2008/05/04/#ubuntu-devel.txt

[02:34] <sven_> hello, world!
[02:34] <sven_> (thought it was appropriate)
[02:35] <tritium> sven_: see /topic
[02:36] <sven_> tritium: thanks, on my way to motu
[02:52]  * calc sees interesting rumors of nvidia releasing docs/open source driver
[02:52]  * calc thinks hell will freeze over before that happens
[02:53]  * calc has seen nvidia's 'attempts' before ~ 10 years ago
[03:00] <johanbr> calc: Well, the situation is a bit different now, with the competition already having open specs/drivers.
[03:04] <wgrant> Nouveau is doing an excellent job even without Nvidia's help.
[03:05] <johanbr> Yes, but a reverse-engineered driver will always be playing catchup.
[03:06] <wgrant> This is true.
[03:06] <calc> johanbr: yea
[03:07] <calc> johanbr: from what i recall at the time there were open drivers for the competition as well, ati has an on again off again open source strategy
[03:07] <calc> it looks like now that amd owns them they may stay foss friendly
[03:08] <calc> and amd's help appears to have convinced intel to release documentation along with their drivers :)
[03:11] <wgrant> Is there anyone else around other than NVIDIA, ATI and Intel?
[03:11] <calc> Via, but not really
[03:12] <Nafallo> sis?
[03:12] <calc> they claimed to be going foss but phoronix had an interesting article mentioning their foss efforts appear to be binary only drivers
[03:12] <wgrant> Hahhah SiS.
[03:12] <calc> SiS is what ~ < 0.01% of the market
[03:12]  * jdong hugs Nafallo 
[03:13] <jdong> that was the cheer-up I needed for the long day I've had
[03:13] <Nafallo> :-)
[03:13] <calc> SiS has always existed but I have only seen one time
[03:13] <jdong> calc: the last time I saw them it didn't leave a good impression :)
[03:13] <wgrant> I've got one SiS chipset here.
[03:13] <Nafallo> ooh
[03:13] <Nafallo> I know another one...
[03:13] <wgrant> For some reason, Compiz decides it's OK to start on it.
[03:13] <Nafallo> it's m something :-)
[03:13] <calc> and aiui SiS binary drivers suck so hard it probably wouldn't help to have docs
[03:13] <wgrant> It works sort of.
[03:13] <Nafallo> matrox
[03:13] <wgrant> Hah.
[03:13] <Nafallo> I think
[03:14] <calc> i have seen bug reports about SiS can't even do regular 2D right
[03:14] <wgrant> calc: That's right.
[03:15] <calc> matrox is all but dead in the market, they have very specific area that they still sell into
[03:15] <wgrant> What's that? CAD workstations?
[03:16] <jdong> doom95 gamers.
[03:17] <calc> wgrant: airports apparently
[03:17] <calc> wgrant: and financial large number of screens type installs
[03:17] <calc> not much use for 3d, etc
[03:17] <wgrant> Ah.
[03:17] <wgrant> I thought they were for 3D stuff.
[03:17] <wgrant> A rather niche market, anyway.
[03:18] <calc> the last time (i know of) that matrox was foss friendly was ~ 1998-2000 timeframe
[03:18] <calc> back when their video cards were still relevant on the desktop
[03:19] <calc> some of the first 3d support under linux was for the matrox g200
[03:20] <calc> looks like they do quad head on one slot without fan
[03:20] <Nafallo> they are still good for multiscreen setup indeed :-)
[03:20] <calc> yea
[03:21] <Nafallo> which I could have a fun setup like that ;-)
[03:23] <calc> there is also that company XGI
[03:23] <calc> never seen anything from them though
[03:24] <calc> doesn't look like they are sold by many companies either
[03:36] <calc> hmm finally 10GbE over regular copper
[03:36] <Nafallo> has been around for a while :-)
[03:37] <calc> Nafallo: not over copper
[03:38] <Nafallo> yes it have :-)
[03:38] <calc> eh?
[03:38]  * calc looks it up
[03:38] <Nafallo> several months
[03:39] <calc> says the standard was released in 2006 and intel announced their card on thu (afaict)
[03:40] <calc> but maybe someone else released cards before that, dunno
[03:40] <Nafallo> connectix have sold 10GbE solutions for some time :-)
[03:41] <calc> the only connectix i can find info about is the one that wrote the mac software
[03:42] <calc> but yes there was 10GbE for a long time, but over fiber, since at least 2002
[03:42] <Nafallo> http://www.connectix.co.uk/
[03:42] <Nafallo> :-)
[03:44] <calc> ok
[03:46] <calc> of course desktops are barely able to saturate regular GbE currently
[03:47] <Nafallo> very true
[03:49] <protonchris> Anybody had any luck getting a intrepid pbuilder chroot going?
