2008/02/07/#ubuntu-devel.txt

[00:31] <rtc> Did someone already fix the issue noted by Knuth in his "Flame About 64-bit Pointers" (http://www-cs-faculty.stanford.edu/~knuth/news.html)?
[00:34] <rtc> Ah, found it in the bugtracker...
=== Ubulette_ is now known as Ubulette
[02:10] <Vadi> Quick question, but how does dependency resolution work when I remove a package? I ran into this problem - installed the emacs package, it for 67mb of other stuff. Removed it, and it only cleared 40kb of stuff.
[02:11] <Vadi> I had to manually go into synaptic and remove the things it dragged in. Not good, because there's a ton of other programs that can probably leave packages about and I won't know about them...
[02:11] <mahmoud_> !offtopic | Vadi
[02:11] <ubotu> Vadi: #ubuntu is the Ubuntu support channel, #ubuntu+1 supports the development version of Ubuntu and #ubuntu-offtopic is for random chatter. Welcome!
[02:12] <Vadi> Uh huh. Exactly which do I go to?
[02:12] <mahmoud_> join #ubuntu
[02:12] <Vadi> That's a support channel, I was already there.
[02:12] <Vadi> I need a technical explanation, please.
[02:13] <ScottK2> That doesn't make this a support channel, but in a console window do sudo apt-get autoremove and that'll get rid of it (assuming you aren't on Dapper).
[02:13] <Vadi> I did it, but that didn't work.
[02:13] <Vadi> It said everything was okay.
[02:14] <Vadi> But, oh well, whatev. I'll file a bug report
[02:16]  * Hobbsee waves
[02:19] <RAOF> Howdie, Hobbsee!
=== Traxer is now known as Traxer|off
=== Traxer|off is now known as Traxer
[04:11] <Aloha> whats a sprint?
[04:12] <lifeless> its faster than a walk
[04:12] <lifeless> sorry ;)
[04:12] <Aloha> heh
[04:13] <lifeless> in general
[04:13] <lifeless> http://en.wikipedia.org/wiki/Sprint
[04:13] <lifeless> but in Ubuntu
[04:13] <lifeless> http://en.wikipedia.org/wiki/Hackathon#Sprints
[04:13] <Aloha> lifeless, thanks :)
[04:16] <Aloha> lifeless, what gathering are ubuntu sprints centered on?
[04:18] <Aloha> oh ijust realized soyuz is russian for launchpad
[04:22] <AfterDeath> Aloha: in the general sense, a sprint is using a burst of energy to run as fast as possible (unlike normal running, where you would run at a pace that would allow you to keep running for a long time)
[04:22] <Aloha> AfterDeath, i know. i was wondering about software sprints. i used to run track
[04:22] <AfterDeath> ah
[04:29] <jdong> Aloha: sprint's an overpriced poor coverage north american cellphone provider :)
[04:29] <Aloha> jdong, i use AT&T ;)
[04:29] <superm1> jdong, well the nextel side of things is overpriced at least, but my coverage is great
[04:30] <jdong> Aloha: I use AT&T too... it's great at my home in MI
[04:30] <jdong> Aloha: but here on the MIT campus, I can freaking use Skype to make calls from more places than I can use my cell
[04:32] <superm1> jdong, you need to hack together some sort of mic for the touch and then port over an SIP app
[04:32] <superm1> thats a much better project than what you mentioned in -motu the other day :)
[04:33] <jdong> superm1: well first I need to fix some.. er.. packaging issues.. with the MIT intro to EECS laptops :)
[04:33] <jdong> superm1: my predecessor went kinda metapackage-happy....
[04:33] <superm1> haha
[04:33] <jdong> and to make matters worse, several metapackages contain VITAL INIT SCRIPTS
[04:34] <jdong> :D
[04:34] <superm1> running what distro?
[04:34] <jdong> superm1: Feisty :)
[04:34] <StevenK> !jdong
[04:34] <ubotu> <Hobbsee> jdong: yes, but you're FULL OF CRACK!
[04:34] <jdong> Ubuntu is well loved around here
[04:34] <superm1> jdong, just ditch feisty and port it all to hardy?
[04:34] <jdong> the next version of the campus public workstation OS will be based on Ubuntu
[04:34] <jdong> superm1: wish I can, but we use a proprietary kernel module atm
[04:35] <StevenK> For what?
[04:35] <jdong> superm1: fortunately, we have one bright student working on reverse engineering that
[04:35] <jdong> StevenK: NIDAQ data acquisition box
[04:35] <superm1> jdong, the module isn't available on newer kernels?
[04:35] <jdong> the thing hardlocks and/or panics the kernel when it isn't happy
[04:35] <StevenK> Nice
[04:35] <superm1> jdong, to build against newer kernels that is
[04:35] <jdong> superm1: from what I understand, it can give RMS heart attacks from 3 floors down
[04:35] <jdong> :)
[04:35] <Aloha> jdong, that sucks. i don't have to worry about coverage too much. I live on an island
[04:35] <superm1> haha
[04:36] <jdong> superm1: but yeah.... it is full of problems the least of which is we have to force-dep on the original released feisty kernel
[04:36] <jdong> (YES... it's a BINARY kernel module....)
[04:36] <superm1> oh that's really unfortunate
[04:36] <jdong> there's no such thing as "compiling" 'round here :D
[04:36] <superm1> you guys tried to contact NIDAQ?
[04:36] <jdong> once this rev enged one kicks off, I'm gonna be happy
[04:37] <jdong> superm1: I'm not sure, this is the first day I've been submersed into the inner world of the 6.01 course
[04:37] <superm1> ah
[04:37] <jdong> superm1: last term I served as a lab assistant primarily answering student questions on lab assignments and python code
[04:37] <jdong> this term.... I'm kinda managing all the course laptops and stuff too
[04:38] <superm1> jdong, but for the machiens that don't end up needing support for the data acquisition box, they could go to a newer release could they not?
[04:38] <jdong> superm1: well they all will by the 4th week of labs
[04:38] <jdong> superm1: as I said, upgrading Ubuntu is one of the smaller concerns
[04:38] <jdong> right now, there's 4 padlocked RHEL lab workstations that are hardlocked
[04:38] <jdong> the datestamp on one of them is 2000
[04:39] <jdong> apparently it's padlocked so well nobody knows how to get in :D
[04:39] <jdong> overall though it's a finely run class. Just shows how fun technology can be at times :D
=== asac_ is now known as asac
[06:49] <pitti> Good morning
[06:49] <ion_> Hi
[06:54] <Hobbsee> morning pitti!
