hate firefox translations system
[04:01] <ogra> yeah 
[04:02] <ogra> seb128++++++++++
[04:02] <seb128> jsgotangco: as everything else needing gtkmozembed atm
[04:02] <jsgotangco> gee
[04:02] <jdub> jsgotangco: could be switched to xulrunner
[04:02] <ogra> jsgotangco, it could use xulrunner as well at some point
[04:02] <danimo> BenC: hi, can I annoy you with some intel HDA driver fun?
[04:02] <seb128> I expect switch for next cycle
[04:02] <glatzor> Chipzz: on the underlaying code side we could use your math fenction. I don't want to include the whole child/parent stuff that I propsed earlier on ubuntu-devel, too.
[04:02] <seb128> Debian has switched for GNOME
[04:02] <danimo> BenC: your last changes in -19 broke them for me, in that I can no longer hear any sound
[04:02] <glatzor> match function
[04:03] <ogra> yay ...
[04:03] <danimo> BenC: I want to debug it as far as I can, I just need a point to start from
[04:03] <ogra> sad we didnt already ...
[04:03] <jsgotangco> hmmm
[04:03] <BenC> danimo: I believe those are fixed in current git
[04:03] <glatzor> Chipzz: my concern is that we mix up the two dialogs even more.
[04:03] <Chipzz> glatzor: that function needs changing anyway; it doesn't recognize deb-src channels atm
[04:03] <jsgotangco> ogra: epi default in edubuntu is not possible though for now right?
[04:03] <danimo> BenC: when will you create updated packages?
[04:04] <ogra> jsgotangco, we dont gain *anything* by it now
[04:04] <jsgotangco> gee
[04:04] <danimo> BenC: I appended my whinings to 35098
[04:04] <ogra> my only target is to free up CD space at this point :)
[04:04] <Chipzz> glatzor: but that would involve changing SourceEntryTemplate to have a types field instead of the type field
[04:04] <glatzor> Chipzz: you would also have to add a new option to change the channel type binary/source to the add dialog
[04:04] <Chipzz> glatzor: one sec
[04:04] <Chipzz> glatzor: will put my discussion with mvo online in a bit
[04:04] <danimo> BenC: not really related, but audio jack sensing didn't work for me either up to -18, and now I can't really test
[04:04] <jdub> BenC: what are the chances of pulling in an updated sata_promise driver?
[04:04] <ogra> if the edubuntu cookbook gets ready for dapper we need a lot of extra space i dont know where toi take yet
[04:05] <jdub> BenC: or a patch for a particular card to support
[04:05] <glatzor> Chipzz: ok
[04:05] <jsgotangco> ogra: release a dvd heh
[04:05] <jdub> BenC: http://lwn.net/Articles/176208/ (search for TX4300)
[04:06] <ogra> jsgotangco, we *have* a DVD for edubuntu 
[04:06] <ogra> :)
[04:06] <jsgotangco> release disc 2
[04:06] <ogra> its just not our main media
[04:06] <ogra> we'll be fine once i replaced kdeedu with gnome apps ...
[04:06] <ogra> but thats dapper+1 :+
[04:06] <ogra> :)
[04:07] <jsgotangco> yuo're not doing that now
[04:07] <ogra> nope
[04:07] <ogra> no big changes for dapper applies to edubuntu as well :)
[04:07] <ogra> i want a new app selection for dapper+1 ... without kdelibs and their unbelivable huge translation packages
[04:08] <ogra> (either that or a separate language CD)
[04:08] <Chipzz> glatzor: http://chipzz.studentenweb.org/mvo-um
[04:10] <seb128> mvo, doko: https://launchpad.net/distros/ubuntu/+source/fontconfig/+bug/35472
[04:10] <Ubugtu> Malone bug 35472 in fontconfig "Propose to set default font to Bitstream Vera instead of DejaVu" [Normal,Unconfirmed]  
[04:11] <BenC> jdub: I'll look into it
[04:14] <jdub> BenC: thanks!
[04:15] <glatzor> Chipzz: why do you want to extend the template by type?
[04:16] <glatzor> Chipzz: the type could be used from selected
[04:17] <glatzor> Chipzz: do you want to show the type combo/radio in the add dialog, too?
[04:17] <danimo> BenC: what git branch are you looking at?
[04:17] <danimo> (talking about alsa)
[04:17] <BenC> danimo: I'm looking at the ubuntu-2.6 git
[04:18] <danimo> BenC: oh, ok, I didn't know ubuntu was using git, too
[04:18] <danimo> BenC: where is it?
[04:18] <BenC> danimo: https://wiki.ubuntu.com/KernelGitGuide
[04:19] <danimo> tnx
[04:19] <glatzor> Chipzz: another problem that I see are mirrors.
[04:21] <glatzor> So since we are hackish people we could  use "if uri in 'ubuntu.com/ubuntu'" instead of the baseuri of the template
[04:21] <glatzor> that would be fine for dapper
[04:23] <Chipzz> glatzor: just try the patch; it works for deb repo's, but not for deb-src repo's (the templates only include deb as type)
[04:23] <Chipzz> the idea is to not extend the UI at all wrt deb/deb-src
[04:23] <Chipzz> glatzor: mirrors are a problem in the current UI anyway
[04:24] <glatzor> Chipzz: changing the mirror is not supported, but I would not call this a problem
[04:24] <glatzor> Chipzz: the mirrors are/should be detected in the sources treeview
[04:25] <Chipzz> glatzor: yes, but that is not a regression in my patch ;)
[04:25] <jdub> http://lwn.net/Articles/176825/
[04:25] <jdub> ^ ha ha ha
[04:26] <glatzor> Chipzz: oh, you pathed against your installed python files?
[04:26] <Chipzz> yes
[04:27] <Chipzz> I can put a fixed patch online too if you want
[04:27] <Chipzz> lemme do that
[04:27] <glatzor> it is "patch -p4 < /home/dapper/Desktop/um.patch", right?
[04:27] <glatzor> no. it worked so, too
[04:27] <Chipzz> glatzor: just reget the patch ;)
[04:27] <Chipzz> ah ok
[04:28] <Keybuk> ya know, I think I finally understand what WPA does/is now
[04:28] <jdub> Keybuk: kitten-strangling device?
[04:28] <sladen> Keybuk: "broken"?
[04:29] <rcaskey_> any thoughts on straightening out RestrictedFormatsAssistant, RestrictedFormatsOnDemandAssistant, Restricted Formats Solutions, Restricted Formats Solutions, and EasyCodecInstallat?
[04:29] <Kinnison> Kamion: do you want me to move it, or will you?
[04:29] <j^> Keybuk what did you think it is before?
[04:29] <Keybuk> j^: I had no idea
[04:29] <Keybuk> "scary evil kitten-strangling device"
[04:29] <Keybuk> now I think it's a garrotte for kittens
[04:30] <glatzor> Chipzz: I was worong. Thought that you used the templates of the info file. The mirrors are not an issue. 
[04:30] <Chipzz> glatzor: if you can get your branch in, all the better, but I dunnow if its finnished, and I doubt it would egt in anyway
[04:30] <Chipzz> and basically the situation as it is now sucks
[04:31] <Chipzz> so as a little pollish for dapper, I think my patch would be the best (short term, ie dapper) solution
[04:31] <Chipzz> we've been with the current UI for several releases; doing big changes at this point would not be a good idea I guess
[04:33] <Chipzz> like I said to mvo:
[04:33] <glatzor> Chipzz: doing something wrong for monthes doesn't make it right. take a look at malone, how many people mistae the add and edit dialog
[04:33] <glatzor> mistake
[04:33] <doko> seb128: ENOCLUE ... I thought that jdub did change fontconfig again to default to Bitstream?
[04:33] <Chipzz> 14:21 <Chipzz> the point of the patch is to make it very easy to enable multiverse etc
[04:33] <Chipzz> 14:21 <Chipzz> which is something a lot of users will want to do
[04:33] <Chipzz> 14:22 <Chipzz> and I think it would be a significant improvement to make that easier and more polished
[04:34] <Kamion> Kinnison: I won't have time until tomorrow or so, so if you're not too busy, please do
[04:34] <glatzor> Chipzz: and as I said before, I don't see the support for getting my branch in. There was only one reply on this topic.
[04:34] <Kinnison> Kamion: I'll do it now then, I can take a break from g-p-m for a bit
[04:34] <seb128> doko: no such upload according to my dapper-changes mailbox
[04:34] <Chipzz> glatzor: so if your branch can't get in, can'ty we use my patch as a solution for dapper and get your branch in for dapper+1?
[04:34] <glatzor> Chipzz: I also see the need. I justed waited if my branch would be accepted. 
[04:35] <pips1_> Kamion, ping
[04:35] <Chipzz> glatzor: I'm not even sure if my patch will be accepted ;)
[04:36] <Kinnison> Kamion: where exactly do you mean by "by the buttons" ?
[04:37] <danimo> BenC: indeed. great stuff
[04:37] <doko> mvo: I did hear you are the new font specialist?
[04:37] <danimo> BenC: seem like the issues with dell laptops are fixed
[04:37] <Kamion> Kinnison: there are "Cancel", "Back", and "Forward" buttons at the bottom of the screen
[04:37] <Kamion> Kinnison: the breadcrumbs should be vertically aligned with those
[04:37] <BenC> danimo: great, thanks for checking it
[04:37] <pips1_> Kamion,  found a wierd localisation bug: the installer ask me if I want to download language support - this question is in english, rather than german, like all the dialog questions before... I downloaded the debian-installer language file (po) from launchpad, but I could find those english strings in the po file at all...
[04:37] <Kamion> Kinnison: they're in hbuttonbox5 right now
[04:38] <danimo> BenC: well, I'd still have to compile the kernel myself to actually test it
[04:38] <Chipzz> glatzor: maybe we should wait for mvo to join the discussion ;)
[04:38] <Kamion> pips1_: I have to go out just now, but will you still be around in about 45 minutes or so? I can talk with you about it then
[04:38] <pips1_> ok
[04:38] <danimo> BenC: is there an instruction how to build ubuntu packages from the source?
[04:38] <glatzor> seb128: hi. at the moment it is not possible to add "universe" or other components in the softwareproperties by a single click. furthermore many people mix up the "add" and "edit" dialog.
[04:38] <Kinnison> Kamion: so top-right of the text aligned near the top left of the left-hand button?
[04:39] <Kamion> Kinnison: no, it should be at the left of that box, probably yalign=0.5 in it
[04:39] <seb128> glatzor: hey. Better to ping mvo about that :)
[04:39] <BenC> danimo: fakeroot debian/rules binary-debs flavours=686
[04:39] <danimo> ok
[04:39] <BenC> change 686 to whatever you actually need
[04:39] <BenC> debs will end up in debian/build/
[04:40] <Kinnison> Kamion: so, same horizontal position as before, vertically aligned center-to-center with the buttons?
[04:40] <Kamion> Kinnison: so like this:
[04:40] <Kamion> +-----------------------------------------------------+
[04:40] <Kamion> + Step 1 of 7               | Cancel | Back | Forward |
[04:40] <Kamion> +-----------------------------------------------------+
[04:40] <danimo> BenC: 686 is alright
[04:40] <Kamion> Kinnison: yeah
[04:40] <Kinnison> Kamion: okay, it'll need a good shuffle around because of the possibility of pictures
[04:41] <Kinnison> Kamion: I'll get on it right away
[04:41] <Kamion> thanks
[04:48] <glatzor> Chipzz: the add functions is a problem. if you change the channel in the edit dialog, a new channel entry will be generated by the add dialog
[04:48] <glatzor> Chipzz: how do you want to handle this exception?
[04:48] <mdke> scary stuff: opening a bookmark from my gnome places menu opens konqueror
[04:49] <ogra> thats the new ubuntu/kubuntu migration ...
[04:49] <ogra> wait until edubuntu joined in and tuxpaint opens for jpegs ;)
[04:49] <mdke> kde is taking over huh?
