[12:10] <fabbione> hey Keybuk
[12:10] <fabbione> Keybuk: i got a regression in dapper... not sure if it's udev...
[12:11] <fabbione> midi* devices are created in /dev/
[12:11] <fabbione> they really should be in /dev/snd/
[12:14] <Keybuk> k, which midi devices?  snd or oss?  what names?
[12:14] <fabbione> snd
[12:14] <fabbione> names are midi*
[12:14] <fabbione> basically all midi devices on my tivo are in /dev
[12:15] <fabbione> if you need the details i need a few minutes to power it on
[12:15] <Keybuk> ok, fixed
[12:15] <fabbione> thanks
[12:42] <lifeless> Keybuk: the tried and true method of 'wait a week, update, reboot' fixed my udev glitch. Thanks
[12:45] <Keybuk> cool
[02:05] <Keybuk> jbailey: yeah, they're all from musicals
[02:05] <Keybuk> can you name them all though? :p
[02:08] <jbailey> Keybuk: No.  I have to admit that my musical tastes tends to extend to top broadway only, and then only from the last 20 years. =)
[02:09] <Keybuk> they're all pretty big musicals, iirc
[02:09] <Keybuk> well, other than the first one <g>
[02:09] <Keybuk> Buffy/Once More With Feeling - What can't we face?
[02:09] <Keybuk> Oliver - Got to pick a pocket or two
[02:10] <Keybuk> Closer to Heaven - My Night
[02:10] <jbailey> I'm only halfway throught season 2 of buffy. =)
[02:10] <infinity> Once More is season 6.
[02:10] <Keybuk> Rocky Horror Show - Charles Atlas Song
[02:10] <jbailey> I missed Oliver, but do plan to see i.
[02:10] <jbailey> The RHPS I recognised.
[02:10] <infinity> It's also the episode that seems to get many people hooked. :)
[02:10] <jbailey> infinity: My wife keeps hoping I'll get hooked.
[02:11] <Keybuk> Fame - Hard Work
[02:11] <jbailey> infinity: I keep pointing out that they can't act. =)
[02:11] <infinity> jbailey : That's not important.
[02:11] <Keybuk> The Producers - Spring Time for Hitler
[02:11] <jbailey> infinity: Funny, that's her answer. =)
[02:11] <infinity> jbailey : Given that the entire series is a self-referrential spoof, it doesn't need to have acting.
[02:11] <Keybuk> Little Shop Of Horrors - Grow For Me
[02:11] <Keybuk> Chess - One Night In Bangkok
[02:11] <Keybuk> I think that's the list so far :p
[02:12] <infinity> jbailey : In fact, it gets worse (well, sillier) as time goes on, but that just seems to make it better.
[02:12] <jbailey> infinity: Right, and I'm actually starting to enjoy it.
[02:12] <infinity> Sadly, I can't get Zofia hooked.  She's resistant.
[02:12] <Keybuk> Joss Whedon is a modern-day god
[02:12] <infinity> I took a sneaky approach and got her hooked on Firefly instead.
[02:12] <Keybuk> accept it, sit back, and enjoy the fun
[02:12] <jbailey> I just saw the last half of the to be cont'd one in Season 2 where they try to make Spike's girl less insane.
[02:13] <infinity> Then said "hey, this is EXACTLY the same as Buffy.. Same jokes, same everything... You shoud really try Buffy"
[02:13] <infinity> So far, she hasn't bitten.
[02:13] <jbailey> I've been wondering if those were the first SM scenes on primetime television.
[02:14] <infinity> Oh well, I guess "I got hooked on Buffy 'cause I watched it with the teenage daughter of my ex-girlfriend" is never a good way to get one's current girlfriend interested.
[02:15] <jbailey> I think it depends on how long ago the ex was.
[02:15] <jbailey> Since it's been 11 years since mine, Angie just isn't threatened anymore. =)
[02:16] <infinity> This one's long enough ago, but people can be.. Twitchy.
[02:16] <infinity> I'll wait until we're a good 10 years in before I say "Hey, you do that swirly tihng with your tongue JUST LIKE ANNA USED TO."
[02:16] <jbailey> Right.  It's clearly been too long since I've seen you.
[02:16] <infinity> And with that, I realise I shouldn't be drinking and typing, and I should perhaps just go to bed.