[06:54] <pitti> hey Hobbsee, hi ion_1
[07:29] <slytherin> Hi all, we have latest kernel in repos linux-image-2.6.24-7-generic but linux-image-generic still depends on linux-image-2.6.24-5-generic. Is this a known issue? The reason I am asking is that the latest kernel fixes ibook boot issue on hardy but it won't be installed in standard upgrade.
[07:30] <Hobbsee> slytherin: normal for a while, until each image has been out for a bit
[07:30] <Hobbsee> slytherin: has linux-image-* built on all arches yet?
[07:31] <Hobbsee> slytherin: and has l-r-m and l-u-m built on all arches yet?
[07:31] <slytherin> Hobbsee: let me check
[07:31] <mjg59> slytherin: linux-ubuntu-modules isn't built yet, I believe
[07:32] <pitti> no, lum and lrm weren't uploaded yet
[07:32] <slytherin> Hobbsee: hppa build for linux-image has failed, ia64 is currently building.
[07:32] <Hobbsee> slytherin: the above 2 conditions are why, then
[07:32] <slytherin> Thanks for info
[07:33] <slytherin> I am also eager to check if the new driver manager handles broadcom 43xx devices because there is now a new driver.
[07:46] <pitti> slytherin: not yet, but I think I'll manage to add one for b43 in the next week
[07:46] <pitti> slytherin: can I ping you once it's ready, for testing?
[07:47] <pitti> slytherin: would be great if someone could test it before I upload it
[07:48] <TheMuso> pitti: Add me to that list also. I've a mini here iwth a b43 I think it is, so I'd be happy to help with testing also.
[07:48] <pitti> cool
[07:49] <TheMuso> Yeah, a BCM4306
[07:50] <pitti> TheMuso, slytherin: I don't know anything about b43 and b43-fwcutter; from a clean install, is it enough to install b43-fwcutter and it does the rest? if not, which steps are necessary in addition
[07:50] <TheMuso> pitti: All I need do after an install, is install fwcutter, and I think, it works from there.
[07:51] <pitti> ok, great
[07:51] <pitti> that should be easy enough, then I just need the detection
[07:51] <pitti> TheMuso: do you have the current jockey installed?
[07:52] <TheMuso> pitti: I plan on doing a fresh install in the next few days, so will try the procedure again. If its any different, I'll ping you.
[07:52] <pitti> TheMuso: can you please do "jockey-gtk --debug --list >/tmp/jockey.log 2>&1"
[07:52] <TheMuso> pitti: Just updating the box now..
[07:52] <pitti> TheMuso: and send me /tmp/jockey.log? that should help me to verify the current detection
[07:53] <TheMuso> pitti: Will do.
[07:53] <pitti> awesome, thanks
[07:53] <TheMuso> As soon as the mini update.
[07:53] <TheMuso> updates
[07:54]  * TheMuso gets dinner while waiting...
[08:08] <slytherin> pitti: I will be happy to test. It is jut that I am not sure if the firmware needed with new driver is same as the old driver. Also the old driver is now blacklisted in favour of new one. I will do some manual installation over weekend needed to make the wireless work with latest kernel and then let you know.
[08:14] <pitti> slytherin: great, thanks; maybe you can do the same --debug --list command that I asked above?
[08:14] <pitti> slytherin: the handler will just detect if you have a broadcom wifi, and if so, install b43-fwcutter
[08:14] <pitti> slytherin: if you need anything in addition, please tell me, so that the jockey handler can do it automatically
[08:15] <slytherin> pitti: will do but I will have to wait for the kernel upgrade, right? because current kernel does not boot.
[08:15] <pitti> oh :)
[08:16]  * Hobbsee sighs at her connectino
[08:16] <dholbach> good morning
[08:17] <slytherin> pitti: this is specific to ibook only. related to some old ide drivers. By the way, bcm43xx-fwcutter is in universe. Shouldn't it really be in main and included in CD? because how are you supposed to install it if your connection is not working.
[08:18] <Hobbsee> morning dholbach
[08:18] <dholbach> hey Hobbsee
[08:21] <pitti> slytherin: if it's not working, then it's useless anyway, because it needs to fetch the firmware from the net
[08:22] <slytherin> pitti: You can have firmware saved locally. At least 'Restricted Manager' allowed using a local .so file.
[08:23] <pitti> MacSlow: introducing ubuntu specific strings in patches is ok, as long as the package updates the POT during build
[08:23] <pitti> MacSlow: if the package uses cdbs and gnome.mk, this will happen automatically
[08:24] <pitti> slytherin: hm, true
[08:24] <MacSlow> pitti, the packages in question are libwnck and gnome-panel
[08:24] <pitti> MacSlow: if not, you have to call intltool-update -p
[08:24] <MacSlow> pitti, that one I know
[08:24] <pitti> MacSlow: ah, both use cdbs, and gnome.mk I suppose
[08:25] <pitti> MacSlow: so, just patch the strings, and the POT and LP translations will automatically get them
[08:25] <pitti> so that people can start translating them in LP
[08:26] <MacSlow> pitti, for testing I've provided my own translation for de
[08:26] <pitti> MacSlow: you can add your new translations as a patch, but IMHO it's easier to maintain if you just translate them in LP
[08:26] <iiPing> lo
[08:27] <stgraber> moin
[08:27] <pitti> hi stgraber
[08:27] <MacSlow> pitti, well since I cannot add anything to LP myself (in an official packages) I've to wait for seb128/mvo/you to sponsor/integrate my stuff
[08:28] <iiPing> people i bump a lot, where can i find the ubuntu-app-developer for ubuntu channel?
[08:28] <MacSlow> greetings carlos
[08:28] <carlos> MacSlow: hi
[08:29] <pitti> MacSlow: not? you should become a member of the German translation team then
[08:29] <pitti> hey carlos
[08:33] <TheMuso> slytherin: Actually, I haven't been able to boot recent 2.6.24 kernels on my mini either.
[08:33] <TheMuso> As the mini has notebook components, I'd say they both use the same IDE chip.
[08:36] <pitti> TheMuso, slytherin: FYI, I created the first version of the b43 handler, which should work at least in general; I'll do the testing/fine-tuning together with you later
[08:38] <TheMuso> pitti: Sure.