[04:50] <Chipzz> glatzor: hrrrm need to take a look at that
[04:50] <ogra> (multibuntu)
[04:51] <glatzor> Chipzz: the mirrors are an issue: the edit dialog can only add "new" components of the default archive and doesn't respect the mirrors used
[04:53] <glatzor> Chipzz: this issue will be very drastic fo cdrom users disabling and reenabling a component
[04:54] <glatzor> I thinkt that this approach is too much a kind of a hack.
[04:56] <glatzor> Chipzz: I hope that novice users will use gnome-app-install to install new apps, so they won't see the s-p at all
[04:57] <Kinnison> Kamion: pushed, take a peek
[05:11] <Keybuk> dholbach: ping
[05:11] <Chipzz> glatzor: anyway, I'm working atm, will look at it later
[05:33] <danimo> BenC: bad luck, compilation fails in the hda directory. investigating
[05:39] <danimo> BenC: sound/pci/hda/patch_analog.o fails to build
[05:40] <danimo> BenC: sound/pci/hda/patch_analog.c:630: error: syntax error before 'ad1986a_laptop_eapd_mixers'
[05:40] <BenC> ok
[05:41] <danimo> BenC: my C knowledge is a bit rusty, plus I don't know alsa too well. would be great if you had a look
[05:43] <BenC> I will, thanks
[05:44] <mjg59> BenC: That bcm compile fix looks bogus
[05:44] <mjg59> It's supposed to be checking the revision of the phy, not the core
[05:44] <danimo> BenC: just ping me by the time you sorted things out, and I will git pull the update
[05:44] <danimo> BenC: being without sound sucks ;)
[05:45] <mjg59> BenC: Probably ought to be bcm43xx_current_phy(bcm)->rev == 8
[05:45] <BenC> mjg59: I didn't see any other references to the same thing
[05:45] <mjg59> BenC: No, there aren't any
[05:45] <mjg59> Hm. Let me just recheck that to make sure.
[05:47] <mjg59> BenC: Nope, definitely supposed to be checking the phy revision, not the core revision
[05:47] <BenC> ok, changed
[05:48] <mjg59> Thanks
[05:49] <mjg59> BenC: dce0ca36f2ae348f005735e9acd400d2c0954421 - it seems to have got 945 support in the same diff 
[05:49] <mjg59> (It's labelled as the powernow-k8 fix)
[05:50] <BenC> weird, goof on my part then
[05:52] <Kamion> Kinnison: much better, thanks! merging and pushing
[05:53] <dholbach> Keybuk: pong
[05:53] <Keybuk> dholbach: on Palm Pilots
[05:53] <Keybuk> does my comment make sense?
[05:54] <dholbach> Keybuk: didn't read it yet - which bug number was that (just if you have it)
[05:54] <dholbach> got it
[05:54] <Keybuk> basically I'm suggesting that instead of having the current system
[05:54] <Keybuk> the ideal UI is pretty much Serpentine/SJ on CD writers
[05:54] <Keybuk> use HAL to look for USB-connected pilots
[05:54] <Keybuk> then give them a drop-down like:
[05:55] <Keybuk> Device: [ Palm Pilot 9000 on USB           ] 
[05:55] <Keybuk>         : Palm Pilot 8200 on USB           :
[05:55] <Keybuk>         : Pilot plugged into Serial Port 1 :
[05:55] <Keybuk>         : Pilot plugged into Serial Port 2 :
[05:55] <Keybuk> where the first two were "detected" automatically, and the other two just hard-code /dev/ttyS0, /dev/ttyS1 for stuff we can't detect
[05:55] <Keybuk> that way we don't need /dev/pilot, and I can get rid of it
[05:56] <dholbach> hrmmmmm, since upstream is not very active, this is the "ideal" case
[05:56] <Keybuk> (I'm likely to get rid of /dev/pilot anyway, as it doesn't work)
[05:58] <Kinnison> Kamion: you're welcome
[05:58] <Kinnison> Kamion: anything else I can do for you UI-wise?
[05:59] <Kamion> Kinnison: pretty much anything in the "braindump from UI sprint" section of UbuntuExpress/ToDo is fair game
[06:00] <Kamion> some of those should be reasonably accessible, some are harder - should be fairly obvious which, I think
[06:00] <Kinnison> Kamion: I'd like to get this g-p-m upload finished since I've been working on it for nearly four days (with interruptions :-)
[06:00] <Kinnison> Kamion: once I've got this done, I'll go look at the ToDo
[06:01] <dholbach> thanks Keybuk, i just thought we might apply some udev magic and make palm devices happy again
[06:01] <Kamion> Kinnison: sure, fortunately I'm actually beginning to run out of HYPER-URGENT things
[06:01] <Keybuk> dholbach: sadly there is no udev magic
[06:01] <Keybuk> ignoring not-on-USB pilots for a moment, which don't get the symlink
[06:02] <Keybuk> it seems that USB pilots contain two serial port interfaces
[06:02] <Keybuk> only one of which is useful for gnome-pilot
[06:02] <Kamion> from a functionality point of view, the timezone->country->locale logic, apt-setup, and network configuration are the missing pieces I know of
[06:02] <dholbach> Keybuk: ok, thanks
[06:02] <Keybuk> because of the way these things work, the events for each serial port device get processed by a separate udevd process
[06:02] <Keybuk> so without having some kind of "BIG UDEV LOCK" there's no way to provide "enumeration" of them
[06:02] <Keybuk> (which is why we can't just have /dev/pilot, /dev/pilot1, etc.)
[06:03] <Keybuk> and helpfully, there's no difference in identification between the two interfaces
[06:03] <Keybuk> they're just both identified as a serial interface
[06:03] <Keybuk> so there's nothing we can put in the udev rule to select the right one
[06:03] <Keybuk> and even better, it's usually the second one
[06:03] <Keybuk> so we could just do "only make the symlink for odd-numbered interfaces"
[06:04] <Keybuk> but that obviously breaks if you have a USB serial converter plugged in first <g>
[06:04] <Keybuk> as then the pilot would have ttyUSB1 and ttyUSB2
[06:04] <Keybuk> the icing on the cake is that old pilots have just one serial interface
[06:04] <Keybuk> *breathe*
[06:04] <Keybuk> so yeah, that's why /dev/pilot is broken, and why I don't think it can be fixed
[06:05] <Keybuk> (it's also why /dev/cdrom, /dev/dvd, etc. are broken if you have multiple drives)
[06:05] <dholbach> poor palm users :/
[06:05] <Keybuk> and I'm more inclined to suggest the software is fixed in the first place
[06:06] <mjg59> The applications should just be fixed to try all available pilot devices
[06:06] <mjg59> Palms have names, IIRC
[06:06] <Keybuk> mjg59: not only do palms have names, but they have unique vendor and device ids for each model
[06:06] <Keybuk> so the software could even intelligently offer features if it bothered to use HAL
[06:07] <mjg59> Never mind. I'm sure opensync will make everything ok.
[06:08] <xhaker> if there was a "we need a hero" song, it would suit this exact moment
[06:08] <mjg59> http://gregdek.livejournal.com/4008.html is cool
[06:08] <xhaker> mjg59: i don't find it cool, except the ubuntu love in the comments
[06:09] <xhaker> mjg59: you mean that or the OIN?
[06:09] <mjg59> The OIN stuff
[06:10] <mjg59> It'll be interesting to see what else they're willing to go for
[06:11] <xhaker> somehow, i feel they're playing the "silly-patent-game", but yeah.. aslong as they protect
[06:12] <Keybuk> until we (the Dental community) play the silly patent game, and excel at it, we can't expect to influence the rules
[06:13] <xhaker> Keybuk: cool laying down of things. I guess it's the option we have. Ohh, Sony is part of OIN :S
[06:14] <dholbach> hum, how can I tell gdb to write output to a file?
[06:14] <xhaker> redirect it? gdb .... > file.. sorry if this sounds silly, not used to console debugging
[06:15] <Keybuk> which kind of output?
[06:15] <Keybuk> output of the process?
[06:15] <dholbach> backtrace
[06:15] <Keybuk> output of debugging session?
[06:15] <Keybuk> set logging file FILENAME
[06:15] <dholbach> i tried to feed it something like    --command="/usr/share/bug-buddy/gdb-cmd"
[06:15] <dholbach> ahhhh
[06:16] <Keybuk> set logging on too
[06:16] <Keybuk> (my favourite gdb trick for debugging curses apps is setting the process output to a different tty :p)
[06:16] <Chipzz> Keybuk: how do you do that?
[06:17] <Keybuk> tty TTYNAME
[06:17] <Keybuk> oddly enough
[06:17] <dholbach> Keybuk: thanks a lot
[06:17] <Keybuk> handiest with screen
[06:17] <Chipzz> Keybuk: or with dual head ;)
[06:17] <Keybuk> indeed
[06:18] <Chipzz> ie 'tty /dev/tty2' in gdb?
[06:18] <Keybuk> yup
[06:19] <Keybuk> with screen it's handiest because you can create a shell-less tty
[06:19] <xhaker> Keybuk: what if you set logging filename to /dev/tty2 ?
[06:19] <Keybuk> xhaker: that works too -- except then you can't interact with gdb *and* your debugged app
[06:19] <xhaker> Keybuk: thanks
[06:20] <Keybuk> (assuming curses)
[06:20] <Keybuk> if not curses, then logging to another tty is quite acceptable
[06:22] <danimo> BenC: I think I can see what went wrong
[06:22] <danimo> BenC: first of all, there's a "struct" missing in line 630, that one was easy
[06:22] <Keybuk> meh, I have too many "Personal Projects" to start, and too little time or motivation for any of them
[06:23] <danimo> BenC: but that doesn't seem to be everything
[06:24] <sladen> kinnison/mdz: Can I have a UVF for powernowd (0.96 in Ubuntu, 0.97 in Debian) to fix power-saving on SMP systems (eg. all the new dual cores)
[06:24] <danimo> BenC: ah_ *t_* vs *.t:630
[06:25] <Kamion> sladen: send mail to me/mdz
[06:25] <Kamion> er, me+mdz
[06:25] <Kamion> with changelog
[06:28] <danimo> oh, strike the struct, just reverse the _t mistake :)
[06:34] <BenC> danimo: s/t_new/new_t/
[06:34] <danimo> BenC: yes, exactly
[06:34] <danimo> BenC: I just realized that the struct is superflous because the definition was turned into a typedef
[06:34] <danimo> BenC: compiles now, please apply the change
[06:35] <BenC> done, pushing
[06:35] <danimo> tnx
[06:35] <BenC> thanks for tracking it down
[06:36] <danimo> no problem
[06:36] <danimo> BenC: it's just funny to see how upstream likes to mess with datastructures just inbetween minor releases
[06:36] <danimo> that said, their new way of doing it is of course cleaner
[06:39] <danimo> <- is a C++ whimp anyway
[06:39] <marseillai> hi! i've report this bug https://launchpad.net/distros/ubuntu/+source/network-manager/+bug/36086 and i want to know if i can do something more to help ...........
[06:39] <Ubugtu> Malone bug 36086 in network-manager "network manager can't activate eth1 interface" [Normal,Unconfirmed]  
[06:43] <Kamion> pips1: sorry, forgot to get back to you earlier
[06:43] <Kamion> pips1: which installer - the traditional text-mode installer on the install CD?
[06:43] <ogra> Kamion, edubuntu
[06:44] <ogra> the traditional one ...
[06:45] <Seveas> marseillai, just wait for a developer to spot your bug and place a reaction...
[06:45] <Kamion> pips1: OK, this is my fault - I forgot to include pkgsel in the list of packages I included in the merged debian-installer translation files
[06:46] <Kamion> pips1: I'll get that corrected straight away. Thanks for the heads-up!
[06:46] <marseillai> oki Seveas thanks
[06:46] <pips1> Kamion, ah, ok, cheers
[06:48] <mdke> is it known that OOo writer crashes whenever you scroll with the mouse or sidebar?
[06:49] <giftnudel> mdke: not for me
[06:49] <mdke> hrph
[06:50] <mdke> it's hard to remember not to scroll
[06:50] <giftnudel> what do you do? just scroll?
