[12:53] <pitti> Good night everybody
[01:35] <daniels> mdz: not drink six beers at once? BRILLIANT!
[01:47] <sivang> how do I install debugging symbols for gnome 
[01:47] <sivang> ?
[03:22] <chrisa> Any netinst isos for the rc?
[03:37] <mjg59> Why the christ have we picked up Dick Morrel?
[03:38] <daniels> dude, be thankful.
[03:38] <jdub> yeah, ugh
[03:38] <daniels> try on the following names: dan jacobson, john hendrickson
[03:39] <mjg59> He's anti-GPL and he's a cock
[03:39] <mjg59> Dan Jacobson is mad but harmless in comparison
[03:40] <daniels> morrell is an idiot, yes
[03:41] <mjg59> http://groups.google.com/groups?hl=en&lr=&c2coff=1&safe=off&threadm=slrn9nkfpp.g4a.mjg59%40vavatch.jesus.cam.ac.uk&rnum=1&prev=/groups%3Fq%3Dmorrell%2Bgarrett%2Blinux%26hl%3Den%26lr%3D%26c2coff%3D1%26safe%3Doff%26selm%3Dslrn9nkfpp.g4a.mjg59%2540vavatch.jesus.cam.ac.uk%26rnum%3D1 - that was it
[03:41] <mjg59> Fucking URLs. Hrngh.
[03:47] <mjg59> Maybe I should IRC less when drunk.
[03:49] <daniels> (fwiw, he took offence when I asked him to remain on-topic, and spent that 14 minutes ranting about who he was and why I couldn't afford to piss him off.  i just asked him to remain on-topic a couple of times, and directed him towards #offtopic.)
[04:01] <jdub> Zindar: ah, erik :)
[04:01] <jdub> ww
[04:03] <mjg59> daniels: To be honest, I'm inclined to think life would be easier without him
[04:06] <mjg59> dickmorrel is not signal
[04:06] <mjg59> BE SIGNAL
[04:06] <jdub> BE THE SIGNAL!
[04:07] <mjg59> I feel this is important.
[04:11] <daniels> http://www.theage.com.au/articles/2004/10/18/1097951587553.html
[04:12] <hornbeck> man, seven cars
[04:29] <daniels> elmo: can you please let me know how many packages build-depend on: xlibs, xlibs-dev, xlibs-static-pic
[05:02] <fabbione> morning guys
[05:03] <fabbione> daniels: why do we need to know? ;)
[05:03] <fabbione> daniels: in one way or another even the kernel build-dep on them :P
[05:09] <daniels> heh
[05:10] <daniels> fabbione: well, when we get the xorg stuff in, i'd like to start busting up build-deps
[05:10] <daniels> i'd be happy for xlibs/xlibs-dev to not exist
[05:11] <fabbione> daniels: i think we can create a meta package for it
[05:11] <fabbione> that was my idea at least
[05:11] <fabbione> for the beginning it would be acceptable
[05:12] <fabbione> and give people a few weeks to do the transition
[05:12] <fabbione> i want to kill Xfree86 in one shot
[05:12] <fabbione> no double packages 
[05:13] <daniels> yeah, dude, I already have a set of metapackages :)
[05:14] <fabbione> no you have nothing :P
[05:14] <daniels> i bootstrapped xlibs (with full debian packaging) on to my laptop yesterday
[05:14] <daniels> ii  libx11-6                       6.2.1+cvs.200408040713-0ubuntu X Window System protocol client library
[05:14] <daniels> only took about two hours -- some stuff is still a little broken due to the fact /usr/lib/X11 is no longer a symlink
[05:14] <daniels> but yeah, it's all good otherwise
[05:15] <fabbione> eheh
[05:17] <fabbione> ok new X seems to be all good
[05:18] <fabbione> daniels: but did you bootstrap each lib onm theirown?
[05:18] <fabbione> or just xc/lib in one shot?