[03:50] <Nafallo> calc: hehe. they've started to think about 40GbE and 100GbE now :-)
[03:54] <jdong> protonchris: yeah I've got one already
[03:55] <jdong> you need to use intrepid's debootstrap but otherwise it was a simple matter
[03:55] <protonchris> Ok.  I thought that the debootstrap for intrepid was backported to hardy.
[03:55] <jdong> not yet that I know of
[03:55] <jdong> oh nvm it's there now
[03:56] <jdong> wasn't there last night yet
[03:57] <protonchris> It doesn't work for me.  Dies when "Installing core packages"
[03:59] <protonchris> It runs into trouble with libc6_2.7-10ubuntu3_i386.deb
[04:05] <wgrant> It's due to a missing dep in debconf, AFAICT.
[04:08] <protonchris> ok
[04:13] <illSleepWheniDie> hi all
[04:14] <illSleepWheniDie> can all the indians here say Here!
[06:16] <Hobbsee> stgraber: your blog is broken.
[06:17] <Hobbsee> stgraber: Fatal error: Call to undefined function drupal_submit_form() in /data/www/stgraber.org/www/modules/akismet/akismet.module on line 707
[06:18] <wgrant> Hobbsee: It *is* PHP...
[06:18] <Hobbsee> wgrant: this is true
=== hunger_t_ is now known as hunger
[09:56] <stgraber> Hobbsee: akismet was broken, I disabled it for now.
[10:24] <qense> is there an ALSA developer online? Or maintainer :)
=== qense is now known as qense|lunch
=== qense|lunch is now known as qense
[12:13] <fursund> Hello... In the early alpha-betas of hardy, there was this nifty mount archive - right click in nautilus - tool, do you know where I can download and install that?
=== bureflux is now known as afflux
[12:23] <qense> Is there an ALSA maintainer online?
=== gnomefre2k is now known as gnomefreak
[13:28] <Hobbsee> stdin: ah, cool
[13:40] <alex-weej> kernel doesn't boot on a mate's laptop
[13:40] <alex-weej> i noticed -17 came out recently
[13:40] <alex-weej> any way to test it with a disc image? do we still make daily cd's after release?
=== nand_ is now known as nand
[14:13] <_MMA_> qense: crimsun will kinda be the guy to talk to. But since it's Sunday I would try back tomorrow. Or at *least* later today. (4 hours or so).
[15:06] <qense> what package is responsible for managing /proc/acpi/fan?
[15:06] <Hobbsee> qense: check with dpkg -S ?
[15:07] <qense> I thought /proc didn't contain files
[15:08] <stdin> the kernel manages all files in /proc
[15:08] <qense> ok
[15:08] <qense> I'm triaging a bug wher fan support disapeared in hardy
[15:09] <qense> everyting was alright in hardy
[15:09] <qense> do you think it's a 2.6.24 issue?
[15:09] <qense> it's filed again acpi-support
[15:11] <qense> the bug number might help you help me. ;) bug 226422
[15:11] <ubottu> Launchpad bug 226422 in acpi-support "Laptop gets very hot" [Medium,Confirmed] https://launchpad.net/bugs/226422
[15:38] <johanbr> qense: Yes, that's a kernel issue.
[15:38] <qense> ok thx I've already changed the package to linux
[15:40] <johanbr> Is frequency scaling working?
[15:41] <qense> I haven't asked him yet
[15:41] <qense> I'll do
[15:41] <alex-weej> when i plug a DVI cable into my notebook, nothing happens. is something supposed to happen or are we not quite there yet?
[15:41] <qense> he says the fan is running all time
=== blueyed_ is now known as blueyed
=== asac_ is now known as asac
[17:03] <hwilde> Driver 'sd' needs updating - please use bus_type methods
[17:04] <hwilde> Driver 'sr' needs updating - please use bus_type methods
[17:04] <hwilde> bootup looks good until those errors
[17:05] <qense> does it actually start or holds it at that point?