[02:17] <jbailey> I keep forgetting that your online tact matches your in person tact ;)
[02:17] <infinity> Tact?
[02:17] <infinity> What's.... Tact?
[02:18] <infinity> That's kinda the opposite of "saying what I mean, when I mean it, especially when I think you need to hear it", right?
[02:19] <jbailey> The first one is important.  The second one is harder.  The third one is usually when you need a walk around the block first.
[02:26] <Keybuk> heh, I got accused of not having any tact over the weekend
[02:26] <Mithrandir> Keybuk: any idea how to emulate what run-init is doing, using a shellscript?
[02:26] <Mithrandir> I can
[02:26] <Keybuk> we were in a club in manchester and had taken some whingy young queens with us, and one of them was moaning why he hadn't received any "shag tag" messages yet?  And I replied immediately with just "Because you're ugly?"
[02:26] <Mithrandir> argh
[02:26] <zul> hey
[02:26] <Keybuk> Mithrandir: no, never looked what that does
[02:26] <Mithrandir> Keybuk: hmkay.
[03:36] <jbailey> infinity: Next initramfs-tools upload, please consider sucking in all of Max's manpages.
[03:36] <jbailey> infinity: It seems to be currently most of the diff.
[03:37] <infinity> Yup.
[03:37] <infinity> I'm going to pull in a fair chunk of his diff on my next upload.
[03:37] <infinity> And do the DEBUG thing.
[03:37] <infinity> And discover the 36 hour day, too.
[03:37] <infinity> YAY.
[03:40] <jbailey> infinity: Consider polyphasic sleeping.
[03:40] <jbailey> infinity: And then do something else with the spare time other than work. =)
[03:40] <infinity> "spare time"?
[03:40] <infinity> I have about as much of that as I have tact.
[03:40] <infinity> The two may relate.
[03:41] <jbailey> True. =)
[03:41] <jbailey> squicky dreams, infinity.
[03:42] <infinity> May all your felching involve bendy straws.
[04:13] <makx> infinity cool
[04:15] <makx> big noise in our diff is the quoting of variables done by jonas
[04:15] <makx> i've not been consequent on this since 0.32
[04:39] <Mithrandir> infinity: also, please remove the "exit 1" when detecting circular dependencies with a panic.
[04:40] <Mithrandir> Keybuk: why does init-premount/udev source initramfs.conf?
[04:40] <Mithrandir> Keybuk: it means overriding boot= on the boot loader command line won't work correctly.
[04:41] <Mithrandir> Keybuk: any chance of having /usr/bin/udevinfo and /lib/udev/cdrom_id in the initramfs?  (Needed for casper)
[04:42] <Mithrandir> Keybuk: also, making boot=casper an alternative for boot=local in ./scripts/init-premount/udev would make me happy.
[05:58] <Mithrandir> hmm, so, no response.  Would anybody be inconvenienced if I uploaded an udev with those changes?
[06:58] <Keybuk> Mithrandir: because init doesn't export BOOT
[06:59] <Keybuk> if casper needs things in the initramfs, it should copy them itself with its own hook
[07:00] <Mithrandir> Keybuk: then it should use the same logic as init, not just source a file.
[07:00] <Mithrandir> uhm, even files which aren't part of the casper package?  That'll break if something else copies them too.
[07:00] <Keybuk> actually, init should export that variable :)
[07:00] <Keybuk> Mithrandir: Depends
[07:00] <Keybuk> it won't break, the copy of the file will be the same
[07:00] <Mithrandir> ok, then it's fine.
[07:01] <Keybuk> I'll fix udev and initramfs-tools so that variable is exported rather than sourced
[07:01] <Keybuk> I think I only went with the source to avoid changing two packages at once
[07:01] <Keybuk> and didn't notice the boot= bit
[07:01] <Mithrandir> ook
[07:01] <Mithrandir> could you make boot=casper as an alternative for boot=local in udev as well?
[07:02] <Keybuk> sure
[07:02] <Mithrandir> thanks.
[07:02] <Keybuk> so, here's the $10,000,000 question
[07:02] <Keybuk> the bits you're using udev helpers for
[07:02] <Keybuk> do those really belong in the udev initramfs script?
[07:02] <Keybuk> ie. are you just getting a root filesystem mounted?
[07:03] <Mithrandir> yes, I am, but I don't think it's a usecase that udev itself should necessarily support.