[06:50] <ogra> get a bigger display ? 
[06:50] <Burgwork> mdke, there is a bug filed
[06:50] <mdke> lol
[06:50] <mdke> Burgwork, ah good
[06:51] <mdke> got the number handy?
[06:51] <Burgwork> just a sec
[06:51] <mdke> i'll search, np
[06:51] <Burgwork> https://launchpad.net/distros/ubuntu/+source/openoffice.org/+bug/36121
[06:52] <Ubugtu> Malone bug 36121 in openoffice.org "crashes on more than one page" [Normal,Needs info]  
[06:52] <Diziet> dholbach: ping
[06:52] <dholbach> Diziet: pong
[06:52] <Diziet> dholbach: You were asking for a firefox-dbg package.
[06:52] <dholbach> yes
[06:52] <dholbach> It'd make the world a happier place.
[06:52] <Diziet> There are basically two ways of doing that that aren't mad.
[06:52] <Diziet> Firstly, make the existing build system build everything twice.
[06:53] <Diziet> Secondly, make a script which transforms a firefox source package into a firefox-dbg source package.
[06:53] <dholbach> what about   dh_strip --dbg-package=firefox-dbg  ?
[06:53] <Diziet> Neither of these are very nice :-).
[06:53] <Diziet> To debug it properly you want to compile it with -O0.
[06:53] <Diziet> And against debug/static versions of the other libs.
[06:54] <Diziet> But I could do the --dbg-package thing fairly easily I suppose.
[06:54] <dholbach> I'm not quite sure I understand the technicalities behind that, but wouldn't be the dh_strip-version a good first step? before building a noopt version?
[06:54] <Diziet> Right.
[06:55] <Diziet> (Is --dbg-package new ?  I remember us talking about this kind of thing in Docklands.  I didn't realise it had been put into debhelper.)
[06:55] <dholbach> I can't remember it not being there, but that doesn't say anything :))))))
[06:55] <mdke> thanks Burgwork 
[06:56] <Diziet> Well, thanks for pointing it out :-).
[06:56] <Kamion> Diziet: older than the Docklands sprint, but relatively new
[06:56] <Burgwork> mdke, np
[06:57] <Kamion> hmm, October 2003. shows what my notion of "relatively new" is I guess ...
[06:57] <Diziet> :-)
[06:57] <dholbach> Kamion: that's why I can't remember a state of the world without it ;-)
[07:06] <danimo> BenC: works now as advertised
[07:06] <BenC> nice
[07:06] <dholbach> Diziet: thanks for looking into it
[07:06] <danimo> BenC: incl. audio jack sensing
[07:08] <Diziet> Anyone here have an ia64 box ?
[07:09] <dholbach> bddebian: it's nice if your heating is broken :)
[07:09] <Diziet> OK, failing that, I'll get the admins to install it on halley for me.
[07:09] <bddebian> dholbach: :-)
[07:10] <slomo> infinity, Keybuk: what's that? two uploads with the same version and different changes? ;) which one will be used now? (initramfs-tools)
[07:11] <Kamion> that would be ... a Soyuz bug!
[07:11] <Keybuk> slomo: We  Soyuz!
[07:11] <Keybuk>  _  ___           _              _
[07:11] <Keybuk> | |/ (_)_ _  _ _ (_)___ ___ _ _ | |
[07:11] <Keybuk> | ' <| | ' \| ' \| (_-</ _ \ ' \|_|
[07:11] <Keybuk> |_|\_\_|_||_|_||_|_/__/\___/_||_(_)
[07:11] <Kamion> in fact, it's bug 31038
[07:11] <Ubugtu> Malone bug 31038 in launchpad-upload-and-queue "two accept messages for different udev 079-0ubuntu9 uploads" [Normal,Unconfirmed]  http://launchpad.net/bugs/31038
[07:11] <Keybuk> (bet that doesn't trigger his notify)
[07:11] <dholbach> hahahaha
[07:12] <infinity> Wow, where do I get hilight-on-figlet technology for my IRC client?
[07:12] <Kinnison> infinity: hire a lackey to watch channels for you and tell you when someone figlets
[07:12] <Treenaks> http://www.audiorelief.co.uk/shop/product_info.php?products_id=31
[07:13] <Kinnison> Keybuk: known bug, blahblahblah, probably the latter, no way to know until the archive cycles mumblemumblemumble, sorry
[07:13] <ogra> Treenaks, thanks
[07:14] <Keybuk> Kinnison: heh
[07:14] <infinity> Kinnison: S'ok, there was apparently an ubuntu27 uploaded shortly after the clashing ubuntu26s, so we know what will win this time. :)
[07:14] <Keybuk> infinity: I'm not sure we do actually know that <g>
[07:14] <infinity> Keybuk: Yours just hit -changes...
[07:15] <infinity> Keybuk: You did the same thing... Twice? :)
[07:15] <infinity> Keybuk: Oh, 26 was doing prereqs...
[07:16] <Keybuk> right, 26 I modified the lvm and md scripts in the package
[07:16] <Keybuk> which I realised meant initramfs-tools had to depend on udev again
[07:16] <Keybuk> because of the PREREQ bug
[07:16] <Keybuk> so I decided it was time to purge them
[07:16] <infinity> Heh.
[07:16] <Keybuk> thus 27, which moved the scripts to the right packages and dropped them (and a boatload of deps) from initramfs-tools
[07:17] <infinity> Keybuk: Want to send me the -27 .dsc and .diff.gz, so I can do -28 and upload it before the archive cycle?
[07:18] <Keybuk> infinity: I already did
[07:18] <Keybuk> infinity: see #ubuntu-kernel
[07:18] <Keybuk> (and it's a .tar.gz, because initramfs-tools is native, no? :p)
[07:18] <infinity> Ah, rock.
[07:18] <infinity> Err, yeah.  tar, diff, whatever.
[07:18] <infinity> Brain not all here.
[07:18] <infinity> Been up for 22 hours or so.
[07:21] <Keybuk> Kamion: may have just come across a bug caused by the installer not recognising certain ethernet/wifi cards
[07:22] <Kamion> Keybuk: similarly, I trust you've checked which of your two libnjb uploads actually hit the archive?
[07:22] <Kamion> Keybuk: any idea why?
[07:22] <Keybuk> Kamion: I didn't think we'd hit a daily cycle since then yet
[07:22] <ogra> Von: 	Doug <899sergio@fcta.com>
[07:22] <ogra> An: 	daniel.holbach@ubuntu.com
[07:22] <ogra> Betreff: 	Hello, new job for you
[07:22] <ogra> Datum: 	Thu, 23 Mar 2006 14:56:19 +0300  (12:56 CET)
[07:22] <ogra> Mailer: 	Microsoft Office Outlook, Build 11.0.5510
[07:22] <Keybuk> Kamion: not sure yet; need some details.
[07:22] <Kamion> Keybuk: we certainly have, that was hours ago
[07:22] <ogra> dholbach, hey, i'm getting your spam !
[07:23] <dholbach> tssssss
[07:23] <dholbach> ogra: what will my new job be like?
[07:24] <ogra> but if you want to channel several billions from a african bank through your account i can forward it ;)
[07:24] <ogra> Requirements:
[07:24] <ogra> - basic knowledge of computer (email),
[07:24] <ogra> - free time (approx 1 hour per day),
[07:24] <ogra> - good communication skills
[07:24] <ogra> :)
[07:24] <ogra> - bank account to withdraw/receive funds,
[07:24] <ogra> - ability to cash money orders and certified checks (only for USA, UK and DE),
[07:25] <ogra> *giggle*
[07:26] <dholbach> ogra: :)
[07:26] <Keybuk> Kamion: the bug seems a little strange
[07:27] <Keybuk> the user had an /etc/iftab line that caused udev to hang, because they had two devices that matched the rule
[07:27] <Keybuk> I thought I'd explicitly checked for that in netcfg
[07:28] <Keybuk> Kamion: is there a way I can get the user to run that bit of netcfg from an installed system?
[07:30] <Kamion> no, sorry
[07:48] <mjg59> I note that nobody is subscribed nvidia-glx bugs by default...
[07:50] <Ubugtu> Malone bug 36296 in gnome-power-manager "Problem with the systray and gnome-power-manager" [Normal,Unconfirmed]  http://launchpad.net/bugs/36296
[07:50] <ogra> somehow it stops making sense to me very precisely in the middle ...
[07:50] <infinity> mjg59: I thought I got all LRM bugs..?
[07:51] <mjg59> infinity: Hm. I just assigned something to nvidia-glx, but it didn't seem to add you
[07:52] <Pygi> infinity: seems that new patched L-R-M don't work??
[07:52] <infinity> Meh, oh well.  I can't actually FIX nvidia-glx bugs anyway. :)
[07:52] <mjg59> infinity: Some time, we should really get in touch with nvidia and "suggest" that we get a direct route to their development people
[07:52] <infinity> Pygi: That's what Seveas was saying... But it's the exact same patch I applied when I built the other packages for you, so I'm at a loss.
[07:53] <Pygi> infinity: perhaps something in kernel makes it stop workin'? :-/
[07:53] <infinity> Pygi: It's concievable.
[07:53] <infinity> We'll have to look deeper when it's not 6am. :)
[07:53] <Keybuk> nvidia-glx scares me
[07:54] <Pygi> infinity: k, great ... please ping me once you have time, ok ? ^_^
[07:54] <Keybuk> mostly because I have to use it, because the default X drivers won't even detect my gfx card
[07:54] <mjg59> (Rather than nvidia responding to my bug reports with a job offer)
[07:54] <KaiL> uhm,. who (re)added xscreensaver to ubuntu-desktop? ;)
[07:54] <mjg59> ogra: I /think/ it means "When I boot on battery, the systray icon doesn't appear"
[07:55] <ogra> KaiL, nobody
[07:55] <KaiL> ogra, at least ubuntu-desktop depends on both, gnome-screensaver (running) and xscreensaver
[07:55] <ogra> mjg59, yeah, it was the combination with the next sentence that totally confused me
[07:55] <infinity> KaiL: No it doesn't...
[07:56] <KaiL> uh, wait..
[07:56] <ogra> KaiL, that would be news to me
[07:56] <Keybuk> KaiL: doesn't here
[07:56] <KaiL> it's only the data - sorry :)
[07:56] <Keybuk> are you sure you're not getting confused by the fact it depends on the xscreensaver *hacks*, not the daemon itself
[07:56] <Amaranth> mjg59: They offered you a job? :)
[07:56] <ogra> unless mvo secretly added it today ;)
[07:57] <mjg59> Amaranth: They sent me an email titled "Resume", but meant "Resum"
[07:57] <Burgwork> mjg59, ogra I think that is my rant on g-p-m on hiding the icon when battery is full and plugged in
[07:57] <infinity> mdz: ping
[07:57] <mjg59> Burgwork: No
[07:57] <sivang> mjg59: joining canonical ? :)
[07:57] <Kamion> janimo: hmm, can you do the same libxml-perl -> libxml-parser-perl change to xfmedia, if possible?
[07:57] <mjg59> Burgwork: And that's an option in the preferences
[07:57] <mjg59> sivang: Nope
[07:57] <Kamion> sivang: 18:54 < mjg59> (Rather than nvidia responding to my bug reports with a job offer)
[07:57] <ogra> Burgwork, "To see it, you have to plug and unplug your ubuntu box to the sector." 
[07:57] <janimo> Kamion,sure
[07:57] <Kamion> ta
[07:57] <ogra> Burgwork, did you write that ? 
[07:57] <sivang> Kamion: :-)
[07:57] <sivang> mjg59: wow! 
[07:58] <Burgwork> ogra, umm. I spaek Engrish beetter than that
[07:58] <Keybuk> mjg59: did you give them the same "I'd rather create an army of fruit flies to do my bidding than work for you" speech?
[07:58] <ogra> Burgwork, right ...
[07:58] <mjg59> Keybuk: Not quite...