[08:49] <TheMuso> pitti: You should have mail from me sometime soon.
[08:52] <pitti> \o/
[08:52] <pitti> brb, testing new udev
[08:56] <TheMuso> pitti: Actually, discard that. Seems the bcm43xx module wasn't loaded for some reason.
[08:57] <slytherin> TheMuso: it is blacklisted in favour of new driver. check /etc/modprobe.d/blacklist
[08:57] <slytherin> TheMuso: and the not booting problem is fixed in latest kernel uploads.
[08:57] <TheMuso> slytherin: Yeah I'm using 7.11.
[08:57] <TheMuso> slytherin: Is the newer driver in lum?
[08:59] <slytherin> TheMuso: the driver is in kernel itself. But I am not sure which firmware it needs. Check dmesg, it should have some message related to b43 firmware
[08:59] <mjg59> b43 needs revision 4 firmware. b43legacy needs revision 3.
[08:59] <TheMuso> slytherin: Whats the module name
[08:59] <mjg59> bcm43xx uses revision 3
[08:59] <MacSlow> seb128, greetings
[09:00] <seb128> hi MacSlow
[09:00] <slytherin> and I guess b43 will need b43-fwcutter instead of old bcm43xx-fwcutter
[09:00] <mjg59> The issue isn't the application
[09:00] <mjg59> bcm43xx-fwcutter can extract version 4 firmware
[09:01] <mjg59> But you need different firmware for b43 and b43legacy, and we need to support both
[09:01] <mjg59> b43 can't drive everything bcm43xx can
[09:01] <TheMuso> ah I see
[09:02] <mjg59> It's made even more difficult by the latest hardware needing even newer firmware
[09:03] <slytherin> damn, I thought bcm43xx chipsets are not used anymore.
[09:04] <slytherin> TheMuso: by the way, the module names are b43 and b43legacy. But I am not sure which one handles which chipset
[09:05] <TheMuso> slytherin: Yeah I worked it out.
[09:06] <TheMuso> pitti: Oh ok, the one I sent you is the right one.
[09:07] <mjg59> b43legacy handles old cores. b43 handles new cores. You can't tell from the PCI ID alone.
[09:09] <TheMuso> Well on my mini, b43 gets loaded, and there is no firmware change needed.
[09:10] <Iulian> Good morning.
[09:12] <slytherin> mjg59: then how can one decide which to use if not from PCI ID? Or do we trust the autodetection?
[09:14] <mjg59> Does it use the ssb bus implementation yet?
[09:18] <MacSlow> *sigh*
[09:21] <StevenK> MacSlow!
[09:21] <MacSlow> can someone tell me what's wrong with the way I created http://people.ubuntu.com/~mmueller/02_expose_wm_keybindings.patch, which not apply to libwnck-2.21.90 when doing dpkg-buildpackage -rfakeroot ?
[09:21] <StevenK> MacSlow: Can you pastebin the error you get?
[09:22] <StevenK> MacSlow: It also doesn't patch configure
[09:23] <MacSlow> isn't configure always generated when one starts a package-build?
[09:23] <mjg59> No
[09:23] <StevenK> MacSlow: Not always, depends on the package
=== cjwatson_ is now known as cjwatson
[09:30] <cjwatson> pitti: http://paste.ubuntu.com/4282/
[09:30] <cjwatson> pitti: output of jockey-gtk --debug --list for me; no mention of b43, weirdly
[09:31] <cjwatson> MacSlow: configure is in fact usually *not* detected to avoid build-depending on autoconf et al
[09:31] <pitti> cjwatson: right, no ssb: modaliases, hmm
[09:31] <cjwatson> which sometimes causes problems - generally regarded as better to have the maintainer generate that locally and verify that it's correct
[09:31] <cjwatson> MacSlow: s/detected/generated/
[09:31] <MacSlow> StevenK, pitti, cjwatson: http://paste.ubuntu-nl.org/55063
[09:32] <pitti> cjwatson: does the b43 module get autoloaded for you? or did you stuff it to /etc/modules?
[09:32] <MacSlow> StevenK, pitti, cjwatson: http://paste.ubuntu-nl.org/55063
[09:32] <cjwatson> pitti: gets autoloaded, along with bcm43xx
[09:32] <MacSlow> I created the patch for libwnck like this
[09:32] <asac> calc: any info on ooo + xul yet?
[09:32] <cjwatson> pitti: I had to blacklist bcm43xx to avoid problems
[09:32] <pitti> cjwatson: right, bcm43xx is blacklisted by default now
[09:33] <asac> cjwatson: i think latest update blacklisted it
[09:33] <asac> ack
[09:33] <TheMuso> cjwatson: My log didn't have b43 either
[09:33] <cjwatson> MacSlow: there should be a log somewhere with the output of patch
[09:33] <pitti> hm, so I wonder how b43 can autoload if /sys does not have the corresponding modalias
[09:33] <cjwatson> MacSlow: it may be that you didn't generate it against the immediate parent patch, and collided
[09:34] <pitti> MacSlow: unpack the source to a fresh and clean directory, cdbs-edit-patch 02_expose_wm_keybindings.patch, resolve the conflicts
[09:35] <TheMuso> pitti, cjwatson, my ifconfig -a also gives this: http://www.pastebin.ca/894827
[09:35] <TheMuso> This is with b43 autoloaded.
[09:36] <mjg59> b43 doesn't use PCI modaliases
[09:36] <mjg59> It uses ssb ones
[09:36] <MacSlow> cjwatson, I guess you mean these http://people.ubuntu.com/~mmueller/02_expose_wm_keybindings.patch.level-0.log http://people.ubuntu.com/~mmueller/02_expose_wm_keybindings.patch.level-1.log http://people.ubuntu.com/~mmueller/02_expose_wm_keybindings.patch.level-2.log
[09:36] <mjg59> ssb has the PCI modalias
[09:36] <mjg59> Broadcom uses a bus called ssb - depending on what sort of host bus you have, the ssb bus will be bridged onto that
[09:36] <StevenK> MacSlow: Yeah, that's them.
[09:37] <StevenK> MacSlow: You probably have them locally now, too. :-)
[09:37] <cjwatson> pitti: /sys/bus/ssb/devices/ssb0:0 is my Broadcom card, though
[09:37] <mjg59> The PCI device is just a PCI to ssb bridge
[09:37] <MacSlow> StevenK, of course
[09:37] <mjg59> So ssb gets autoloaded, and then generates modalias events for the Broadcom cores
[09:37] <MacSlow> StevenK, still I don't understand why that fails
[09:38] <mjg59> pitti: Does that explain things?