[17:05] <hwilde> hodls
[17:05] <hwilde> holds
[17:06] <hwilde> strange part is there is no sd device
[17:06] <hwilde> I don't even know why it's loading that
[17:07] <qense> weird
[17:08] <hwilde> http://pastebin.com/m31bc6bcd   lines 49 & 59
[17:08] <hwilde> then is just says   sda:
[17:08] <hwilde> and sits there
[17:08] <Chipzz> hwilde: errr, you DO know that sd has NOTHING to do with SD cards, do you?
[17:09] <qense> does /var/log/boot contain anything?
[17:09] <hwilde> Chipzz, I dunno it's right in line with it looking for sda in the boot messages.  check my pastebin
[17:09] <hwilde> qense, I dunno it doesn't boot up for me to check the log
[17:10] <qense> can you access the harddisk at another way?
[17:10] <Chipzz> hwilde: no interest. just mentioning because you may be seriously mistaken
[17:10] <Chipzz> sd is a scsi driver
[17:11] <Chipzz> which in recent (for some value of recent) has started to be used by IDE too
[17:11] <hwilde> strange I don't have scsi either
[17:11] <Chipzz> + kernels
[17:11] <hwilde> oh it does load scsi for the cdrom
[17:11] <hwilde> hmm it's a hard lockup, ctrl+alt+del doesn't even respond
[17:11] <Chipzz> not only that
[17:11] <hwilde> brb
=== fta_ is now known as fta
[17:12] <Chipzz> hwilde: it should also be mentioned that this channel is not for bug-reports. And since this is a kernel bug, that's one more reason not to ask here
[17:13] <Chipzz> you should, at the very least, ask in #ubuntu-kernel instead of here
[17:13] <hwilde> well... I wan't intending to submit a bug report
[17:13] <hwilde> and if it says a driver needs updated, that sounds like development
[17:13] <hwilde> but thanks for pointing me in the right direction
[17:13] <Chipzz> it isn't
[17:14] <Chipzz> but people should be able to help you in #ubuntu-kernel
[17:14] <Chipzz> if you follow that reasoning, just about everything is "development" ;)
[17:15] <hwilde> sure but that same logic could be used to say that everything is a kernel issue
[17:15] <Chipzz> no :)
[17:15] <Chipzz> it can't :)
[17:16] <Chipzz> if I write a python program with a bug in it, how is that a kernel bug? :)
[17:16] <Chipzz> anyway
[17:16] <hwilde> oh right this isn't gentoo
[17:16] <Chipzz> make sure to check the bugtracker for similar bugs first
[17:16] <Chipzz> chances are someone else already reported it
[17:16] <Chipzz> oh and one more thing
[17:17] <Chipzz> It's currently weekend. Chances are you will not receive much response on #ubuntu-kernel today
[17:17]  * hwilde stares at Chipzz 
[17:17] <Chipzz> yes?
[17:17] <hwilde> you have been most helpful sir
[17:17] <hwilde> thank you for the encouragement
[17:18] <Chipzz> just stating some facts :P
[17:18] <Chipzz> ubuntu developers like to have the weekend off too ;)
[17:18] <hwilde> that's why the irc god invented /away mode
[17:18] <Chipzz> and no need for "sir" ;P
[17:19] <hwilde> sir yes sir, thank you for correcting my salutations!
[17:19] <hwilde> hehe
[17:19] <Chipzz> hwilde: there's also a reason god/irc invented topics. Yet people seem to ignore them on a constant basis :P
[17:20] <Chipzz> (that was meant as a general comment, not aimed at you ;))
[17:21] <Chipzz> anyway, good luck with your bug!