[07:58] <Kamion> sivang: not to run mjg59 down or anything, but it's not too uncommon - I've had job offers from two companies I've worked with on Ubuntu stuff since joining Canonical
[07:59] <Keybuk> Kamion: and how many of those were search engines? :p
[07:59] <Kamion> *cough*
[07:59] <mjg59> Haha
[07:59] <Kamion> (that would be one)
[07:59] <mjg59> Google were attempting to hire the entire world at LCA
[07:59] <Keybuk> again?
[07:59] <Keybuk> they were trying to do that in Canberra too
[07:59] <mjg59> There were four of them there this year
[07:59] <Treenaks> Guess they're doing it everywhere then
[08:00] <mjg59> They bought beer for everyone at the conference one evening
[08:00] <mjg59> All night
[08:00] <Kamion> mjg59: jesus
[08:00] <mjg59> Only worked out at about 4000NZD, so...
[08:00] <Kamion> so thruppence or so?
[08:00] <Amaranth> o_O
[08:00] <Amaranth> dude, i need to go to these conferences
[08:00] <Amaranth> free beer from google!
[08:01] <mjg59> I think I managed free beer every night
[08:01] <Treenaks> Amaranth: in exchange for your most intimate details for them to search through :P
[08:01] <mjg59> And a helicopter ride
[08:01] <Keybuk> siretart: ping
[08:01] <Amaranth> that's why you do things for people throughout the year, so everyone owns you a beer
[08:01] <Treenaks> mjg59: Yeah but you're kind of a well-know guy...
[08:01] <sivang> Kamion: Well, that's because you guys are of this caliber :)
[08:02] <mjg59> Kamion: Pretty much
[08:04] <sivang> Kamion: what were you working with nvidia on regarding ubuntu? 
[08:04] <mjg59> So, uh, who moderates ubuntu-announce?
[08:04] <Treenaks> jdub
[08:04] <infinity> Okay, can someone tell me why the world is trying to hax0r my sshd all of a sudden?
[08:04] <dholbach> Treenaks: i realized what you meant an hour ago... sorry :)
[08:05] <dholbach> Treenaks: i mean the other spam mail 1-2-3 weeks ago
[08:05] <dholbach> s/other//
[08:05] <Treenaks> dholbach: oh :)
[08:05] <Pygi> infinity: bah, probably some random scripts....
[08:05] <Kamion> sivang: I wasn't; nvidia never offered me a job
[08:05] <KaiL> hmm, shouldn't gnome-screensaver allow at least a _few_ configurations for the screensaver? ;)
[08:06] <Kamion> mjg59: apparently jdub on not remotely enough coffee, repeatedly
[08:06] <mjg59> sivang: Me? I just sent them a couple of bug reports...
[08:06] <infinity> Pygi: Oh, I'm sure hitting /my/ IP is random, but it's usually not random when I hear no activity for weeks, then thousands of attempts in a day.
[08:06] <sivang> mjg59, Kamion : okay, sorry for the noise, going back to my hacking now :)
[08:07] <Pygi> infinity: just make sure you change the port to some "not normal", for example 1022 or somethin'
[08:07] <infinity> Pygi: <shrug>.. I'm just firewalling it off completely for now.
[08:07] <Pygi> infinity: ah, changing port makes it good enough...
[08:15] <sivang> Kamion: hmm, acutally thinking of it one search engine does have a branhc over ireland somewhere :)
[08:18] <Kamion> sivang: that would be more useful if I lived in Ireland though ...
[08:23] <Keybuk> Kamion: pcmciautils missing dep on lsb-base
[08:25] <Keybuk> not that I guess it's necessary these days ... given an Essential package depends on it
[08:25] <Keybuk> but still
[08:26] <Kamion> not really, given that it has compatibility code for the installer
[08:27] <Kamion> I can put the dependency in if you *really* want, but it's a bit false
[08:27] <Keybuk> right
[08:28] <Keybuk> I just noticed it because I was looking for whatever versioned-dep we need this week :p
[08:28] <Keybuk> and I picked on that because it was reasonably recent and I trusted you to get it right <g>
[08:28] <Kamion> if it needs a versioned dep, I can add that with more justification, I guess
[08:28] <Keybuk> you use log_action_begin_msg which was added in 3.0-6 or wherever
[08:29] <Kamion> oh, heh, that's quite different in Debian, I guess Per rearranged that
[08:29] <Kamion> yeah, 3.0-6, will add the dep
[08:34] <Riddell> Kamion: how come the powerpc live CD has a bunch of extra language packs the other arches don't?
[08:34] <ogra> Riddell, look at your seeds :)
[08:34] <pitti> Riddell: the seeds support per-arch selection
[08:34] <Riddell> ogra: that's what I'm querying
[08:35] <Kamion> Riddell: because at one point that made sense for size reasons
[08:35] <pitti> Riddell: when we released breezy, we just filled all CDs up to the rim
[08:35] <Kamion> it probably doesn't any more
[08:35] <ogra> i reverted it for edubuntu, made ppc behave way more sane
[08:35] <pitti> Riddell: I have a script which calculates kubuntu and ubuntu cumulative size for language packs, which can be used to figure out the seeds
[08:35] <pitti> in case you need it
[08:36] <pitti> (it also respects our 'priority' langauges)
[08:36] <ogra> pitti, i'd be intrested ...
[08:36] <Riddell> I'm also wondering why Bengali seems to be a priority langauge, do we have a lot of Bengali speaking users I don't know about?
[08:36] <Kamion> it's by count of speakers
[08:36] <pitti> ogra: http://people.ubuntu.com/~pitti/bzr/langpack-o-matic/langpacksize
[08:36] <ogra> pitti, TA !
[08:37] <Keybuk> pitti: when you're done with libnl, https://wiki.ubuntu.com/MainInclusionReportWpaSupplicant
[08:38] <pitti> oh, sorry for dragging that, will do now
[08:38] <pitti> Keybuk: you are fine with wpa-supplicant's quality? last time I tried it it just fell apart
[08:39] <Diziet> Gah, debugging the `clean' target in the firefox build system is really unpleasant.  Everything takes aaaages.
[08:40] <Kamion> Riddell: according to ethnologue there are about 70 million Bengali speakers, which is roughly on a par with German
[08:40] <Kamion> although actually the d-i languages.xml file has a different figure, it says 189 million
[08:40] <Kamion> I'm inclined to trust d-i on this, I know a lot of work has gone into that file
[08:40] <pitti> ISTR that it were way more, too
[08:41] <Riddell> wikipaedia says 196 million native,
[08:41] <pitti> Kamion: but it probably depends on the source
[08:41] <Keybuk> pitti: meh, I don't *like* it ... but it's a dependency of n-m 0.6
[08:41] <Kamion> German is at 101 million, for comparison
[08:41] <Kamion> pitti: that's a "corrected" figure
[08:41] <Keybuk> and it does work better than xsupplicant in my experience
[08:41] <pitti> Keybuk: can't tell, neither worked for me at debconf 5 :)
[08:41] <Riddell> Keybuk: what's the status of n-m 0.6?
[08:41] <pitti> Keybuk: they broke differently
[08:41] <Keybuk> wpasupplicant was the one that worked for me at debconf
[08:41] <KaiL_> Riddell, afaik nobody really know, how many they are..
[08:41] <janimo> wikipedia says 270 millionranked as the 4th most spoken
[08:41] <Kamion> and the number of native English speakers is listed at around 309 million
[08:42] <Keybuk> and with NM providing the configuration, it has worked perfectly here today in all my testing
[08:42] <KaiL_> but aren't that mostly very poor people?
[08:42] <pitti> Keybuk: I guess it just doesn't at powerpc then
[08:42] <KaiL_> I miss spanish a bit on that list...
[08:42] <Keybuk> it may be that it doesn't work with linux-wlan-ng of course
[08:42] <pitti> Keybuk: but that would be fine for me, if it works on i386
[08:42] <Keybuk> there's no specific mention of wlan-ng in the source
[08:42] <Kamion> KaiL_: on principle we're not discriminating against language support just because the speakers happen to be poor
[08:42] <KaiL_> oh, there es _is_ prioritized - only not let landed on the live-CD..
[08:43] <pitti> Keybuk: if WPA is driver specific, that might be, too
[08:43] <Kamion> es is the next one down after fr
[08:43] <Keybuk> pitti: it has to be supported by the driver, fwict.  see the patch to madwifi
[08:43] <Kamion> we'll probably add it and a few others in too
[08:43] <KaiL_> Kamion, the question is only about how big is the chance, that this lang is searched for..
[08:44] <pitti> Keybuk: could you test libnl on any !i386 arch?
[08:44] <Keybuk> pitti: I can give it a go on powerpc this evening
[08:44] <pitti> would be interesting
[08:45] <Kamion> KaiL_: consider that Canonical and related organisations have put a fair bit of effort into getting free software and Ubuntu into areas not traditionally known as high-tech; look at the Freedom Toaster and the Xhosa translation work
[08:45] <Lure> pitti, Keybuk: will wpasupplicant also include patches required for n-m (in test repository)
[08:45] <Keybuk> is with a linksys orinoco pcmcia card, would be a fun test
[08:45] <Keybuk> Lure: see -changes
[08:45] <Keybuk> (yes)
[08:45] <Keybuk> Kamion: *three* dailies today?  what broke? <g>
[08:45] <pitti> Keybuk: the world will *love* you for that :)
[08:45] <Kamion> Keybuk: the third was Mithrandir's, I think - no idea
[08:46] <Kamion> Keybuk: the first was broken code on cdimage, my fault
[08:46] <KaiL_> Kamion, right, but what is more usefull: a lang with 50Mio speakers, of which 805 have a PC or a lang with 100Mio speekers, of which maybe 5% have a PC..?
[08:47] <Keybuk> Kamion: uhhh, what versioned dep on module-init-tools did you just add? :p
[08:47] <pitti> Keybuk: ok, I trust you for the QA assessment; libnl is fine
[08:47] <Keybuk> ah, just >= 3.2.2 that's ok then
[08:48] <Kamion> KaiL_: how about a language with 100 million speakers, of which not many have a PC yet, but when they do, guess who shows up with full support for them
[08:48] <Kamion> look to the future! :)
[08:49] <Evaso2> hi guys there is islsm driver packaged in ubuntu?
[08:49] <pitti> janimo: evince-gtk? yay source code duplication.. :/
[08:49] <siretart> Keybuk: pong
[08:49] <Keybuk> Evaso2: not that I'm aware; what is it?
[08:49] <janimo> pitti, yes unfortunately
[08:49] <janimo> there;s more of that kind
[08:49] <KaiL_> Kamion, good idea, if there'a chance, that this happenes in a to to far future..
[08:49] <Keybuk> siretart: ETOOLATE :p
[08:49] <siretart> hehe
[08:50] <Kamion> plus I don't think we can objectively measure number of speakers with PCs; you mentioned Spanish, of which there are a large number of poor speakers in Latin America
[08:50] <janimo> pitti, we talked about this about two weeks ago
[08:50] <Kamion> and yet Brazil is one of the centres of Canonical development
[08:50] <Keybuk> siretart: I just uploaded wpasupplicant with your ctrl_ap patch, a prettyfied init script and wpagui split out
[08:50] <Evaso2> Keybuk: is the prism54 driver for the softmac cards
[08:50] <Kamion> so go figure
[08:50] <siretart> was it about wpasupplicant? I'm currently preparing the next upload to debian
[08:50] <Keybuk> Evaso2: yes.
[08:50] <pitti> janimo: no chance for multi-building the evince package with a different patch set or so?
[08:50] <pitti> janimo: or, rather, configure options?
[08:50] <Keybuk> siretart: wanted to check there was nothing I missed, and it was all ok
[08:50] <siretart> Keybuk: wpagui is split out for some time in debian. I thought that slomo already fakesynced that work some time ago
[08:50] <janimo> pitti, seb did not want that
[08:51] <janimo> and also cdbs does not cope with multibuild
[08:51] <Evaso2> Keybuk: so there isn't support for it in ubuntu?