[09:38] <MacSlow> I did use cdbs and created the paches as always
[09:38] <StevenK> mjg59: What does ssb expand out to? Wikipedia failed me.
[09:38] <mjg59> Silicon Sonics Backplane
[09:38] <StevenK> MacSlow: Because the patch doesn't apply for some reason.
[09:38] <TheMuso> pitti: Mine is /sys/bus/ssb/devices/ssb0:0
[09:38] <StevenK> MacSlow: Either too much fuzz, or failed hunks
[09:39] <cjwatson> mjg59: ssb doesn't seem to expose modalias files in /sys
[09:39] <cjwatson> which is what jockey is looking for
[09:39] <mjg59> cjwatson: That's potentially the case
[09:39] <mjg59> I'm in San Francisco right now, so I don't have access to a Broadcom right now
[09:39] <cjwatson> MacSlow: level-1 is the relevant one; three failed hunks for some reason
[09:39] <pitti> TheMuso: wlan0_rename> udev bug, I filed it a while back
[09:40] <MacSlow> it drives me nuts when additional levels of "infrastructure" suddenly make my created patches not apply anymore
[09:40] <StevenK> MacSlow: But it just runs patch ...
[09:40] <pitti> mjg59: yes, thanks; I just need to find a way how to grab it out of /sys then
[09:41] <TheMuso> pitti: ah ok. My mini is on ethernet, and I never use wireless on it, so thats alright for now.
[09:42] <pitti> TheMuso: can you tar up your /sys and send it to me?
[09:42] <TheMuso> pitti: Can do.
[09:42] <pitti> TheMuso: if I cannot rely on moaliases, I need to find a b43 specific way of autodetection
[09:42] <pitti> TheMuso: thanks muchly
[09:42] <pitti> TheMuso: I'll do "moalias present" || "b43 ssb custom detection" then
[09:42] <mjg59> pitti: The b43 and b43legacy modules do, at least, have the revisions they each support in them
[09:43] <MacSlow> e.g. just running: patch -p1 <../02_expose_wm_keybindings.patch (while in a freshly grabbed libwnck-2.21.90 works flawless)
[09:43] <StevenK> MacSlow: Ah ha. Then the first patch affects it
[09:43] <pitti> mjg59: right; I hope I can pick out the vendor/product/revision IDs somewhere from /sys and compare it against the modalias lines from modinfo b43
[09:43] <slytherin> pitti: in my case b43 got loaded automatically and dmesg had a message that it needs firmware. That is on ibook G4. But I don't have access to the machine right now so can not give more info.
[09:44] <slytherin> pitti: I will be able to help on weekend though.
[09:44] <MacSlow> StevenK, but I started cdbs with the first patch as parameter
[09:44] <pitti> slytherin: thanks; if all goes well, I'll already have a handler to test for you
[09:44]  * mjg59 sighs at the kernel
[09:44] <MacSlow> even then it applied in the cdbs-session
[09:44] <mjg59> It's being mean :(((
[09:44] <StevenK> MacSlow: Wierd. I'd follow pitti's suggestion of using cdbs-edit-patch
[09:45] <StevenK> mjg59: Show me on the doll where the nasty kernel touched you
[09:45] <pitti> mjg59: (BTW, jockey does not only look for pci modaliases)
[09:45]  * StevenK hides
[09:45] <mjg59> pitti: Sweet
[09:46] <pitti> . o O { and I thought that modalias paradigm would be a reliable standard nowadays... }
[09:46] <mjg59> pitti: If ssb isn't exposing the modaliases in sysfs (it's certainly generating the uevents) then it probably needs fixing
[09:47] <TheMuso> pitti: on its way
[09:47] <TheMuso> brb
[09:47] <pitti> TheMuso: cheers
[09:48] <StevenK> pitti: Could I impose on you to process bug #189797?
[09:51] <Hobbsee> bug #189797
[09:51] <Hobbsee> ubotu: ping
[09:51] <ubotu> ping yourself ;-) really the diodes all down my left side are sore
[09:51] <pitti> *chuckle*
[09:51] <StevenK> Hah
[09:52] <Hobbsee> oh, LP is timing out
[09:52] <Hobbsee> ubotu: go back to parking cars!
[09:52] <pitti> StevenK: synced
[09:52] <Hobbsee> [20:52] <ubotu> Error: I am only a bot, please don't think I'm intelligent :)
[09:53] <StevenK> pitti: Thanks!
[09:54] <StevenK> pitti: Can I bug you to let it out of binary NEW after it builds? :-)
[09:54] <pitti> please bug me, yes
[09:54]  * StevenK pastes that quote on the front page of wiki.ubuntu.com
[09:54] <StevenK> :-P
[09:58] <MacSlow> f**king crap!
[09:59] <MacSlow> I did cdbs-edit-patch 01_workspaces_default_name.patch
[09:59] <MacSlow> applied the patches to the .new-variant
[10:00] <MacSlow> from that created the new patch-diff between .new and the origianal
[10:01] <MacSlow> then I tried to apply the resulting patch against a fresh grab of libwnck-2.21.90 I also did cdbs-edit-patch 01_workspaces_default_name.patch for
[10:01] <MacSlow> failed again
[10:01] <TheMuso> pitti: I hope you can find something in there thats useful.
[10:02] <soren> Could an archive team member please take a look at netcat-openbsd? Pretty please?
[10:04] <pitti> TheMuso: aaah
[10:04] <pitti> TheMuso: $ cat ./devices/pci0001:10/0001:10:12.0/ssb0:1/uevent
[10:04] <pitti> MODALIAS=ssb:v4243id0807rev02
[10:05] <pitti> TheMuso: did you tar this up as root or as user? and with -p?
[10:05] <TheMuso> pitti: as user, without -p. I can do it again...
[10:05] <pitti> TheMuso: as user is great
[10:05] <pitti> TheMuso: can you please check the permissions of /sys/devices/pci0001:10/0001:10:12.0/ssb0:1/uevent on your system?
[10:06] <TheMuso> pitti: -rw-r--r--
[10:06] <pitti> cool
[10:06] <TheMuso> -rw-r--r-- 1 root root 4096 2008-02-07 20:03
[10:09] <pitti> mjg59: does it sound like a good idea to consider 'uevent' files which have a MODALIAS= line instead of/in addition to looking at 'modalias' files in /sys/devices ?