[05:18] <daniels> fabbione: each on their own
[05:19] <daniels> fabbione: started with x-common, then just ran check/build/install through all the packages
[05:19] <fabbione> similar to the approach i used...
[05:19] <fabbione> sounds like at least
[05:20] <daniels> anyway, end result is that everything's all good, except for xterm
[05:20] <daniels> it's probably the app-defaults link that's broken
[05:21] <fabbione> daniels: i started drafting a plan for the 2 weeks
[05:21] <fabbione> we have 2 fronts that we need to fight
[05:21] <fabbione> one is the xresprobe/autoconfig stuff
[05:21] <jdub> 1. Who will order the pizza?
[05:21] <daniels> ME!
[05:21] <jdub> 2. Will Daniel be allowed to sleep inside the house?
[05:21] <fabbione> one is beating upstream with a cluebat and kill it into the many deb packages
[05:21] <daniels> jdub: (hotel)
[05:22] <daniels> fabbione: dude, you missed the confcall on Friday
[05:22] <daniels> fabbione: despite almost sending me to sleep, I got useful stuff done on it
[05:22] <fabbione> jdub: no my house isn't a house yet... it's a building site
[05:22] <daniels> fabbione: i got 'big big vendor merge' on the agenda for x11r7
[05:22] <fabbione> daniels: i don't care to talk with X.org "prime donne"
[05:23] <daniels> heh, most of them aren't prima donnas
[05:23] <fabbione> daniels: (first women)
[05:23] <daniels> they just like to argue semantics for a long time
[05:23] <daniels> but that aside, it was very useful, and yeah, so if we start collecting all our patches, making notes on them and stuff, we can empty our debian/patches into upstream
[05:24] <jdub> pouring patches on the fire :)
[05:24] <daniels> anyway, I have to run out the door now
[05:24] <daniels> actually, ten minutes ago
[05:24] <daniels> but nevermind
[05:24] <daniels> i'll talk to you about this later
[05:24] <daniels> unless you really want to call me ;)
[05:25] <fabbione> daniels: no no.. i will have to suffer your ugly face for 2 weeks here :P
[05:25] <fabbione> i don't really need to talk to you too ;)
[05:33] <fabbione> mdz, jdub: i am ready to upload X
[06:01] <fabbione> and linux-restricted modules to fix the nvidia-glx error
[07:08] <fabbione> oooook
[07:08] <fabbione> X and linux-restricted modules are up
[07:55] <daniels> fabbione: heh!
[07:55] <fabbione> daniels: ?
[07:56] <daniels> fabbione: your talking-to-me remark
[07:56] <daniels> fabbione: just had to go into the travel agent here and finalise my tickets
[07:56] <daniels> about to head back home
[07:56] <fabbione> ahhh
[07:56] <fabbione> eheh
[10:17] <pitti> Good morning everybody! So quiet today here...
[10:25] <pitti> Keybuk: I have a small packaging problem: hal 0.2.98 shipped a conffile, but hal 0.4 does not ship it any more; unfortunately dpkg does not delete the conffile on package upgrade, even if it is unmodified
[10:25] <pitti> Keybuk: is there a recommended way to remove it in the maintainer scripts?
[10:26] <Keybuk> dpkg should delete the conf file ?
[10:26] <Keybuk> unless it was never registered with dpkg
[10:26] <pitti> Why not, if it's unmodified?
[10:26] <pitti> no, the old deb ships it and it was in debian/conffiles
[10:27] <pitti> Since I don't want to change dpkg, is there a "best practice" how to handle this?
[10:27] <pitti> Keybuk: I thought about providing a md5sum list and delete the file in the preinst if one md5sum matches
[10:27] <pitti> Keybuk: similar to ucf
[10:27] <Keybuk> I mean that dpkg *should* delete the conffile, so why isn't it?