[08:51] <janimo> AFAIK
[08:51] <Keybuk> siretart: not that I could tell, the latest source from ubuntu just had it in the main thing
[08:51] <Keybuk> Evaso2: prism54 is in Ubuntu
[08:51] <crimsun> siretart: no, he asked me to, but I didn't due to the bugs in BTS
[08:51] <pitti> janimo: not very well, true
[08:51] <slomo> siretart: no... crimsun wanted to do it
[08:51] <siretart> oh
[08:51] <siretart> I see
[08:51] <Evaso2> Keybuk: i'm talking about softmac
[08:51] <janimo> pitti, otherwise I'd try the same source package approach with gcalctool for example
[08:51] <crimsun> (0.4.7-4 is not ready for our main, so I didn't)
[08:51] <Keybuk> Evaso2: I don't know what that is, sorry
[08:51] <Evaso2> Keybuk: http://jbnote.free.fr/prism54usb/
[08:52] <siretart> well, I spent now about an hour fixing some debian bugs about wpasupplicant
[08:52] <Keybuk> siretart: Debian's package seemed somewhat different from Ubuntu's?
[08:52] <slomo> siretart: i asked him to do it as there were greater changes in packaging and you and crimsun were noted in the changelog. so i thought it would be far easier for one of you to merge the changes
[08:52] <siretart> Keybuk: we worked on getting it merged, so that ubuntu's package doesn't need to be diverged anymore
[08:53] <siretart> slomo: no problem. I wasn't aware that this hasn't happened yet
[08:53] <janimo> pitti, similar situation with xubuntu-system-tools which is very close to gnome-system-tools
[08:53] <janimo> on the plus side is still better than duplication by rewriting from scratch
[08:54] <siretart> Keybuk: I'll look at your upload in a minute and merge it with debian's svn
[08:54] <mjg59> Evaso2: Last time I checked, the prism54 softmac code didn't work. Has that changed?
[08:54] <sivang> janimo: who's the upstream authors?
[08:54] <siretart> Keybuk: I wanted to have 0.4.8 in ubuntu anyway, but havn't yet prepared a UVF exception request
[08:54] <KaiL_> wpa going into main? maybe even on the final dapper-CD? ;)
[08:54] <janimo> sivang, gnome devels genarlly I dunno
[08:54] <Kamion> er, Brazil => Portuguese of course, not Spanish. sigh, obviously not quite awake ...
[08:54] <Evaso2> mjg59: yes it is quite usable without encryption
[08:54] <Pygi> Kail_: hehe ;)
[08:54] <Keybuk> siretart: other than the prettyfying, I don't think there's anything interesting in the upload
[08:55] <siretart> Keybuk: you suggested that wpasupplicant should only run via pre-up.d scripts
[08:55] <Keybuk> I copied the patch from the DNM repository, and the control file snippet for wpa_gui
[08:55] <Keybuk> the rules changes were "minimal"
[08:55] <Keybuk> right, that was already done in Ubuntu
[08:55] <siretart> Keybuk: this doesn't cover all use cases for wpasupplicant
[08:55] <mjg59> Evaso2: Ok. If you can point me at the source for the working version, I'll see what I can do
[08:55] <Kamion> janimo: you might want to bump evince-gtk to 0.5.2, BTW
[08:55] <Keybuk> why not?
[08:55] <janimo> Kamion, yep planning that
[08:55] <siretart> Keybuk: there are people who want to use wpasupplicant as roaming daemon, waiting for the interface to come up
[08:55] <Keybuk> siretart: why is that a valid usecase though? :p
[08:56] <siretart> Keybuk: I've prepared a new package where there is a switch to choose what mode to use for wpasupplicant
[08:56] <siretart> Keybuk: as a replacement for waproamd
[08:56] <Keybuk> that's like saying dhclient3 should be running all the time waiting for the interface to come up <g>
[08:56] <Keybuk> ahh, I think we don't have to care about that in Ubuntu
[08:56] <Keybuk> "one true way to run it" is fine
[08:56] <Keybuk> and actually makes it a damned sight easier to integrate with other things
[08:57] <siretart> Keybuk: well. therefore I built a switch in /etc/default/wpasupplicant where the user can choose what mode he wants to use
[08:57] <siretart> Keybuk: it is NOT enabled by default. 
[08:57] <Keybuk> yeah, the trouble with that switch is it makes it harder for things like NM to start and stop wpa_supplicant and modify its configuration :)
[08:57] <Keybuk> and makes it harder for us to support it
[08:58] <Keybuk> as it means we have to find out how the user is running it
[08:58] <siretart> Keybuk: hm. how is nm actually starting nm?
[08:58] <siretart> argl
[08:58] <Keybuk> if we know that it's only run on a running interface, and only takes care of that single interface, like dhclient3, then it's easier to support
[08:58] <siretart> Keybuk: hm. how is nm actually starting wpasupplicant? via init script?!
[08:58] <Keybuk> directly I think
[08:58] <siretart> Keybuk: then thats no problem anyway
[08:58] <Evaso2> mjg59:  http://jbnote.free.fr/prism54usb/data/code/tarballs/islsm-workbench-latest.tar.bz2
[08:58] <pitti> Keybuk: wpasupplicant needs LSBified init scripts
[08:58] <Lure> siretart: n-m calls it directly
[08:58] <pitti> Keybuk: s/s$//
[08:58] <Keybuk> pitti: it does :p
[08:58] <Keybuk> pitti: wait for the next cron.daily
[08:58] <siretart> Keybuk: just leave the init scripts alone, they are disabled by default
[08:59] <pitti> Keybuk: ah, cool
[08:59] <Keybuk> siretart: well, I was going to take the init script and default file out
[08:59] <Keybuk> and then change the pre-up and post-down scripts to invoke and kill wpa_supplicant directly
[08:59] <Keybuk> depending on how that interacts with network-manager
[08:59] <pitti> Keybuk: ok, pacakging looks good, I'll do a quick source code review before I approve it (but after meeting...)
[08:59] <Evaso2> mjg59: firmware http://daemonizer.de/prism54/prism54-fw/
[08:59] <siretart> Keybuk: have you had a look at our integration in the experimental tree?
[08:59] <crimsun> Keybuk: that sounds sane enough.
[09:00] <Keybuk> siretart: has it changed from what's in Ubuntu?
[09:00] <Evaso2> mjg59: or here for an explanation of firmware http://jbnote.free.fr/prism54usb/PrismFirmwares.html
[09:00] <siretart> Keybuk: well, we support 3 modes there
[09:00] <Keybuk> the current stuff in Ubuntu is kinda buggy <g>
[09:00] <siretart> Keybuk: with the default being 'do not start it at all'
[09:00] <Keybuk> I'm not disagreeing that it's good in Debian to do that
[09:00] <Keybuk> however in Ubuntu I've been systematically purging all multi-mode operation from everything
[09:00] <siretart> Keybuk: one mode is as waproamd replacement, as ever in debian. the second is like you suggest
[09:00] <Keybuk> and making things only work one way
[09:00] <Evaso2> mjg59: do u need somthing else?
[09:00] <mjg59> Evaso2: No, that should be ok
[09:01] <mjg59> Oh, argh, another copy of the madwifi tree
[09:01] <siretart> Keybuk: the third is with direct integration to /etc/network/interfaces like wireless-tools
[09:01] <mjg59> WHY MUST PEOPLE BE SO MEAN?
[09:01] <Keybuk> siretart: isn't that just a pre-up/post-down script?
[09:01] <Keybuk> that's how wireless-tools intergrates with ifupdown
[09:01] <Pygi> mjg59: ask people? ;)
[09:01] <siretart> Keybuk: http://svn.debian.org/wsvn/pkg-wpa/branches/wpasupplicant-0.5/debian/README.modes?op=file&rev=0&sc=0
[09:02] <Keybuk> it just has if-up scripts that get environment variables
[09:02] <siretart> that file describes the planned state
[09:02] <siretart> mom phon
[09:02] <Keybuk> siretart: ok, I like the first one
[09:02] <Keybuk> I'd probably take that, and strip the other two from the package in Ubuntu
[09:02] <crimsun> well, seeing how I made the crufty if-p{re-up,ost-down}.d script, cut away
[09:03] <crimsun> (at the time I wanted to make the least invasive change to get it working)
[09:04] <mjg59> Keybuk: Ok, here's a fun one for you
[09:04] <mjg59> Keybuk: prism54 softmac and fullmac cards are identical except for the amount of RAM in them
[09:04] <mjg59> Keybuk: The softmac driver will drive both, but the fullmac driver is more functional
[09:05] <Keybuk> siretart: do you not think that just supporting one true mode is better?
[09:05] <Keybuk> even from a "I have to maintain this" point of view, it'd make your bug-triaging life much easier
[09:05] <siretart> Keybuk: mom, I'm on phon. just a sek
[09:05] <mjg59> Keybuk: The fullmac driver should refuse to load on softmac cards (since there's not enough ram for the firmware)
[09:05] <Keybuk> mjg59: heh, why is the driver different?
[09:05] <mjg59> Keybuk: softmac cards do host-ieee80211
[09:06] <mjg59> fullmac ones do it on the card
[09:06] <mjg59> Oh, no, hang on - the fullmac driver only loads firmware on ifup
[09:06] <mjg59> Hnngh

[09:06] <Keybuk> was going to mention that
[09:07] <mjg59> (Why does it not load firmware on modprobe? Why? Why? WHY?)
[09:07] <KaiL_> Network-Manager looks really nice...
[09:07] <Keybuk> mjg59: we have the source for that?  might be worth fixing
[09:07] <Keybuk> that causes no end of pain
[09:07] <mjg59> Keybuk: We have the source, yes
[09:08] <Keybuk> iirc. it has no MAC address until the firmware has been loaded
[09:08] <Keybuk> which is cute
[09:11] <Evaso2> mjg59: what kind of prism card do u havE=
[09:11] <siretart> ok. back
[09:11] <mjg59> Evaso2: None
[09:11] <siretart> Keybuk: I don't think that there is really 'one true mode'
[09:12] <mjg59> Oh horrid horrid driver
[09:12] <Keybuk> siretart: what does the always-running daemon give them?
[09:12] <siretart> Keybuk: I do agree that wpasupplicant should not be started via the init script by default
[09:12] <Keybuk> that the ifup/down mode doesn't?
[09:12] <siretart> Keybuk: roaming support. you don't need to care about where you are.
[09:12] <Keybuk> how does it give them that?
[09:12] <siretart> Keybuk: you just specify all your networks, and it just associates. thats neat
[09:13] <Keybuk> even when the interfaces are down? :p
[09:13] <siretart> Keybuk: when the interfaces are down, it waits for them to come up
[09:13] <Keybuk> right
[09:13] <Fawzib> hello, I'm trying to make a deb for a python application, but all the samples are for compiled apps not interpreted ones. Is there a skeleton debian/rules for python programs or a site with samples?
[09:13] <Keybuk> so if you put all the networks in a configuration file, and referenced that in /etc/network/interfaces using wpa-conf ... doesn't that give you the same behaviour?
[09:14] <siretart> Keybuk: the first mode in that README.modes does not give you roaming. thats only in the second mode
[09:14] <Keybuk> right
[09:14] <Keybuk> so what does the third mode give you that the first two don't
[09:14] <Keybuk> ?
[09:14] <siretart> Keybuk: the first one is intended for making it easier for gnome-system-admin to provide an interface to wpa
[09:14] <Keybuk> one question about the experimental package, btw
[09:14] <Keybuk> why does the svn include doc/ examples/ wpa_gui/ and wpa_gui-qt4 directories?
[09:14] <Keybuk> should those be in the diff.gz verbatim?
[09:15] <siretart> they come from the orig tarball
[09:15] <mjg59> Keybuk: Ok, I think I can manage this (so that the fullmac driver will fail on probe for softmac cards)
[09:15] <mjg59> Keybuk: So, can you then ensure that the fullmac driver is always loaded before the softmac one?