[10:09] <ion_> dholbach: D’oh, i knew i forgot something. Thanks. :-) (Re: the missing (LP: #...) in the changelog)
[10:09] <mjg59> pitti: Erm. Can we guarantee that the uevent will only be the modalias?
[10:09] <dholbach> ion_: no problem
[10:09] <mjg59> Devices can generate arbitrary uevents?
[10:10] <pitti> $ find /sys/devices -name uevent|xargs grep MODALIAS|wc -l
[10:10] <pitti> 93
[10:10] <pitti> $ find /sys/devices -name modalias|wc -l
[10:10] <pitti> 93
[10:10] <pitti> doesn't seem to make a differnence on my box
[10:10] <mjg59> I haven't checked the defined smeantics
[10:10] <pitti> TheMuso: I take it above commands deliver more results for the udevent line on your system?
[10:10] <mjg59> But if it's guaranteed to always provide the modalias, then sure, just check that
[10:11] <pitti> mjg59: well, to save a lot of file i/o I could probably just do it for the b43 module
[10:11] <pitti> the other things we support (nvidia, fglrx et al) don't seem to need it
[10:12] <pitti> or, generally just for stuff below /sys/devices/.../ssb*/...
[10:14] <pitti> (that seems like the best option to me)
[10:26] <hubuntu> hello guys.. Wil network manager 0.7 be included in hardy? to coupe with netman 0.65 for an LTS release is probably not the best idea?
[10:27] <hubuntu> fedora 8 has it already and its rocking.... Will netman 0.7 make it before the feature freeze?
[10:31] <cjwatson> hubuntu: at present it's unlikely since the rewrite doesn't look like it'll be in good enough shape for us by feature freeze
[10:31] <cjwatson> although the decision is not final
[10:32] <cjwatson> hubuntu: LTS is not, in general, an argument for switching to new versions of software that are a complete rewrite and are still as yet rather lightly tested
[10:32] <hubuntu> then it will probably end in backports? I
[10:32] <soren> Quite the contrary, really.
[10:33] <cjwatson> backports obviously has a much less strict policy
[10:34] <hubuntu> I know, I'm just all excited about 0.7 and been waiting for a long time for its inclusion... Will the gnome-obex-vfs be included by default? Phones and ubuntu work perfectly now with it, shouldn't it probably be included by default?
[10:36] <hubuntu> it's just a shame that the bluetooth applet is there with the option to browse the phone and it just doesn't work out-of-the-box and has to be installed manually...
[10:36] <seb128> hubuntu: obex has not been ported to gvfs yet so it doesn't work at all in hardy
[10:37] <hubuntu> cjwatson: yeah, but backports are not supported...
[10:38] <cjwatson> that's correct
[10:38] <hubuntu> seb128: so we are going for gvfs and drop bluetooth support? Anyone thinking about blueman as an option? I do not know if it supports gvfs but it sure works better than anything else i have tried...
[10:39] <hubuntu> ok then hardy+1 will be the one to fiunally bring VPN to the masses :)
[10:39] <seb128> hubuntu: dunno about bluetooth
[10:45] <slytherin> hubuntu: For bluetooth work you should probably keep a watch on this blog, http://www.hadess.net/ :-)
[10:47] <asac> hubuntu: are you using my nm 0.7 packages from ppa?
[10:47] <hubuntu> I have been giving it a try and things seem to work fine with several pcmcia cards
[10:48] <hubuntu> d-link and netgear cards.. some with restricted drivers other with atheros and one with some open driver i don'æt remember
[10:48] <hubuntu> wpa enterprise wortk like a dream
[10:48] <hubuntu> ;)
[10:48] <asac> hubuntu: do you have fedora?
[10:48] <hubuntu> no
[10:48] <hubuntu> well I've seen it
[10:49] <hubuntu> and tried it a bit
[10:49] <asac> ok ... would be curios to know if we lack some features compared to their release
[10:50] <Hobbsee> ah yes, i need to try out the mangler again
[10:53] <hubuntu> I am justthere was an update i haven't checked on fedora 8: http://liquidat.wordpress.com/2008/01/24/networkmanager-enterprise-encryption-eduroam-style-works-again/
[10:58] <hubuntu> good work with those packages asac :) Netman really is the new Chuck Norris and you are the ubuntu Chuck Norris ;)
[10:58] <hubuntu> that is logic, right?
[10:58] <asac> lol
[11:01] <slytherin> asac: which PPA has latest network manager?
[11:01] <hubuntu> well it seems bluetooth got even better with gvfs :) - gnome-obex-send is dead, long live bluetooth-sendto. @ http://hadess.net gives me peace of mind
[11:01] <hubuntu> his
[11:01] <asac> slytherin: ~asac
[11:01] <asac> slytherin: https://edge.launchpad.net/~asac/+archive
[11:02] <slytherin> asac: oops I forgot, PPA doesn't yet support powerpc arch right?
[11:02] <hubuntu> jo'er..
[11:06] <asac> slytherin: nope ... you need libnl + wpasupplicant + nm + nm-applet iirc
[11:07] <slytherin> asac: that means I will first have to set the pbuilder on my ibook and then build these packages.
[11:07] <Hobbsee> asac: is it safe just to add the repo, or?
[11:11] <hubuntu> asac you should follow this guy closely: http://blogs.gnome.org/dcbw/ But i guess that's old news for you...
[11:11] <hubuntu> thanks for all your help :)
[11:11] <asac> Hobbsee: for now you need to look what packages i have in there. I will use the ppa of ~network-manager team in future.
[11:12] <Hobbsee> asac: right.  when will that change over?
[11:13] <asac> Hobbsee: next time i upload a nm to ppa. not ETA yet
[11:13] <Hobbsee> ok
[11:20] <pitti> thekorn: current p-lp-bugs uses any() and thus crashes on python 2.4
[11:20] <pitti> thekorn: would you mind if that was rewritten to be 2.4 compatible or shall we just hack it locally?
[11:22] <pitti> thekorn: I just manually added a definition for any() to commentsbase.py now :)
[11:37] <YokoZar> Is "Switch Page Direction" still on the right click menu for Hardy FireFox?
[11:39] <Hobbsee> doesn't look like it
[11:50] <TheMuso> pitti: Sorry, been on the phone. I'll run sed commands and give you the results.