[07:03] <Mithrandir> it's mount a cdrom and mount an cloop image off there.
[07:03] <Mithrandir> see http://people.ubuntu.com/~tfheen/bzr/casper/simplifiedlivecd/
[07:03] <Keybuk> well, if udev has to know about "casper", it should probably know about all things
[07:04] <Keybuk> hmm
[07:04] <Mithrandir> specifically, http://people.ubuntu.com/~tfheen/bzr/casper/simplifiedlivecd/debian/casper.initramfs
[07:04] <Keybuk> don't use "modprobe -q cloop", use "modprobe -Qb cloop"
[07:05] <Mithrandir> I don't think that makes sense to make part of udev, even if written a bit more generaliised.
[07:05] <Keybuk> don't use the while/sleep combo, use "udevplug -W"
[07:05] <Mithrandir> I just copied this off the previous casper stuff, but will do
[07:05] <Keybuk> same for the unionfs modprobe
[07:05] <Keybuk> ok, I take your point ... everything in this script is basically "mounting" the filesystem
[07:06] <Keybuk> I'll add casper to the udev script as a copy of local
[07:06] <Mithrandir> it's similar to scripts/nfs and scripts/local
[07:06] <Keybuk> *nods*
[07:06] <Mithrandir> so since udev knows about those, I think it should know about casper as well.
[07:06] <Keybuk> right
[07:06] <Keybuk> agree
[07:06] <Mithrandir> coolie
[07:07] <Keybuk> I like your CD-ROM hunt
[07:07] <Keybuk> Kamion: we should use something like that in d-i rather than the devfs-reliance
[07:07] <Mithrandir> I copied it off code you sketched out for me. :-P
[07:07] <Keybuk> oh, right, yeah :P
[07:08] <Mithrandir> isn't it just modprobe -Q rather than -qQ?
[07:08] <Keybuk> -Qb
[07:08] <Keybuk> -Q is -q plus even more quietness
[07:08] <Keybuk> -b allows the user to blacklist whatever follows
[07:08] <Mithrandir> ah
[07:08] <Keybuk> (ie. it allows it to be utterly consistent)
[07:08] <Mithrandir> that'll blow up, though.
[07:08] <Mithrandir> but sure.
[07:08] <Keybuk> so?  we assume if they do blacklist something, they had a bloody good reason
[07:09] <Keybuk> it's good to be consistent about it
[07:09] <Mithrandir> can I assume that /dev/cloop0 will always be the name of cloop0, or do I have to query udev/sysfs for it?
[07:12] <Mithrandir> reload/bzr pull
[07:12] <Keybuk> you should probably query it, I guess
[07:12] <Keybuk> the user could commit fantastic pedicide and rename them
[07:12] <Keybuk> or I could
[07:12] <BenC> Keybuk: hey, question for you in -devel
[07:12] <Keybuk> BenC: not that I know of, why?
[07:13] <Keybuk> ide-generic, bad
[07:13] <Keybuk> ide-core, probably ok
[07:13] <Mithrandir> keybuk: in which case I'd hop on the first plane to .uk and beat you with a stick. :-P
[07:13] <Keybuk> indeed
[07:13] <BenC> mainly because with ide.ko modular, none of the ide= hd*= kernel command line options are honored
[07:13] <BenC> since it expects them as module options
[07:14] <Keybuk> ah, heh
[07:14] <Keybuk> I don't know of any problems with making that built-in
[07:14] <Keybuk> I wish we built-in more stuff *cough* unix.ko
[07:14] <BenC> why isn't unix.ko built-in anyway?
[07:14] <Mithrandir> I think it's a legacy of Xu's reluctance to build anything into the kernel
[07:15] <BenC> I've no problem doing it, but I don't know the reason for having it modular in the first place
[07:15] <Mithrandir> he'd probably made ELF into a loadable module, if he could.
[07:15] <BenC> lol, ok, unix and ide will be built-in in -8.10
[07:15] <Keybuk> \o/
[07:16] <BenC> modularity can only go but so far
[07:16] <Mithrandir> Keybuk: what's the Right way to find an unused cloop node?