[09:15] <mjg59> (Will fail in the case of multiple cards, but, well)
[09:15] <Keybuk> siretart: hmm?  so I don't need to check those out?
[09:15] <Keybuk> mjg59: no, I have no way of forcing driver order at this time
[09:15] <mjg59> Keybuk: Well, we're going to need that
[09:15] <mjg59> (For the ahci case as well)
[09:16] <Keybuk> mjg59: like I said; I'd rather we fixed the modules <g>
[09:16] <siretart> Keybuk: you just check the svn out and build it with svn-buildpackage
[09:16] <Keybuk> assume I don't have that
[09:16] <mjg59> Keybuk: In this case, loading the softmac driver on a fullmac card is a perfectly reasonable thing to do
[09:16] <mjg59> It's just not what we /want/ to happen by default
[09:16] <Keybuk> I'm trying to understand the shape of the svn
[09:17] <Keybuk> the blah/blah/debian/ I'm used to, I just export the debian
[09:17] <Keybuk> but why are there the other trees next to it?
[09:17] <siretart> err, but this was a by hand upgrade, perhaps there has been some mistake while upgrade
[09:17] <siretart> I focused on the 0.4.8-1 upload. the experimental was kel's work from yesterday. need to catch up there
[09:18] <Keybuk> ok
[09:18] <Keybuk> cause I rather like the shape of that experimental one
[09:18] <Keybuk> and am gonna try for a UVF for it :)
[09:18] <siretart> well, we are not sure how many systems that is gonna break
[09:18] <siretart> therefore we are uploading that crack to debian/experimental and poke ppl to test that :)
[09:18] <Keybuk> no supported Ubuntu systems, wpa isn't in main yet <g>
[09:18] <siretart> glad that you like that work :)
[09:19] <mjg59> Keybuk: So what we want is for prism54 to be preferred over prism54_softmac, which would appear to be entirely a load order decision
[09:19] <fabbione> pitti: it is more complicate than that
[09:20] <fabbione> pitti: what i need is: query for bugs with conditions X/Y/Z -> get the list -> change status -> add comment
[09:20] <fabbione> pitti: there i no way to do that in LP for more than one bug
[09:20] <Keybuk> mjg59: but we can't guarantee the load order for people with one device of one type, and another of the other type
[09:20] <Keybuk> and trust me, SOMEONE will do this
[09:20] <mjg59> Keybuk: Yes. What do you propose instead?
[09:20] <siretart> Keybuk: whats the status with nm? can I look at your nm 0.6 branch?
[09:21] <mjg59> Keybuk: We can either make it work for most people, or we can leave it broken for everyone
[09:21] <pitti> fabbione: right, debbugs can search for way more properties than lp ATM
[09:21] <siretart> I'd love to see how I can help on the wpasupplicant side for integration
[09:21] <fabbione> pitti: only till a few hours back :)
[09:21] <pitti> fabbione: but it depends on what 'conditions' are
[09:21] <fabbione> pitti: my grep sk1ll5 are on it :)
[09:23] <Keybuk> siretart: I don't have a "branch", I just have a bunch of code on my disk
[09:23] <siretart> Keybuk: another option would be to integrate whats outlined in README.modes in an wpasupplicant_0.4.8-0ubuntu1 package 
[09:23] <fabbione> pitti: in about 100 lines of shell i have LP caching and local search in bugs headers :)
[09:23] <Keybuk> mjg59: like I said, I'd rather we fixed the modules to work whichever order they're loaded
[09:23] <siretart> but I'd rather see some more testing for that
[09:23] <Keybuk> siretart: given that the config changes are the bulk of the UVF request, I think that's probably defeating the spririt
[09:23] <siretart> I tried that today, its horrible to debug
[09:23] <pitti> fabbione: heh, I do that in my ubuntu-bugs maildir, easier :)
[09:23] <Keybuk> the upstream bit looks mostly like "fixed some stuff and added supprot for this algorithm"
[09:24] <siretart> nearly all options are case sensitive, and you won't get any assistance in debugging :/
[09:24] <mjg59> Keybuk: Which, in this case, would be how?
[09:24] <fabbione> pitti: assuming you have all of them.. yes..
[09:24] <fabbione> pitti: but not if you need to adopt something
[09:24] <mjg59> Keybuk: I mean, yes, ideally, the softmac driver should have all the functionality of the fullmac one. But it doesn't, and it won't in time for dapper
[09:24] <pitti> fabbione: ubuntu-bugs@ has them all, yes
[09:24] <Evaso2> pitti: i think that the dmraid udeb is available for debian. do u plan to backport it to ubuntu?
[09:24] <siretart> Keybuk: right. but the upgrade from 0.4.7 to 0.4.8 should be done anyway. quite focused bugfix release from upstream. the packaging and integration changes are disputable, sure
[09:25] <Keybuk> siretart: the packaging and integration changes are great though!  they make my life much easier <g>
[09:25] <Keybuk> mjg59: right
[09:25] <siretart> :)
[09:25] <pitti> Evaso2: I think you want to talk to Kamion for installer issues
[09:25] <infinity> Actually, me.
[09:25] <Keybuk> the trouble is I'm not sure we have time to patch and debug modprobe before dapper either <g>
[09:25] <pitti> or that ;)
[09:25] <infinity> That particular bug has been assigned to me (for now) for the initramfs integration stuff.
[09:25] <infinity> The installer stuff will come nearly for free (but not quite)
[09:26] <Evaso2> pitti: are u maintaining the user space tool?
[09:26] <mjg59> Keybuk: So we can either cripple the softmac driver so it can't load on fullmac hardware (which would be less than ideal, because it does have a couple of features the fullmac one doesn't), or we can ensure a specific load order
[09:26] <Keybuk> modprobe patch would need
[09:26] <Keybuk> 1) config entries
[09:26] <Keybuk> 2) sorting of alias expansion
[09:26] <pitti> Evaso2: I don't think so, but which tool? 
[09:27] <mjg59> Keybuk: Does the device ID get passed to the environment when modprobe.d commands are executed?
[09:27] <siretart> Keybuk: ok. let me propose this: I look at your upload now and see what can and should be merged to debians svn here: http://svn.debian.org/wsvn/pkg-wpa/trunk/?rev=0&sc=0
[09:27] <infinity> dmraid.  It's all in the same package.
[09:27] <infinity> (same source package, that is)
[09:27] <siretart> and report back. this package should do it in dapper as well
[09:28] <Keybuk> siretart: there's not much to integrate back
[09:28] <Keybuk> just the init script change
[09:28] <Keybuk> and that's just s/echo/log_action_*_msg/
[09:28] <siretart> ok
[09:29] <Keybuk> why am I only getting 8KB/s from cdimage?
[09:30] <Keybuk> ah
[09:30] <Keybuk> because my boyfriend left bittorrent running
[09:30] <Keybuk> *kills*
[09:30] <Keybuk> oh look, 200KB/s
[09:31] <LaserJock> Keybuk: because everybody loves Dapper?
[09:33] <mjg59> Keybuk: We could just have something in modprobe.d that modprobes the other module if certain criteria are met?
[09:33] <Evaso2> pitti: the dmraid package
[09:33] <siretart> crimsun: care to test my wpasupplicant_0.4.8-1 package? (pkg-wpa/trunk)
[09:34] <Keybuk> mjg59: that'd work too
[09:34] <siretart> crimsun: I've introduced a new variable INTERFACE in /etc/default/wpasuplicant and like to hear a second opinion
[09:34] <Keybuk> mjg59: have "install prism54_softmac" and "install prism54_fullmac" config options that both just load the right one
[09:34] <mjg59> Keybuk: Yeah
[09:35] <Evaso2> mjg59: there is another issue some netgear prism54 card (wg511 made in china) have the wrong fullmac pci id
[09:35] <mjg59> Evaso2: ?
[09:36] <Evaso2> mjg59: the chinese assemblers have missed to set the right id on the pci id memory :)
[09:36] <siretart> btw, are these log_action_{begin,end}_msg commands in init scripts supposed to work in debian as well out of the box?
[09:36] <mjg59> Evaso2: What's it set to?
[09:36] <jdthood> siretart: Yes
[09:37] <crimsun> siretart: I'd love to, but I'm occupied until 2300 GMT
[09:37] <Keybuk> siretart: yes
[09:37] <Evaso2> mjg59: 0000:01:00.0 0280: 1260:3890 (rev 01)
[09:37] <mjg59> Evaso2: And what's the problem with that?
[09:38] <siretart> great. thanks
[09:38] <siretart> crimsun: thats okay. I'll polish further then
[09:38] <Evaso2> mjg59: ubuntu actually try to load the wrong prism54 module associated to this id
[09:38] <mjg59> Evaso2: I don't understand.
[09:38] <Diziet> Damn, it's still stripping firefox-bin.
[09:38] <mjg59> Evaso2: Do you mean there's a softmac card with the same PCI ID as a fullmac one? If so, that's the problem we've just been discussing
[09:40] <Evaso2> mjg59: ah ok, but the problem is it has the wrong PCI id, it is only an hardware assembly problem
[09:40] <sivang> pitti: so, are the db2 debs already testable (thinking of the certification test suite)
[09:40] <sivang> pitti: ?
[09:40] <mjg59> Evaso2: I still don't understand what the problem is here (other than what we've already discussed)
[09:41] <Evaso2> mjg59: on the pragmatical software side nothing 
[09:43] <Pygi> keybuk, poke ...
[09:44] <Keybuk> Pygi: 'sup?
[09:44] <Pygi> Keybuk: I heard that you plan to get N-M "in" by next monday...we should rather postpone that period
[09:44] <Pygi> Keybuk: There is one major issue that we need to solve
[09:44] <Keybuk> what issue is that?
[09:45] <Pygi> Because N-M does constant background-scanning, and madwifi-old doesn't support it (-ng does, but we don't want to support that) disconnects happen
[09:45] <Pygi> this is not good, and should be fixed
[09:45] <Keybuk> that's not a regression from the current situtation though
[09:45] <Pygi> there are also some minor issues, like bad icons :-/
[09:46] <Keybuk> so I don't see how that should block the update
[09:46] <Pygi> Keybuk: yes, true, but we should really solve that first
[09:46] <Lure> Keybuk: I agree
[09:46] <lifeless> mjg59: is the hibernate button still considered known-non-functional ?
[09:46] <Keybuk> I don't see the need for a "first"
[09:46] <Lure> we will just get more testers
[09:46] <Keybuk> uploading 0.6 to main doesn't mean we can't continue to fix bugs afterwards
[09:46] <Pygi> Keybuk: ah, ok, your call after all, but I just don't think it's ready for inclusion
[09:46] <Kamion> better to get 0.6 in early if we're going to get it in at all
[09:46] <Keybuk> we have 10 weeks until release to do that
[09:46] <mjg59> lifeless: No, should work now
[09:47] <Pygi> Kamion: ah, ok
[09:47] <lifeless> mjg59: should I file a bug then ?
[09:47] <Keybuk> "get it in, then worry about whether it's comfortable"
[09:47] <Keybuk> *ahem*
[09:47] <Kamion> postponing it for one piece of hardware makes it more likely that we won't be able to include it after all
[09:47] <mjg59> lifeless: What are you running, and what does it show up in xev as?
[09:47] <Keybuk> anyway, the disconnection issue isn't that bad
[09:47] <Keybuk> you don't notice it unless you're running wavemon or something
[09:47] <lifeless> dell X1 still
[09:48] <lifeless> KeyPress event, serial 26, synthetic NO, window 0x4600001,
[09:48] <lifeless>     root 0x4c, subw 0x0, time 685560661, (627,207), root:(633,305),
[09:48] <lifeless>     state 0x0, keycode 165 (keysym 0x0, NoSymbol), same_screen YES,
[09:48] <Pygi> Keybuk: a lot of people reported that...but ok..
[09:48] <Pygi> Kamion: agreed, but still .. perfection is what counts ^_^
[09:48] <Keybuk> Pygi: sure; but then that situation exists with the current 0.5 packages too
[09:48] <mjg59> lifeless: Running gnome-power-manager?