[11:50] <pitti> TheMuso: thanks
[11:51] <TheMuso> pitti: 53 for the first find command, and 51 for the second.
[11:51] <pitti> TheMuso: ok, then you have two which are undetected by scanning for modalias; I guess those are the two ssb ones
[11:52] <pitti> TheMuso: when I'm done with wrestling with GTK, I'll tackle this to fix it generally in jockey
[11:52] <TheMuso> pitti: No rush, I'm about to go to bed anyway, so feel free to email/leave an IRC message if further testing/commands are neded.
[11:52] <pitti> TheMuso: I'll put your /sys layout of the ssb devices into the test suite
[11:53] <pitti> TheMuso: I will, thanks (or ask slytherin)
[11:59] <calc> asac: been working on apt/dpkg backporting
[12:01] <asac> calc: ok, today?
[12:04] <calc> asac: i'll try to get to it today, i'm still working on the backports but will get on the xulrunner stuff once they are done
[12:04] <asac> backports?
[12:04] <calc> asac: i'm having to backport apt/dpkg lzma patches (once i find them all) to dapper and edgy for the buildds/servers
[12:05]  * calc is going back to sleep now, his son had woken him up :-\
[12:06] <asac> calc: ok. sleep well
[12:07] <thekorn> pitti: py-lp-bugs is supposed to work under py2.4, so I will remove this 'any'/'all' statements
[12:07] <pitti> thekorn: ah, danke
[12:08] <thekorn> gerne
[12:14]  * torkel hugs pitti for finally fixing #148003
[12:14] <pitti> \o/
[12:15] <Ng> win 141
[12:16] <pitti> Ng: >= 141 channels? wow, that's what I call a busy IRC client :)
[12:17] <Ng> pitti: or a lazy irc operator. once I get past 19 windows and can't use the alt key anymore, it becomes irrelevant how many there are, so I just don't bother closing them ;)
[12:19] <pitti> Ng: heh, that's why I try to keep them <= 10
[12:20] <ion_> I use query autoclose.
[12:23] <StevenK> pitti: With irssi you can use alt + q w e .... for above ten
[12:23] <pitti> ah, neat
[12:23] <Amaranth> ooh
[12:23] <StevenK> Heh
[12:23]  * Hobbsee likes alt+a
[12:24] <pitti> alt+a is what I use for these, too
[12:24] <pitti> since I already put commands on alt+letter
[12:24]  * mjj29 has bound all the letters and numbers to channels
[12:24] <ion_> I had all the letters bound to window change back in the day. Nowadays i try to limit the amount of channels. :-)
[12:24] <mjj29> and alt-` for active-window
[12:24] <ion_> I have tab bound to active window.
[12:24] <mjj29> ion_: just tab?
[12:24] <ion_> Yep
[12:25] <mjj29> what about tab-completion
[12:25] <ion_> I don’t find it useful on IRC. In shell, it is really useful.
[12:27] <ion_> (At least when you have a shell where you can type /u/s/d/compi/RE and get to /usr/share/doc/compiz-core/README with a small amount of keypreses)
[12:28] <Mithrandir> ion_: it's useful for nick-completion, I find. I've pondered binding it to word-completion too, but suspect I'd develop zshenia in my irc client too.
[12:29] <Hobbsee> Mithrandir: is this a problem?
[12:30] <Mithrandir> Hobbsee: it wears out my tab key.
[12:30] <StevenK> zsh users only need Tab and Enter anyway
[12:30] <Mithrandir> on the other hand, I should be able to just bind all keys on my keyboard to tab and make do with that.
[12:30] <ion_> Nicks are typically short, and i prefer to talk to somedude instead of _someDUDE77^ no matter which nick he chooses to use. Also, every so often you complete to the wrong nick and only notice it after hitting enter. :-)
[12:30]  * StevenK high fives Mithrandir 
[12:30] <Mithrandir> ion_: I'm lazy. :-P
[12:31] <Mithrandir> and people suck at speling Mithrandir
[12:31] <StevenK> pitti: fbreader hasn't built on ia64, if you can cope with that, could you let it out of binary NEW?
[12:33] <Hobbsee> Mithrandir: and in the event of that, you then bind the tab function to your capslock.  problem solved.
[12:34] <pitti> StevenK: done
[12:34] <ion_> My caps lock is control, but i’ve been considering making it an escape instead. :-)
[12:34] <Mithrandir> mine's compose.
[12:35] <ion_> My right hand Windows® key is compose.
[12:35] <Mithrandir> laptop doesn't have a windows key, so. :-)
[12:35] <Hobbsee> mine does.  you need a bigger laptop
[12:36] <Hobbsee> or to bind it to right alt or something.
[12:36] <Keybuk> my caps lock key is CAPS LOCK
[12:36] <Mithrandir> I use right alt for random things.
[12:36]  * Hobbsee used to bind things to insert, long ago.
[12:36] <Hobbsee> Keybuk: yes, but you're boring :P
[12:37] <ion_> I use the right alt mostly for “, ” and –.
[12:37] <Keybuk> about the only useless key on the keyboard for me is "Pause/Break"
[12:38] <pitti> and the menu key for me; I have never used it
[12:38]  * ion_ just noticed his “Logitech” key opens a calculator. :-)
[12:38] <ion_> An idea: make it open irb in a terminal!
[12:39] <Mithrandir> I have a calculator key on my desktop keyboard.
[12:39] <Mithrandir> and a house key.  And one with a letter (as in, what you put in a mailbox) on it.
[12:40] <Amaranth> I have a calculator key on my laptop
[12:40] <charles_> "By pressing down a special key It plays a little melody"
[12:40] <Amaranth> It's handy, even enables numlock for me
=== Amaranth_ is now known as Amaranth
=== \sh_away is now known as \sh
[13:12] <pitti> TheMuso, cjwatson: if you have a moment, could you please try running http://people.ubuntu.com/~pitti/tmp/detect-test.py and pastebin the output for me? it's the /sys walking ripped out of jockey, now with SSB support
=== cjwatson_ is now known as cjwatson
[13:12] <pitti> it passes my test suite with TheMuso's /sys layout, but I'd like to test it on real hw first
[13:15] <cjwatson> pitti: http://paste.ubuntu.com/4287/
[13:16] <pitti> cjwatson: rocking' that works; it even detects that the module b43 is linked to it
[13:16] <pitti> cjwatson: thanks!