[10:27] <pitti> oh, I don't know
[10:28] <pitti> I tried that out yesterday, with a modified conffile and an unmodified
[10:28] <Keybuk> can you throw me both packages
[10:28] <pitti> you already have hal 0.2.98
[10:28] <Keybuk> yeah, but I need to uninstall/reinstall a bit so it's good to have the .deb standing by without having to hunt for it :p
[10:28] <pitti> I send you a link to hal-0.4 when it's ready
[10:29] <pitti> okay, I put the old deb to the same place
[10:40] <pitti> Keybuk: the stuff is at http://www.piware.de/hal/
[10:40] <pitti> Keybuk: for upgrading to hal 0.4.0 you need libhal-storage0 deb
[10:41] <Keybuk> right
[10:41] <pitti> Keybuk: the conffile in question is /etc/hal/device.d/fstab-update.hal
[10:41] <pitti> Keybuk: the old hal installs it as example file, but I want to get rid of it (we have pmount)
[10:41] <Keybuk> ahh, lftp, how we love thee ... mget *.deb :p
[10:42] <Keybuk> so I install the single 0.2.98 deb, and upgrade by installing the others?
[10:42] <jdub> argh
[10:42] <jdub> forgot how loud my ibook was
[10:42] <Keybuk> jdub: bong!
[10:42] <pitti> Keybuk: yes, but you need only libhal-storage
[10:42] <pitti> Keybuk: not all the other debs
[10:43] <pitti> Keybuk: the problem is that above conffile is not shipped by hal 0.4.0 any more, so it should disappear
[10:43] <Keybuk> ok, well in 0.2.98 it's in the package and listed as a conffile
[10:43] <pitti> and I guess you did not modify it
[10:44] <Keybuk> just waiting for the ol' chroot to finish updating
[10:46] <pitti> Hi carlos
[10:46] <pitti> I finally managed to take a look at your application yesterday evening :-)
[10:46] <carlos> pitti: hi
[10:47] <carlos> pitti: I saw it :-), btw, the TODO was not a mark that says.. From here is not done it was just for that question
[10:47] <pitti> carlos: I know
[10:47] <pitti> carlos: but the other stuff was okay
[10:47] <carlos> ok
[10:48] <carlos> I have the other practical questions, I will try to send you all remaining things soon
[10:48] <carlos> pitti: thanks
[11:10] <pitti> Hi mvo_!
[11:11] <mvo_> hi pitti 
[11:11] <mvo_> hi to all others
[11:12] <seb128> hello mvo_ 
[11:13] <mvo_> hi seb128 
[11:22] <Keybuk> seb128: is it me, or is gnome-terminal 2.8 a little late? :p
[11:23] <seb128> ah ah
[11:23] <seb128> just a bit late :)
[11:23] <Keybuk> and 2.8.1 just after it
[11:23] <Keybuk> amusing
[11:24] <Keybuk> pitti: (best Mal voice) well now, ain't that just an oddness
[11:31] <Keybuk> D000002: fork/exec /var/lib/dpkg/info/hal.postinst (dpkg: error processing hal (--install):
[11:31] <Keybuk>  subprocess post-installation script killed by signal (Segmentation fault), core dumped
[11:31] <Keybuk> Errors were encountered while processing:
[11:31] <Keybuk>  hal
[11:34] <pitti> Keybuk: just back from lunch - argh, what's that?
[11:34] <Keybuk> oh, just another wonderful example of dpkg's amazing bug-free-ness
[11:35] <pitti> Keybuk: odd, it works fine here; but this can hardly be a bug in the hal deb, can't it?
[11:35] <Keybuk> pitti: try installing with: dpkg -D7777 -i ... :)
[11:36] <pitti> D000020: deferred_configure `/etc/dbus-1/system.d/hal.conf' (= `/etc/dbus-1/system.d/hal.conf') useredited=-1 distedited=-1 what=2
[11:36] <pitti> D000002: fork/exec /var/lib/dpkg/info/hal.postinst ( )
[11:36] <pitti>  * Restarting system message bus...