[07:17] <Keybuk> sodomy non sapiens
[07:18] <Keybuk> don't know much about those
[07:18] <Mithrandir> heh
[07:18] <crispin> Keybuk: btw, I only have about 30 mins now, tomorrow and wed I will have longer
[07:18] <Keybuk> losetup or something
[07:18] <Keybuk> crispin: that's ok, am ready for you now
[07:18] <crispin> k, so /dev/hda doesn't exist
[07:18] <Keybuk> crispin: cat /proc/modules ... and list off the first few for me
[07:18] <crispin> ide_generic, ide_cd, cdrom, sata_nv, generic
[07:19] <Keybuk> right
[07:19] <Keybuk> so that's screwyfic
[07:19] <BenC> heh, all three port architectures already had unix.o built-in
[07:19] <BenC> powerpc already had ide-core built-in
[07:19] <BenC> now they all match
[07:19] <Keybuk> crispin: echo $BOOT $ROOT
[07:19] <crispin> /dev/sda1
[07:19] <Keybuk> BOOT="" ?
[07:20] <Keybuk> . conf/initramfs.conf
[07:20] <crispin> yep
[07:20] <Keybuk> echo $BOOT
[07:20] <crispin> local
[07:20] <Keybuk> cat /proc/cmdline
[07:21] <crispin> root=/dev/sda1 ro quiet splash
[07:21] <Kamion> BenC: you may need to tweak the d-i bits a bit when building in ide-core
[07:21] <Kamion> I'll certainly have to tweak debian-installer to compensate
[07:21] <Keybuk> crispin: ls scripts/init-premount
[07:21] <BenC> Kamion: yeah, I'll check that
[07:22] <crispin> udev thermal
[07:22] <BenC> Kamion: for ide-core, just match whatever ppc is doing (it has it built-in already)
[07:22] <Keybuk> crispin: ls scripts/init-top
[07:22] <crispin> usplash
[07:23] <Keybuk> ok ...
[07:23] <Kamion> BenC: right, I can fix the d-i side easily enough, although it's not as simple as matching because the pkg-lists files for the various architectures are often quite divergent
[07:23] <Keybuk> crispin: tail -2 scripts/init-premount/udev
[07:23] <crispin> \t;;\nesac
[07:23] <Mithrandir> Kamion: do you want me to fix the cdrom-detect similarly to what's in new-casper?
[07:24] <Kamion> Mithrandir: sure, if you like
[07:24] <Kamion> I was about to go and look at that
[07:25] <Kamion> ah, yeah, /sys/block querying. if you could keep the current behaviour as an else so that it can be trivially merged, that would be a bonus
[07:25] <Kamion> I made os-prober use /sys/block if it's there recently
[07:25] <Keybuk> crispin: ok ... then I'm bloody confused what's loaded ide-cd, cdrom and ide-generic
[07:25] <Kamion> (though not in preference to /dev/discs; should probably fix that)
[07:26] <Keybuk> crispin: ls /sys/bus/ide
[07:26] <crispin> devices drivers
[07:27] <Keybuk> crispin: ls /sys/bus/ide/devices
[07:27] <crispin> 1.0 1.1
[07:27] <Keybuk> k...
[07:27] <crispin> (btw that list of modules I attached to the bug is what I have loaded here)
[07:27] <Keybuk> could you reboot, and this time remove quiet and splash from the kernel command-line and add "break=premount" to it instead
[07:28] <crispin> yep, done, am at a prompt
[07:28] <Keybuk> ok
[07:28] <Keybuk> scripts/init-premount/thermal
[07:28] <Keybuk> (to stop laptop dying)
[07:29] <crispin> this isn't a laptop ....
[07:29] <Keybuk> could you now check /proc/modules and see what's there (should be not much)
[07:29] <Keybuk> k, do it anyway <g>
[07:30] <Mithrandir> Keybuk: that means we need /lib/udev/cdrom_id in udev-udeb too
[07:30] <crispin> I have 9 modules in there, nothing hardware related, it appears to be just the fb stuff and a couple of others
[07:30] <Keybuk> Mithrandir: k, can you bug me an e-mail
[07:30] <Mithrandir> Keybuk: sure
[07:30] <Keybuk> crispin: ok ... cat /proc/cmdline
[07:31] <crispin> root=/dev/sda1 ro break=premount
[07:31] <Keybuk> right
[07:31] <Keybuk> run udevd --daemon
[07:31] <BenC> Kamion result is that ide-core udeb is going away
[07:31] <Keybuk> and for the next command or two, watch for anything being printed on the console
[07:31] <Kamion> BenC: right, hence necessary adjustment in d-i
[07:31] <Keybuk> if it talks about ide, it's bad
[07:31] <crispin> ok, udevd run, nothing on console
[07:32] <Keybuk> ls /sys/bus/ide/devices /sys/bus/scsi/devices /sys/bus/i2o/devices /sys/bus/usb/devices /sys/bus/ieee1394/devices
[07:33] <crispin> I get 'no such file or directory' for all of them, /sys is mounted
[07:33] <Keybuk> good
[07:33] <Keybuk> udevplug -Cmem -Cmisc -Ctty -Cvc
[07:34] <crispin> run, nothing on console
[07:34] <Keybuk> k, ls /sys/block
[07:34] <Keybuk> is there anything that's not ram* ?