[13:16] <cjwatson> np
[13:30] <tbf> glatzor: since you packaged policykit for gutsy... do i have to take some special actions, before i can use policykit?
[13:30] <tbf> glatzor: currently there seems to be no 'org.freedesktop.PolicyKit.AuthenticationAgent' service on the session bus
[13:32] <glatzor> tbf: I only backported them from debian experimental some monthes ago
[13:32] <tbf> glatzor: hmm. ok.
[13:32] <glatzor> tbf: sorry, but I haven't looked at them for a time.
[13:32] <tbf> thanks
[13:50] <ScottK> pitti: I was wondering if there was any chance of MIR processing today?  I still have faint hopes of integrating amavisd-new in the mail server tasksel before feature freeze...
[13:55] <pitti> yeah, that was my plan; I'll start after the team meeting
[13:55] <pitti> ScottK: doko wanted to process some today, too
[13:56] <ScottK> Thanks.  He seemed skeptical about amavisd-new.  I'd rather you looked at it. ;-)
=== doko_ is now known as doko
[14:42] <emgent> heya
[14:46] <pygi> siretart, poke
[15:04] <pitti> cprov, mvo: hm, the symlink should have been mirrored by now, but it's still not on a.u.c.
[15:05] <mvo> pitti: ok, I will mail elmo then
[15:05] <pitti> I guess symlinks pointing outside of /ubuntu don't work
[15:05] <pitti> so I'll do a copy instead
[15:05] <mvo> pitti: he mentioned that the mirror script used may not pick it up
[15:05] <pitti> ah
[15:05] <pitti> mvo: ok, please do that first then
[15:05] <mvo> pitti: I will and will CC you
[15:05]  * pitti hugs mvo
[15:08]  * mvo hugs pitti
=== greeneggsnospam is now known as jsgotangco
[15:27] <dholbach> superm1: does this look interesting to you:  http://www.murrayc.com/blog/permalink/2008/02/07/gnome-lirc-properties-a-gui-to-configure-infra-red-remote-controls/ ?
[15:50] <superm1> dholbach, yeah it looks quite interesting. It obviously won't take advantage of the patch that I'v got locally (upstream hasn't accepted it yet) for "include" files in lircd.conf
[15:50] <dholbach> superm1: ah nice
[15:51] <superm1> but that is going the right direction, and fairly similar to what upstream eventually had in mind
[15:52] <superm1> dholbach, hopefully next cycle that person authoring that gets in contact with us and we can work together on making it fit well
[15:52] <dholbach> superm1: that sounds awesome
[15:57] <tseliot> superm1: did you read my email?
[15:57] <tseliot>  dholbach: hi
[15:58] <dholbach> hi tseliot
[16:11] <superm1> tseliot, should I be reading someone else's mail :)?
[16:12] <superm1> tseliot, ah yeah you were thrown in a different filter
[16:12] <superm1> that's why I didn't see it
[16:12] <tseliot> ok
[16:12] <superm1> tseliot, that is with the latest public release?
[16:12] <superm1> or with something else?
[16:12] <tseliot> yep
[16:13] <tseliot> and I also tried the tarball from phoro git
[16:13] <superm1> tseliot, you're on list - can you please try with a testing release?
[16:13] <tseliot> ok
[16:13] <superm1> there has been a significant amount of changes to the packaging scripts between last month and the next upcoming one
[16:14] <tseliot> I need to get Envy to work with it on Hardy
[16:14] <superm1> right
[16:14] <superm1> well that exact issue cropped up before due to the way that dh_shlibs was handling things
[16:15] <superm1> and it should have been corrected within the last couple of weeks
[16:15] <tseliot> I'm downloading the latest test release right now
[16:17] <tseliot> superm1: let's chat in private
[16:17] <superm1> okay
=== ogasawara_ is now known as ogasawara
[16:23] <Catachan> Hello, would this be the proper channel to ask abou how to change the color of the screen that pops up between logging in, and the actual Desktop?
[16:25] <sistpoty|work> Catachan: #ubuntu
[16:25] <Catachan> okay
[16:25] <Catachan> thank you
[16:25] <sistpoty|work> np
=== kyle__ is now known as kylem
[16:42] <pitti> doko: looking at amavisd MIR now
[16:46] <doko> soren: could you have a look at kvm building with gcc-4.2?
[16:47] <soren> doko: I'm realiably informed it won't work.
[16:47] <soren> doko: Well, on i386, anyway.
[16:48] <doko> soren: there are patches floating around, iiuc
[16:48] <pitti> ScottK: just replied to bug 183418; would you mind doing the upload soon, so that we can actually promote it?
[16:48] <ubotu> Launchpad bug 183418 in amavisd-new "MIR for amavisd-new" [High,New] https://launchpad.net/bugs/183418
[16:49] <ScottK> pitti: Will do.  Thanks for the advice.
[16:49] <ScottK> pitti: I'll upload it in the next 24 hours.  Up to my eyeballs in $WORK today.
[16:50] <doko> pitti: looking at the pam/radius MIRs
[16:53] <soren> doko: Only bad ones.
[16:53] <soren> doko: Although..
[16:53] <tgelter>  hey all, is it currently possible to use the nvidia driver with hardy?
[16:53] <tgelter> I encounter errors when I try to install nvidia-glx-new. I also am unable to load the nvidia kernel module after a successful install of the driver straight from nvidia
[16:54] <soren> doko: Fabrice (the qemu guy) wrote a new code generator that should alleviate the need for gcc-3.x. It's x86 only, so far, but that's all kvm should worry about.
[16:54] <tgelter> oh, in relation to my comment about nvidia, I'm running x86_64 hardy
[16:54] <soren> doko: I can check next week. This week is completely booked.
[16:54] <doko> soren: thanks!
[16:56] <soren> doko: er.. When was gcc-3.4 demoted to universe?
[16:57] <soren> doko: That's a *Very* recent change, isn't it?
[16:57] <doko> soren: no, was done at the sprint
[16:57] <soren> doko: That counts as very recent in my book. :)
[16:57]  * soren grumbles a bit
[16:58] <soren> I mentioned kvm needing it on several occasions.
=== thekorn_ is now known as thekorn
[17:07] <pitti> tgelter: that's the same setup that I run here, and n-glx-new works fine for me
=== \sh is now known as \sh_away
[17:12] <tgelter> pitti: brb and then I'll have a question for you
[17:28] <tgelter> pitti: pastebin.com/m34edd00
[17:29] <pitti> tgelter: ah, that indeed does look like a packaging bug; tjaalton?