[09:48] <Keybuk> and is only fixable by shipping madwifi-ng from what I understand it
[09:49] <Keybuk> or stopping n-m scanning on atheros
[09:49] <lifeless> yes, suspend w button works.
[09:49] <Pygi> Keybuk: not really...we can stop scanning if we have connection...
[09:49] <lifeless> erm, : yes, suspend button works.
[09:49] <Pygi> and once it dc/s, start scanning again
[09:49] <mjg59> lifeless: That's not actually strong evidence for that
[09:49] <Keybuk> there are better bugs to try and fix
[09:49] <mjg59> lifeless: Can you check?
[09:49] <lifeless> thats why I added the comma.
[09:49] <Keybuk> like that fact N-M doesn't report state for interfaces it's not controlling
[09:49] <Pygi> Keybuk: ah, ok
[09:49] <lifeless> there is a gnome-power-manager process, AND the suspend button works.
[09:50] <mjg59> lifeless: Ok. hald-addon-keyboard ?
[09:50] <mjg59> (Running?)
[09:50] <Pygi> Keybuk: hm, well, we can't just take over all the interfaces
[09:50] <lifeless> mjg59: no.
[09:50] <Keybuk> Pygi: reporting state != taking over
[09:50] <mjg59> lifeless: That would be your problem, then
[09:50] <Pygi> Keybuk: yes, understandable
[09:50] <lifeless> doesn't seem to be in the path
[09:51] <mjg59> lifeless: No, it needs to be launched by hal
[09:51] <pitti> Mithrandir: at that time I still was under the impression that the only thing that was left was xmms, which we could demote easily
[09:51] <pitti> seb128: alright, maybe we can get rid of it then, I'd much appreciate
[09:51] <Pygi> Keybuk: ok, so we'll fix bugs after inclusion
[09:52] <seb128> pitti: I'll have a look on some of those stuff if I get borred by bug triage :p
[09:52] <seb128> pitti: the list is not that long, so maybe it's easy ... depending of the app
[09:52] <pitti> yes, I didn't check the particular apps
[09:52] <pitti> imlib should probably go with gtk+1.2 anyway
[09:52] <Pygi> Keybuk: as to what I know, we are to ship madwifi-ng for dapper+1 anyway, along with new n-m
[09:52] <lifeless> mjg59: how do I make that happen ?
[09:52] <_ion> siretart: Pong.
[09:52] <_ion> Hi Pygi
[09:53] <mjg59> lifeless: It ought to happen automatically. Sounds like something is broken for you.
[09:53] <Pygi> _ion: hey, what's up?
[09:53] <mjg59> lifeless: Running latest hal?
[09:53] <Keybuk> Pygi: the main problem with that is that madwifi-ng has regressions from madwifi
[09:53] <Keybuk> like not actually working on amd64
[09:53] <lifeless> 0.5.7-1ubuntu7
[09:53] <janimo> pitti, is putting xfce translations in langpack-base still the plan?
[09:53] <Pygi> Keybuk: yup, I know ... :-/ 
[09:53] <mjg59> lifeless: Do you have /usr/share/hal/fdi/policy/10osvendor/10-keyboard-policy.fdi ?
[09:53] <Keybuk> so assuming upstream get their act in gear by dapper+1, yes, otherwise maybe
[09:53] <pitti> Riddell: kdegraphics is the only package needing imlib; any chance to migrate to something else?
[09:53] <lifeless> yes
[09:53] <Keybuk> depends how enthusiastic infinity and/or mjg59 get about shipping both :)
[09:54] <lifeless> 470 bytes
[09:54] <Mithrandir> mgalvin: hiya, I'm planning on releasing flight-6 on Wednesday.  Any chance you could do your magic and whip up a "what's new and shiny in flight 6" page for me? :-)
[09:54] <Kamion> Pygi: the perfect is the enemy of the good
[09:54] <pitti> Riddell: like the much more modern imlib2?
[09:54] <mjg59> Keybuk: We need to ship madwifi-ng for dapper anyway
[09:54] <Pygi> Kamion: heh ;)
[09:54] <Keybuk> mjg59: we do?
[09:54] <ogra> janimo, but please in separate langpack-xubuntu packages or something 
[09:54] <mjg59> Keybuk: Yes
[09:54] <pitti> janimo: yes, it'll happen automatically now that they are in main
[09:54] <Keybuk> mjg59: infinity was getting cold feet about it yesterday
[09:54] <Kamion> (language-pack-xfce-*)
[09:54] <mjg59> Keybuk: Since it's insane for us to ship without support for modern hardware when there's a working driver
[09:54] <Kamion> (I imagine)
[09:54] <pitti> Kamion: oh?
[09:54] <ogra> Kamion, or that indeed
[09:54] <Riddell> pitti: I'm not sure, I'll take a look sometime
[09:54] <_ion> pygi: Hm, have you and tonio looked at the updated package i sent?
[09:54] <Kamion> pitti: or not. where were you planning on putting them?
[09:54] <pitti> Kamion: no, they are small enough for the base ones, as we figured out recently
[09:54] <Pygi> mjg59: but -ng is likely to be in universe or?
[09:55] <mjg59> Pygi: No
[09:55] <Pygi> _ion: hm, yes, I did..
[09:55] <ogra> pitti, edubuntu is currently 9M oversized ... and i only ship english
[09:55] <Kamion> pitti: oh, ok
[09:55] <pitti> Kamion: some 100 kB or so
[09:55] <mjg59> But it won't be used by default except on cards that are only supported by it
[09:55] <janimo> maybe 200K
[09:55] <mgalvin> Mithrandir: as in 20060329?
[09:55] <janimo> I'll recalculate
[09:55] <Mithrandir> mgalvin: yes
[09:55] <Pygi> mjg59: huh :-/ so we are including -ng in main as well?
[09:55] <Kamion> could be 2MB on Ubuntu CDs then
[09:55] <ogra> oh, ok, if its in the kb area ...
[09:55] <mjg59> Pygi: No, restricted
[09:55] <janimo> too bad abiword and other apps are translated part of gnome-langpack so cannot get those separately
[09:55] <Kamion> *probably* doable but be prepared to reconsider that, if you could
[09:55] <mgalvin> Mithrandir: yea sure I can get it done by then
[09:56] <Mithrandir> mgalvin: thanks :-)
[09:56] <Pygi> mjg59: hm, ok... can we then make madwifi suggested package for network-manager?
[09:56] <mjg59> Pygi: No
[09:56] <_ion> pygi: Well, is the change totally evil? :-)
[09:56] <Mithrandir> mgalvin: I assume you'll coordinate with heno about a page on www.u.c instead of the wiki and such.
[09:56] <lifeless> mjg59: any other thoughts ?
[09:56] <mjg59> lifeless: Can you put up a lshal somewhere?
[09:56] <Pygi> _ion: not totally evil ;)
[09:56] <janimo> Kamion, I'm fine with langpack-xubuntu too if not too much hassle for pitti/the archive whatever
[09:56] <lifeless> mjg59: anything in that that I should censor ?
[09:56] <Pygi> mjg59: ah, ok then ^_^
[09:56] <mjg59> lifeless: No
[09:57] <mjg59> Oh, possibly your serial number if you're paranoid
[09:57] <janimo> if it turns out too much overhea
[09:57] <janimo> d
[09:57] <mgalvin> Mithrandir: np :)... yea, i have access to it, i will move it there when its ready
[09:57] <pitti> Kamion: I'd be fine with either way, it's not difficult to build an extra set
[09:57] <Pygi> _ion: come to #kubuntu-devel, we have to discuss some more things
[09:57] <Kamion> janimo: I'm not overly keen on another seven million packages, but if it's what it takes - we'll see
[09:57] <Mithrandir> mgalvin: thanks again.
[09:58] <lifeless> mjg59: so grep -v serial is ok ?
[09:58] <mjg59> lifeless: Best to eyeball it
[09:58] <mgalvin> Mithrandir: np, I will let you know when its all set :)
[09:59] <pitti> Kamion: me neither, and most of the translations are around 50 kB or so (just the most popular ones were 200 kB), so my gut feeling is that it's not really worth the package overhead
[10:00] <janimo> pitti, anything not explicitely put in gnome or kde langpack and is in main goes to base-langpack?
[10:00] <pitti> janimo: yes
[10:01] <pitti> well, right now, that is; easy enough to tweak
[10:01] <pitti> but it wouldn't fit into either gnome or kde
[10:01] <janimo> if so there are apps which I did not count last time because not xfce proper (graveman, evince-gtk etc)
[10:03] <lifeless> mjg59: people.ubuntu.com/~robertc/lshal.txt
[10:04] <pitti> janimo: evince-gtk doesn't reuse evince.mo? that's evil
[10:04] <janimo> pitti, well I can make it so
[10:04] <janimo> it will reuse it :)
[10:04] <lifeless> mjg59: we're at the restraunt, I will catch up tomorrow and see if there was useful data in there as I imagine ints not a trivial thing to check at this point
[10:04] <pitti> janimo: please do, otherwise all translations need to be done twice
[10:05] <pitti> janimo: so, the pots of both should just be merged (if there are any different strings at all)
[10:05] <janimo> so we'll probably have gnoem langapcks too on xubuntu CD for abiword/evince-gtk whatebver else we may have in common
[10:05] <lifeless> mjg59: ok ?
[10:05] <mjg59> lifeless: Ok
[10:05] <janimo> pitti, ok
[10:05] <mjg59> lifeless: From that, hald-addon-keyboard should have been started. Is evdev loaded?
[10:06] <Pygi> Keybuk: also, new patched L-R-M packages seem to fail for Madwifi...
[10:06] <fabbione> night guys
[10:06] <Pygi> night fabbione
[10:06] <Keybuk> "fail" ?
[10:06] <Keybuk> works ok here
[10:06] <pitti> night fabbione 
[10:06] <Pygi> Keybuk: not work that is...
[10:06] <ogra> ciao fabbione 
[10:06] <janimo> night all
[10:06] <pitti> oh, my main network seems to be back, brb
[10:06] <Keybuk> "fail" and "doesn't work" are both banned terms here
[10:07] <Pygi> Keybuk: hm...perhaps...but it still doesn't work...old L-R-M we have in our repo seem to work with older kernel build, but new ones that are in official repo don't work with new kernel build
[10:09] <Pygi> altought the ones that are in official repo are patched with the same patch we used to patch the older ones...
[10:09] <Keybuk> define "don't work"
[10:09] <Keybuk> I'm running today's kernel and lrm, and I'm on the network
[10:09] <Keybuk> "my network card doesn't work" can be anything from module won't compile, won't load, isn't loaded automatically, won't ifconfig, ifconfigs but won't dhcp, no traffic sent/received, etc.
[10:09] <Pygi> the issue can be seen with WPA
[10:10] <Pygi> the N-M reports that the card cannot do the WPA, while it can
[10:10] <Lure> Keybuk: I suspect madwifi WPA
[10:10] <Pygi> problem is seen with athereos cards aka Madwifi
[10:11] <mjg59> Evaso2: What do the firmware files need to be called?
[10:16] <pitti> Keybuk: wpasupplicant code looks reasonable, I didn't find anything obvious (but I'm too tired already to really dig deep)
[10:17] <pitti> Keybuk: I'm not sure whether it's a piece of sw that we can support for three years, though
[10:17] <pitti> but if you think it's sane, it'll have my blessing
[10:17] <Keybuk> I think it's sane enough
[10:17] <Keybuk> it's as sane as WPA :p
[10:18] <Pygi> Keybuk: thoughts on not working L-R-M?
[10:18] <Keybuk> Pygi: not sure, just setting up my test WPA environment again
[10:18] <pitti> Keybuk: don't scare me off :)
[10:18] <Pygi> Keybuk: ok, thanks ^_^
[10:19] <Keybuk> uh
[10:19] <Keybuk> why isn't write-to-cd working today?
[10:20] <Keybuk> "because Keybuk put the CD in upside down"
[10:21] <Amaranth> Nice.