[07:34] <crispin> no, just ram0 -> ram15
[07:35] <Keybuk> ok
[07:35] <Keybuk> grep 0x01 /sys/bus/pci/devices/*/class
[07:36] <crispin> returned 2 files, 0000:00:09:0 -> 0x01018a and 0000:01:09.0 -> 0x018000
[07:36] <Keybuk> ok....
[07:36] <Keybuk> cat /sys/bus/pci/devices/0000:00:09:0/modalias
[07:36] <Keybuk> cat /sys/bus/pci/devices/0000:01:09:0/modalias
[07:36] <Keybuk> (both)
[07:37] <crispin> urgh, do you need to the full output from these files ?
[07:37] <Keybuk> yes
[07:37] <Keybuk> exactly
[07:37] <crispin> :-)
[07:38] <crispin> the first is: pci:v000010DEd00000065sv0000147Bsd00001C01bc01sc01i8a
[07:39] <crispin> second is: pci:v00001095d00003112sv00001095sd00003112bc01sc80i00
[07:40] <Keybuk> ok, that's kinda interesting
[07:41] <Keybuk> the second one is a Sil 3112 SATA controller?
[07:41] <crispin> sounds about right
[07:41] <Keybuk> the first one is an nVidia NFORCE IDE controller?
[07:41] <crispin> I have a pci sata controller on an nforce m'board
[07:42] <Keybuk> right....
[07:42] <Keybuk> could you: echo blacklist ide-generic > /etc/modprobe.d/test
[07:42] <crispin> the disk hangs off the sata card
[07:42] <crispin> yep
[07:42] <Keybuk> then udevplug -Bpci -Iclass=0x01*
[07:43] <crispin> ok, lots of output on the console this time
[07:43] <Keybuk> k, once done, could you cat /proc/modules
[07:44] <crispin> new stuff in there is scsi_mod, libata, sata_sil, ide_core, amd74xxx, generic, sata_nv, cdrom, ide_cd
[07:44] <Keybuk> ok
[07:44] <Keybuk> you have an IDE CD-ROM drive hanging off your motherboard
[07:44] <Keybuk> but the drives are hanging off your SATA PCI controller?
[07:44] <crispin> yep, 2 of them
[07:45] <crispin> exactly
[07:45] <crispin> (well, one drive off the sata pci controller)
[07:45] <crispin> interestingly, in the console output was:
[07:45] <crispin> ata1: SATA link up 1.5 Gps (SStatus 113)
[07:46] <crispin> ata1: dev 0 not supported, ignoring
[07:46] <Keybuk> right
[07:46] <Keybuk> could you check for me:
[07:46] <Keybuk> ls /dev/hd*
[07:46] <Keybuk> ls /dev/s*
[07:46] <crispin> /dev/hdd /dev/hdc, no s* stuff
[07:47] <Keybuk> k, so to recap
[07:47] <Keybuk> you have a computer with an nVidia NFORCE IDE controller on the motherboard
[07:47] <Keybuk> and an Sis 3112 SATA controller PCI card
[07:47] <Keybuk> there's a CD-ROM hanging off the IDE controller
[07:47] <Keybuk> and the hard-drive you want to boot off hanging off the SATA controller
[07:48] <crispin> yes, (except I have 2 cd rom drives hanging off the m'board ide controller)
[07:48] <Keybuk> both controllers appear find under /sys with the right MODALIAS
[07:48] <Keybuk> and udev loads the right modules for them both
[07:48] <Keybuk> (I think)
[07:48] <BenC> is nforce supported in kernel now, or is it just ide-generic?