[17:31] <tgelter> pitti: note that I upgraded from gutsy, this wasn't a fresh install
[17:31] <pitti> right, or that
[17:32] <tgelter> so what're my options?
[17:32] <pitti> tgelter: can you please file a bug against linux-restricted-modules-2.6.24, including this output?
[17:33] <tgelter> sure thing
[17:35] <tgelter> wouldn't this indicate a problem with nvidia-kernel-common rather than linux-restricted-modules?
[17:36] <pitti> tgelter: oh, sorry, indeed; it's a separate source
[17:39] <tgelter> is there anything to explain why I can't load the nvidia kernel module after installing with the binary installer from nvidia?
[17:39] <tgelter> is there a particular log I should be checking to see if modprobe is returning any errors?
[17:41] <tgelter> ah, looks like someone else already created that bug report
[19:14] <zul> pitti: around?
=== ryu2 is now known as ryu
=== hunger_t is now known as hunger
[20:02] <slangasek> ember: any chance you can follow up on bug #187328?  Now that seahorse has been rebuilt against openldap 2.4, there's a potential for crashes on 64-bit archs in the LDAP code
[20:02] <ubotu> Launchpad bug 187328 in seahorse "seahorse: misbuild on 64-bit architectures due to missing ldap prototypes" [High,Triaged] https://launchpad.net/bugs/187328
[20:15] <ember> slangasek sure, thanks
[20:23] <geser> does somebody know if gimp will stay at 2.4.3 for hardy or will it get updated to 2.4.4?
[20:25] <geser> I try to figure out what to do with ingimp as it wants gimp < 2.4.3 and the latest version in Debian already wants gimp 2.4.4 (but there was a version in unstable generating the right depends on gimp)
=== Kopfgeldjaeger is now known as KGJ|EEE_bestellt
[20:42] <hendrixski> is there a separate tool from apt-ftparchive for creating debug symbol repositores?  because it doesn't seem to pick up ddeb packages :-(
[21:22] <slangasek> doko: are you aware of cwidget being dep-wait on i386 because it's grown an ikiwiki build-depends-indep?
[21:23] <LCID_Fire> Hi
[21:23] <doko> slangasek: now I'm aware of it. thanks
[21:23] <LCID_Fire> Could anyone give me a hand with grep?
[21:24] <ScottK> LCID_Fire: #ubuntu for support.
[21:24] <LCID_Fire> k
[21:25] <slangasek> doko: ok, cheers :)
[21:27] <hendrixski> I've made some debug symbol packages, but I can't seem to put them into a repository.  Is there a way to get dpkg-scanpackages or apt-ftparchive to pick those up?
[21:28] <geser> slangasek, doko: cwidgets needs also libhtml-scrubber-perl from universe to fulfil build-depends-indep
[21:32] <hendrixski> pitti, I've seen your name on a few posts dealing with repositories of ddeb packages.  If you get this, can you direct me to some documentation about how to make them.  Thanks :-)
=== Igorot_ is now known as Igorot
=== Igorots is now known as Knightlust
=== KGJ|EEE_bestellt is now known as Kopfgeldjaeger
[22:39]  * jdong reads Fedora 9's release notes and learns there's now something called PackageKit
[22:39] <jdong> hmm... redhat camp seems really Kit happy
[22:39] <TheMuso> jdong: IMO its shameful.
[22:39] <TheMuso> :p
[22:39] <lifeless> I find the *Kit naming bizarre
[22:39] <TheMuso> lifeless: likewise.
[22:39] <jdong> likewise
[22:39] <jdong> will we get BazaarKit anytime soon? ;-)
[22:40] <jdong> because I really version control belongs on top of dbus and HAL
[22:40] <StevenK> Are they trying to pay homage to Knight Rider or something?
[22:40] <lifeless> we have it, import bzrlib.plugins.dbus
[22:40] <lifeless> kthxbye
[22:40] <jdong> :D
[22:40] <jdong> there's actually a dbus plugin?
[22:40] <lifeless> https://launchpad.net/bzr-dbus
[22:40] <slangasek> KatKit, for managing your pet's diet in a trademark-infringing manner
[22:40] <jdong> StevenK: *Kit feels really Apple-y to me
[22:40] <TheMuso> I think I'll make a speechkit./
[22:41] <TheMuso> Or TTSKi.
[22:41] <TheMuso> TTSKit
[22:41] <jdong> TheMuso: shouldn't that be a part of ConsoleKit?
[22:41] <jdong> :D
[22:41] <jdong> what's the subunit of a Kit?
[22:41] <TheMuso> jdong: No.
[22:41] <lifeless> jdong: (in a word - yes there really is a dbus plugin)
[22:41] <lifeless> jdong: and avahi
[22:42] <TheMuso> avahi I can understand... But not dbus.
[22:42] <jdong> lifeless: I can understand the value of avahi branch announcements
[22:42] <jdong> but... dbus...
[22:42] <lifeless> bzr commit-notify
[22:42] <lifeless> listens on dbus
[22:42] <lifeless> gtk window pops up on each commit/push/pull/uncommit
[22:42] <lifeless> or branch
[22:42] <jdong> interesting
[22:43] <lifeless> bzr lan-notify shares these on the LAN
[22:43] <jdong> what's next? coreutils-dbus?
[22:43] <lifeless> so you commit to a public branch, I get a commit notify back through dbus and gtk
[22:43] <jdong> the ability to pipe things into dbus and then for another coreutil to fetch it out? :D
[22:44] <jdong> I guess I just had a very different conception of what "utopia" meant :)
[22:45] <TheMuso> lifeless: I actually like that.
[22:45] <TheMuso> THat gtk notification stuff.
[22:45] <lifeless> thanks :)
[22:45] <lifeless> TheMuso: I haven't done much/any a11y stuff for it
=== Ubulette_ is now known as Ubulette
=== rzr is now known as rZr
[22:47] <TheMuso> lifeless: Oh I'm not so worried about that, although I should check it out some time. I just like the window notifications of branch updates.
[22:47] <lifeless> do you use it regularly ?
[22:49] <TheMuso> lifeless: I've never used it, but I'm interested enough to check it out.
[22:49] <lifeless> :)
[22:52] <lifeless> bbiab
=== cjwatson_ is now known as cjwatson