[10:21] <Kinnison> Keybuk: tit
[10:21] <Amaranth> I can honestly say I've never done that one. :P
[10:21] <Keybuk> though there's a bug here
[10:21] <Keybuk> I put the CD in my second writer
[10:22] <Keybuk> and it popped up, I clicked Write Data CD, copied the ISO, it asked me whether I wanted to write from image, I did
[10:22] <Keybuk> and the dialog box had my *first* writer selected
[10:22] <Amaranth> it should poke hal to find out which one has a CD in?
[10:25] <Keybuk> it should somehow pass from the "g-v-m sponsored dialog" which writer is being used
[10:27] <Amaranth> wow, dpkg must be killing my HD, i've never had my fan spin up to full speed before
[10:27] <Amaranth> sounds like a small blow dryer
[10:49] <joelbryan> how large is the GNOME Desktop? without any applications just nautilus and metacity?
[10:49] <tseng> how large in what way
[10:50] <_ion> The diameter, in meters.
[10:50] <tseng> yeah.
[10:50] <tseng> its 6 hands high
[10:51] <tseng> joelbryan: a full install of an ubuntu desktop (gnome, firefox, OOo mainly) is a bit under 2gb
[10:51] <torkel> oh, thought it was one foot :-)
[10:51] <milli> darn it all...  ATI chipset in new ThinkPad T60p is unrecognized by radeon/ati driver in xorg...  PCI dev id 1002:71c4
[10:52] <joelbryan> tseng: no apps, just gnome-panel, metacity, nautilus, and human icons
[10:52] <tseng> milli: file a bug, it may be fixable
[10:53] <tseng> joelbryan: i still dont know what indicator of size you are talking about
[10:53] <Amaranth> MB?
[10:53] <tseng> installed on disk?
[10:53] <tseng> in ram?
[10:53] <tseng> on the install cd?
[10:53] <Burgwork> joelbryan, larger than a breadbox :)
[10:53] <joelbryan> flash disk
[10:54] <tseng> i am sticking to "just under two gigs"
[10:54] <tseng> if you want an exact figure you'll need to do your own research
[10:54] <Burgwork> 1.8 was what the hoary and breezy cd cases said
[10:54] <Burgwork> GB that is
[10:54] <Amaranth> yeah
[10:55] <Amaranth> you'll need at least 1.5GB, i think
[10:55] <joelbryan> if I compiled it with --enable-static
[10:55] <Amaranth> um
[10:55] <tseng> static is bigger
[10:55] <Amaranth> that would be a waste
[10:55] <joelbryan> how do you make it under 1 cd?
[10:55] <Amaranth> compression
[10:55] <Burgwork> voodoo
[10:56] <Amaranth> gz and bz2 for the packages on the install cd, squashfs for the live cd
[10:56] <joelbryan> ok hehehe
[10:56] <Amaranth> squashfs-lzma, that is
[10:56] <tseng> you might be able to find a way to copy the livecd image onto a flash disk
[10:56] <tseng> and make it boot
[10:56] <tseng> or similar tactic
[10:56] <joelbryan> tseng: yeah, that was I'm trying to do
[10:56] <Amaranth> i don't think that'd be hard, do flash drives have mbrs?
[10:56] <tseng> Mithrandir might be able to give you a tip or two
[10:57] <tseng> point you in a direction to hack on it
[10:57] <tseng> Amaranth: i dont think so, but modern BIOS has some kind of method of booting off it
[10:57] <joelbryan> I'm trying to make a super small dapper
[10:58] <tseng> there is a Gentoo/Gnome live-usb image
[10:58] <tseng> that might give you some good ideas
[10:58] <milli> tseng: ok
[10:59] <joelbryan> is the live cd uses bz2 compressions too?
[10:59] <tseng> some debs are bz2'd maybe
[10:59] <tseng> the cd uses squashfs
[11:00] <joelbryan> so it's on the fly uncompress when a program is called.
[11:00] <tseng> actually, the livecd shouldnt have deb's on it
[11:00] <tseng> i think part of it might be uncompressed into ram
[11:01] <tseng> you are talking to the wrong people for specific details
[11:01] <joelbryan> ok,
[11:02] <joelbryan> I just want to figure out how to fit dapper under 50MB
[11:02] <tseng> eh
[11:03] <tseng> I am pretty confident that isnt going to happen with GNOME
[11:03] <joelbryan> maybe strip all apps, and just the core, and create a script that has on-the-fly apt-get install clicked-app
[11:03] <tseng> i am telling you, you can only strip so much
[11:04] <joelbryan> so gnome can't compile without those apps?
[11:04] <tseng> the livecd is around 600mb compressed
[11:05] <tseng> if you remove firefox and open office
[11:05] <tseng> you arent getting anywhere close to 50mb
[11:05] <tseng> and one could question what you want with a desktop without a browser or a word processor
[11:06] <joelbryan> they can make it 50MB, if they use the installer cd for on-the-fly "apt-get install clicked-apps"
[11:06] <tseng> who is they
[11:06] <joelbryan> the installer cd would be the main repos
[11:06] <tseng> you cant get a base system with gnome on top in 50mb
[11:07] <joelbryan> how about xfce?
[11:08] <Pygi> just use blackbox ;)
[11:08] <tseng> i think you are really overestimating what fits into 50mb of space
[11:08] <tseng> and need to give it a try to prove it to yourself, I guess
[11:09] <tseng> Installed-Size: 60696
[11:09] <tseng> the linux-image-2.6.15-18-686 is 50mb
[11:09] <tseng> 60*
[11:09] <joelbryan> I think it's possible to make the live cd smaller than 600MB, using the installer cd as the main repos
[11:10] <tseng> thats installed and uncompressed
[11:10] <tseng> but you have to have some manner of kernel and initrd outside the compressed loopback
[11:10] <tseng> "using the installer cd as the main repos" you will have to rephrase
[11:10] <tseng> i have one cd rom drive, I do not understand your idea
[11:11] <joelbryan> tseng: on-the-fly apt-get install clicked-app
[11:11] <tseng> that makes the cd totally useless for people without a network
[11:11] <tseng> or 2 cd drives
[11:11] <joelbryan> in background and just a prompt dialog that says "Please Insert Ubuntu Disk 1"
[11:12] <tseng> you seem to be forgetting that you system is running of the cd in the tray
[11:12] <joelbryan> i have ejected a live cd while running it.
[11:13] <joelbryan> probably because all is loaded in memory
[11:13] <tseng> its past 5pm here, time to head home
[11:13] <tseng> good luck with your project.
[11:13] <joelbryan> thanks, it's 5am here
[11:19] <joelbryan> do you think it's necessary to copy everything to disk? why not apt-get install when needed, that way the installation time for ubuntu is much shorter.
[11:21] <joelbryan> there should be a "Just Desktop" install method, aside from OEM install
[11:28] <Pygi> hm, is it only me, or does it seem we have an outdated madwifi support?
[11:28] <Pygi> me and Tonio_ were just looking, and we are lacking large portions of code
[11:30] <Keybuk> it hasn't been updated in a while, no
[11:30] <Pygi> Keybuk: this is not good :-/
[11:31] <Pygi> Can we do a diff, and apply it?
[11:31] <Keybuk> Pygi: where did you diff from?
[11:32] <Pygi> Nowhere yet ... but we were just looking at one patch, and noticed that we are missing large portions of code :-/
[11:33] <Keybuk> bear in mind that upstream "trunk" is -ng
[11:33] <Pygi> yup, understandable
[11:35] <Tonio_> hi
[11:37] <Keybuk> it's certainly been updated within Dapper
[11:37] <Keybuk> in fact, last update was 1 Dec
[11:37] <Keybuk> so it's not that out of date
[11:37] <joelbryan> deskbar-applet is now included on default
[11:37] <Keybuk> in fact:
[11:37] <Keybuk> r1417 | kelmo | 2006-01-27 13:42:11 +0000 (Fri, 27 Jan 2006) | 5 lines
[11:37] <Keybuk> Fix a compile-time type-warning message in madwifi/ath/if_ath_pci.c on linux
[11:37] <Keybuk> >=2.6.15. Backported from madwifi-ng.
[11:38] <Keybuk> is the only commit to madwifi-old since then
[11:38] <Keybuk> which hardly seems serious
[11:38] <Pygi> Keybuk: ok, thanks
[11:39] <Pygi> we'll have a lot of problems generated by madwifi :-//
[11:44] <Seveas> Keybuk, madwifi devs are crazy, development has stalled on what they call stable and the unstable branch is still quite crackful...
[11:46] <Amaranth> mjg59: you have an intel mini, right? how is the graphics support on that?
[11:48] <mjg59> Amaranth: Unaccelerated linear framebuffer
[11:49] <Amaranth> ouch
[11:49] <Amaranth> i thought drivers already existed for it
[11:49] <Amaranth> or is it an EFI issue?
[11:49] <mjg59> No, the existing i945 drivers require a video BIOS
[11:49] <Amaranth> ick
[11:51] <siretart> Keybuk: I see you assigned the wpasupplicant UVF exception request to yourself
[11:51] <Keybuk> siretart: why not? :)
[11:51] <siretart> Keybuk: I think what wie currently have in debians svn/trunk is suitable for both debian and ubuntu
[11:51] <siretart> but I don't have a wpa secured network here to test
[11:52] <Keybuk> I've gone with experimental :)
[11:52] <Keybuk> I like it
[11:52] <siretart> Keybuk: the package in trunk does install an init script, but doesn't start wpasupplicant by default
[11:52] <siretart> should be fine for integration with nm
[11:53] <siretart> Keybuk: ah, so you merge the ifupdown integration to 0.4.8-1?
[11:53] <Keybuk> (on phone, hang on)
[11:53] <siretart> ok
[11:53] <Keybuk> but yup, basically
[11:54] <Burgwork> anybody else getting duplicated emails for dapper-changes?
[11:54] <seb128> no
[11:54] <seb128> but you might be subscribed to the launchpad package and get the mail from launchpad?
[11:54] <siretart> Keybuk: if you can test this and it works for you, that'll speed up my confidence that it do okay in debian/unstable :)
[11:55] <Seveas> siretart, I am using your 0.4.8 wpa_supplicant here and it works like a charm
[11:55] <siretart> Seveas: thanks :)
[11:55] <Seveas> if there's something newer you want me to test, just poke 
[11:55] <siretart> Seveas: how do you use wpasupplicant? via nm only?
[11:55] <siretart> Seveas: or via init script?
[11:56] <Seveas> currently initscript as there is something broken between madwifi and NM, but it worked a few days ago with NM and madwifi
[11:57] <Seveas> so: no complaints, excellent job with the supplicant
[11:57] <siretart> Seveas: I've cleaned up the initscript massively, this needs testing
[11:57] <Seveas> aight, did you deb-ify it already?
[11:57] <siretart> well, not that massive, but quite a bit. 
[11:57] <siretart> Seveas: it on my latop right now. I'll upload that
[11:57] <Seveas> ok, just poke me when ready
[11:58] <siretart> Seveas: ok, package in place
[11:58] <siretart> not that I didn't change the version number, it is still wpasupplicant_0.4.8-1_i386.deb
[11:59] <Seveas> where can I download it?
[11:59] <siretart> http://siretart.tauware.de/wpasupplicant/wpasupplicant_0.4.8-1_i386.deb
[12:00] <Seveas> ok, my connection may drop during wpa_supplicant restart - so brb
[12:00] <siretart> the directory holds the sources, they come out from here http://svn.debian.org/wsvn/pkg-wpa/trunk/?rev=0&sc=0
[12:00] <Lure> siretart: I am using your 0.4.8 and it works great (WPA2-PSK)
[12:01] <siretart> Lure: via nm or via init script?
[12:01] <Lure> nm 0.6
[12:01] <siretart> ok
[12:01] <Seveas> siretart, "daemon" is run from /etc/rc*.d and "link" from if-pre-up.d?
[12:02] <Seveas> (dpkg prompted for a config file change)