[11:36] <pitti>  * Stopping Hardware abstraction layer...                                                                                                                                     [ ok ] 
[11:36] <pitti>  * Starting Hardware abstraction layer...                                
[11:37] <pitti> works for me
[11:37] <pitti> hmm
[11:38] <mvo_> hi rburton 
[11:38] <rburton> hi mvo_
[11:38] <rburton> so i hear xfs in recent 2.6 kernels is broken
[11:39] <Keybuk> weird
[11:39] <Keybuk> rburton: works ok for me
[11:39] <rburton> i'll probably ubuntufy my home desktop this week
[11:39] <rburton> see if it passes the GirlFriend test
[11:40] <pitti> rburton: you mean the new images?
[11:40] <pitti> rburton: failed miserably for me
[11:41] <rburton> i mean i heard reports of data corruption
[11:41] <pitti> rburton: my gf: "Ugh, what's that?" and after seeing the splash "hey, this gets worse..."
[11:41] <pitti> ah
[11:41] <rburton> oh i see
[11:41] <rburton> i told her about the images, she said she so
[11:41] <rburton> doh
[11:41] <pitti> is your gf able to do data corruption?
[11:41] <rburton> well, now you mention that
[11:41] <rburton> things do just break around her
[11:42] <pitti> ah, so she is the ideal beta tester?
[11:43] <pitti> I've got such a friend as well. Up to now he managed to break every Linux distro I gave him; he is totally frustrated by "this linux shit"
[11:45] <Keybuk> pitti: I suspect this is just another example of dpkg going strange when debug is on
[11:45] <Keybuk> lots of little memory leaks, and double-frees, and stuff
[11:45] <pitti> Keybuk: can you at least reproduce the problem without debugging?
[11:46] <pitti> Keybuk: i. e. the conffile is still present?
[11:46] <Keybuk> yup, got an update log now
[11:58] <jdub> rburton: your GF doesn't like it? what does vicky think?
[11:59] <rburton> jdub: gf/wife/whatever ;)
[11:59] <jdub> no way dude, i heard that slip!
[12:00] <rburton> must... destory... evidence
[12:05] <pitti> rburton: don't forget to send them to the host which stores the irc logs :-)
[12:08] <Keybuk> D000200: oldconffsetflags `/etc/hal/device.d/fstab-update.hal' namenode 0x827d97c flags 4
[12:08] <Keybuk> bleh, well it knows it's only in the old package and now in the new
[12:20] <sabdfl> mdz: around?
[12:21] <daniels> sabdfl: 'morning
[12:21] <sabdfl> hey daniels
[12:21] <daniels> rburton: if you don't have it already, you must get dj platurn -- 'so this is de la heaven'. mix cd if just de la songs and samples.  sensational.
[12:22] <daniels> sabdfl: so, community council is in ... three and a half hours, yeah?
[12:22] <daniels> sabdfl: (missed one meeting because someone said UTC when they said Europe/London)
[12:22] <sabdfl> community council is tuesdays, 1600 UTC
[12:23] <daniels> well, the artwork meeting
[12:23] <sabdfl> this next one is a general community meeting, in 90 minutes
[12:23] <Keybuk> though this tuesday is tech-board
[12:23] <sabdfl> yes
[12:23] <sabdfl> anybody have an ipw2100 handy for testing?
[12:23] <rburton> can anyone sit in on the meetings?
[12:24] <rburton> daniels: oh, cool
[12:24] <Keybuk> reminds me, must beat up Jeff to make an .ics of HoaryReleaseSchedule when it's final
[12:24] <Keybuk> rburton: sure.
[12:24] <daniels> sabdfl: ah right
[12:24] <daniels> ... ninety minutes? fair cop
[12:25] <fabbione> daniels: debian/patches/004_*
[12:25] <fabbione> daniels: how much do we need out of it?