[17:22] <hwilde> actually it finally dropped to busybox
[17:22] <hwilde> says /dev/hda1 does not exist
[17:22] <hwilde> so I think it's just coming up under a different name
[17:24] <Chipzz> hwilde: hrrrm, it should be using UUID's instead of device names
[17:24] <hwilde> yeah I am trying to make an image that can be cloned to multiple of the same devices
[17:25] <Chipzz> ouch
[17:25] <Chipzz> in that case, I suggest you kinda keep a close eye on ubuntu development
[17:25] <hwilde> I never could figure out why they use UUIDs
[17:25] <Chipzz> exactly for this reason :)
[17:25] <Chipzz> when the kernel changed from using ie hda to sda, it wouldn't have mattered
[17:26] <mjg59> hwilde: Because device node names aren't stable
[17:26] <qense> what; s the difference between hda and sda>
[17:26] <Chipzz> qense: the way the kernel sees the device
[17:26] <Chipzz> hda is an ide device
[17:26] <Chipzz> sda is a scsi device
[17:26] <qense> ah
[17:26] <hwilde> well, how would you make /boot/grub/menu.lst and /etc/fstab work on multiple different devices?   the UUIDs are different but /dev/hda1 used to work.  now it's apparently /dev/sda1
[17:26] <Chipzz> or an ide device using scsi emulation
[17:26] <hwilde> but I didn't change it so now the ide is using scsi deriver
[17:27] <Chipzz> hwilde: well in your case I would make sure to read the release notes between upgrades, if you want to stick with hda/sda instead of UUIDs
[17:28] <Chipzz> basically, doing that is unsupported, and if you want to do it, you need to take care of it yourself ;)
[17:29] <ogra> (and no, thats not ubuntu specific, its an pstream kernel change )
[17:29] <ogra> *upstream
[17:30] <Chipzz> btw, unsupported does not mean it won't work. it does mean however, that if it breaks, you get to keep both pieces ;)
[17:30] <ogra> :)
[17:32] <ogra> > I killed Google and the
[17:32] <ogra> > Internet stopped being busy.
[17:33]  * ogra grins about ubuntu-users
[17:33] <qense> why?
[17:34] <ogra> see the quote i pasted :)
[17:35] <qense> well, I think it would make internet even more busy! everyone would have to ask the same question over and over agin
[17:35] <qense> :d
[17:35] <ogra> heh, indeed
[17:36] <qense> #ubuntu would be blown away
[17:37] <hwilde> you should have seen it when ubottu wasnt there :/
[17:39] <qense> at peeks #ubuntu is already terrible
[17:41] <hwilde> depends on how you look at it I guess
[17:42] <hwilde> we could have no users and no interest... then the channel would be quiet
[17:42] <qense> true
=== gnomefre1k is now known as gnomefreak
[18:25] <jeromeg> jdong: hello
[18:26] <jeromeg> do you have a few seconds ?
[18:38] <ogra> hwilde, hey i totally forgot how much fun #ubuntu can be ... thanks for reminding (sadly working for canonical doesnt often leave time for that though)
[19:31] <cjwatson> protonchris: the intrepid bootstrapping problem is http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=479202; I'm inclined to leave it for a while on the assumption that it'll get fixed in Debian soon and we can just sync it, since intrepid isn't really expected to work at the moment
[19:31] <ubottu> Debian bug 479202 in perl-base "fields tries to use Hash::Util" [Unknown,Open]
[19:36] <protonchris> cjwatson: Thanks for the info.  Syncing it sounds like a good idea to me.
[19:36] <cjwatson> once it's fixed in Debian, that'll happen automatically
[19:57] <kees> hm, can someone give-back glib2.0 (2.16.3-2) for i386?  I can't reproduce the test failure, so I'm hoping it a timing bug
[19:57] <kees> https://edge.launchpad.net/ubuntu/+source/glib2.0/2.16.3-2
[20:41] <emgent> heya
[21:08] <bigon> I'm trying to create an intrepid chroot and I get http://paste.debian.net/2194/ should I report that?
[21:24] <pitti> kees: kicked
=== thegodfather is now known as fabbione
=== cr3_ is now known as cr3
=== ryu2 is now known as ryu
[22:07] <cjwatson> bigon: intrepid *will* break semi-randomly for some time
[22:08] <cjwatson> bigon: but in any case that's http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=479202
[22:08] <ubottu> Debian bug 479202 in perl-base "fields tries to use Hash::Util" [Unknown,Open]
[22:08] <cjwatson> bigon: for now, either (a) don't be in a rush to create an intrepid chroot :-) or (b) create a hardy chroot and upgrade it
[22:17] <Kopfgeldjaeger> n8
=== blueyed_ is now known as blueyed
=== TheMuso_ is now known as TheMuso
[23:47] <TheMuso> crimsun: Note that the sample cachign thing is an esound thing, not a sound card thing.
[23:47] <crimsun> TheMuso: yes, but there's nothing that would have changed in PA or ESD to issue it.
[23:47] <TheMuso> True.
[23:59] <ogra> OMG
[23:59] <ogra> transmission is surely the worst crap i have ever seen
[23:59] <LaserJock> uh oh
[23:59] <ogra> it just deleted about 1G of music i downloaded over the last months
[23:59] <LaserJock> ouch