[07:48] <BenC> or are you using nforce drivers from nvidia?
[07:49] <crispin> the sata_nv seems an odd thing to load, the m'board isn't sata
[07:49] <Keybuk> well, he has a sata_nv module -- I assume that's the real thing?
[07:49] <Keybuk> what's amd74xxx ?
[07:50] <BenC> so what's the issue?
[07:50] <Keybuk> hmm
[07:50] <Keybuk> BenC: his hard-drive doesn't appear
[07:50] <crispin> I've no idea, I'm just using the stock ubuntu kernels, nothing changed in them at all
[07:50] <Keybuk> no uevent fromt he kernel for it, looks like the kernel never detects it
[07:50] <Keybuk> (on the sata_sil bus)
[07:50] <Keybuk> crispin: hmm, we don't ship either sata_nv or amd74xxx
[07:50] <Keybuk> are you SURE you've not got some nvidia binary drivers in there?
[07:50] <BenC> we do ship sata_nv
[07:51] <BenC> is this breezy or dapper?
[07:51] <crispin> I do that the restricted modules package installed
[07:51] <Keybuk> oh, sorry, I can see it now
[07:51] <crispin> BenC: daper
[07:51] <BenC> 2.6.15 kernel?
[07:51] <crispin> yep
[07:51] <BenC> sata_nv is in there I know
[07:51] <crispin> BenC: btw this is bug 20807, it has my lspci and other stuff in there
[07:51] <BenC> amd74xx is too
[07:51] <Keybuk> modinfo on sata_nv has
[07:52] <Keybuk> alias: pci:v000010DEd*sv*sd*bc01sc01i*
[07:52] <Keybuk> so it will be loaded for all nVidia storage controllers
[07:52] <BenC> and sata_nv should support ATAPI, so it's doing the right thing to support the cdrom's
[07:52] <BenC> ok, so sata_sil is the problem?
[07:52] <Keybuk> yeah, it looks like it's doing the right things
[07:53] <Keybuk> either sata_sil is the problem, or the combination of both is an unhappy one
[07:53] <crispin> hmm, guys, I need to leave now, I'll be back for longer tomorrow
[07:53] <Keybuk> ok, I'm around most of the day tomorrow
[07:53] <BenC> it sees the disk
[07:53] <BenC> ok
[07:53] <crispin> well, it will be roughly the same time, I am at work till about 6pm GMT
[07:53] <BenC> dmesg shows the disk Keybuk
[07:54] <BenC> sda 160Gigs
[07:54] <BenC> [4294678.458000]  Attached scsi disk sda at scsi0, channel 0, id 0, lun 0
[07:54] <Keybuk> BenC: where did you get the dmesg?
[07:54] <BenC> wait, that's breezy dmesg :/
[07:54] <Keybuk> yeah
[07:54] <Keybuk> breezy!
[07:54] <Keybuk> crispin ata1: SATA link up 1.5 Gps (SStatus 113)
[07:54] <Keybuk> crispin ata1: dev 0 not supported, ignoring
[07:55] <Keybuk> didn't know whether that could be relevant or not
[07:55] <BenC> sounds like it
[07:57] <Keybuk> tomorrow I'll get him to try only loading one of the two drivers, and see what happens then
[07:57] <Keybuk> that'd at least rule out interaction
[07:57] <Keybuk> could be simply that the sata_sil is busted on current kernel
[07:59] <BenC> the only case where the dmesg would look like this is if dev->id is not an ata device
[08:00] <BenC> which is a bitmask check
[08:00] <BenC> this is all ata core stuff
[08:01] <BenC> the message isn't from sata_sil, but I guess it's possible that it is not giving libata the right info
[08:02] <Keybuk> right
[08:02] <Keybuk> interesting
[08:06] <Keybuk> could be, yeah
[08:06] <Keybuk> or that sata_nv has stamped all over it
[08:20] <BenC> [libata sata_sil]  fix merge related issues
[08:21] <BenC> wow, this is a complicated merge
[08:21] <BenC> for git, not for me
[08:22] <Keybuk> hehe
[08:23] <Mithrandir> generating the bootcharts takes forever on the live cd..
[08:41] <BenC> the ata updates will be in -8.10
[08:41] <BenC> Keybuk: I replied to his bug report to test -8.10 when it's built
[08:42] <Keybuk> k