[12:26] <daniels> fabbione: last I checked, all of it
[12:26] <fabbione> -          MANDEFS = AppLoadDefs FileManDefs LibManDefs MiscManDefs DriverManDefs ProjectManDefs $(XORGMANDEFS) $(VENDORMANDEFS)
[12:26] <fabbione> +          MANDEFS = AppLoadDefs ManDefs SyscallManDefs LibManDefs DriverManDefs FileManDefs GameManDefs MiscManDefs AdmManDefs ProjectManDefs $(XOR
[12:26] <fabbione> GMANDEFS) $(VENDORMANDEFS)
[12:26] <fabbione> this is the only line that doesn't really "merge"
[12:26] <fabbione> should i just pristine copy it from xfree86?
[12:27] <daniels> what's -, what's +?
[12:28] <fabbione> daniels: that's what it is in 004_
[12:28] <fabbione> xorg has a different default MANDEFS 
[12:28] <daniels> xfree86, or xorg?
[12:28] <daniels> ah ok
[12:29] <fabbione> 3 way diff :-)
[12:33] <fabbione> daniels: i am going to grab some food
[12:39] <daniels> ok
[12:39] <daniels> when you get back, let me know what we're starting with in xorg and where we need to go
[12:43] <thom> daniels: acpi-support doesn't do suspend/resume, so...
[12:44] <thom> oh, and acpi-support-x40 needs to rmmod ipw2?00 before suspend
[12:45] <daniels> thom: oh, right.  point.
[12:45] <daniels> thom: and, uhm, that bong was apparently fixed in 2.6.8
[12:45] <daniels> (centrino wireless)
[12:48] <thom> daniels: well, i have to have it with the 2.6.8.1-3-15 kernel
[12:48] <thom> otherwise i need to powercycle my machine everytime it comes out of suspend
[12:49] <thom> and 1940 is pbbuttonsd foo
[12:49] <thom> not acpi-support
[12:52] <seb128_> hello thom 
[12:53] <thom> heyhey seb
[12:55] <plovs_work> I am writing a kernel-howto for the wiki, is grub default or should i mention lilo as well?
[12:56] <thom> grub default
[12:56] <plovs_work> thom, ok
[12:57] <daniels> thom: bongtastic.  you should've got an nm.
[01:02] <sabdfl> pitti: pen drive is still failing to automount :-/
[01:02] <pitti> sabdfl: but it worked for the version you downloaded manually? odd, this is the very version I uploaded recently
[01:02] <sabdfl> strange, but true
[01:03] <sabdfl> hoary problem :-)
[01:03] <pitti> sabdfl: does it work sometimes and sometimes not?
[01:04] <sabdfl> not yet
[01:04] <sabdfl> usb 4-4: new high speed USB device using address 4
[01:04] <sabdfl> scsi2 : SCSI emulation for USB Mass Storage devices
[01:04] <sabdfl>   Vendor:           Model: Pen Drive 2.0     Rev: 1.13
[01:04] <sabdfl>   Type:   Direct-Access                      ANSI SCSI revision: 02
[01:04] <sabdfl> SCSI device sda: 253952 512-byte hdwr sectors (130 MB)
[01:04] <sabdfl> sda: Write Protect is off
[01:04] <sabdfl> sda: Mode Sense: 03 00 00 00
[01:04] <sabdfl> sda: assuming drive cache: write through
[01:04] <sabdfl>  /dev/scsi/host2/bus0/target0/lun0: p1
[01:04] <sabdfl> Attached scsi removable disk sda at scsi2, channel 0, id 0, lun 0
[01:04] <sabdfl> USB Mass Storage device found at 4
[01:04] <sabdfl> but it doesn't show up under /media/
[01:04] <pitti> sabdfl: looks good
[01:05] <pitti> sabdfl: can you please do 'lshal > lshal.txt' and send me the result?
[01:05] <sabdfl> pitti: martin.pitt@can...?
[01:06] <pitti> sabdfl: yes
[01:06] <sabdfl> on its way
[01:06] <sabdfl> let me know if it didn't work i was experimenting with piping straight to mutt
[01:08] <pitti> sabdfl: got it
[01:08] <pitti> sabdfl: hmm, /dev/sda* is not contained in the lshal
[01:08] <sabdfl> great, thanks
[01:08] <sabdfl> why not?
[01:08] <pitti> sabdfl: don't know, that's the bug, I suppose
[01:09] <sabdfl> where could i look or check?
[01:09] <pitti> sabdfl: if you have a minute?
[01:09] <sabdfl> sure
[01:09] <pitti> please do sudo killall hald
[01:09] <daniels> sabdfl: you can also pass -a to attach stuff
[01:09] <pitti> sabdfl: then remove the pendrive
[01:09] <daniels> sabdfl: e.g. mutt -s foo -a ~/tmp/bar baz@quux.org
[01:10] <pitti> sabdfl: sudo hald --verbose=yes --daemon=no --drop-privileges
[01:10] <sabdfl> daniels: yes, i was trying to bypass the file and pipe it straight to mutt, thusly:
[01:10] <pitti> sabdfl: if the log flood stops, press enter a few times (to generate some empty lines)
[01:10] <sabdfl> lshal | mutt -s "foo bar" email@sfsdf.com
[01:10] <daniels> sabdfl: ah, cool :)
[01:11] <pitti> sabdfl: then plug in the beast and watch what happens
[01:11] <pitti> sabdfl: I'd be interested in the cut& paste of the debug stuff after the plugin
[01:11] <sabdfl> ALL of it?
[01:11] <sabdfl> and it just popped up.
[01:12] <pitti> sabdfl: what, the device got mounted?
[01:13] <sabdfl> mounted and the window popped up correctly
[01:13] <pitti> sabdfl: re ALL: only the stuff after plugging in the device (that's the reason for the empty lines)
[01:13] <pitti> sabdfl: so, a heisenbug?
[01:13] <sabdfl> maybe i hadn't rebooted since the update?
[01:13] <pitti> sabdfl: might be
[01:13] <pitti> sabdfl: since when your session runs?
[01:13] <sabdfl> let me reboot now and see if it's really working properly
[01:13] <sabdfl> how do i tell?
[01:14] <pitti> uptime shows the time since last reboot
[01:14] <pitti> last|head shows the recent logins
[01:37] <sivang> sabdfl : the data center has PowerEdge machines?
[01:37] <thom> ROCK
[01:37] <thom> sivang: yeah
[01:37] <sabdfl> sivang: i think we have a bit of everything in there now
[01:37] <thom> i have working NetworkManager foo
[01:37] <sabdfl> thom: cool, what was the trick?
[01:38] <thom> apart from the fact they try to use the ESSID as a key and haven't dealt with the fact that essids can have spaces in
[01:38] <thom> sabdfl: writing most of the debian backend over the weekend :-)
[01:39] <sabdfl> most excellent
[01:39] <sivang> thom : what else is there? :) Is it PowerEdge 2800 (glancing on it ad dell's )
[01:39] <thom> which was entertaining since i was at my parents and they don't have wireless
[01:39] <daniels> thom: awesome! nice one
[01:39] <thom> sivang: poweredge, hp dl380s, ibm x345 and bigger
[01:39] <daniels> and g5 xserves
[01:40] <sivang> thom : wow
[01:40] <thom> yeah
[01:40] <thom> and some hp itaniums at some point, i guess
[01:40] <daniels> does level3 know about the itaniums?
[01:40] <sivang> thom : these are also Xeon based? or are they RISCiy ? :)
[01:41] <daniels> they'll need to double their air conditioning capacity
[01:41] <sivang> all of them are buildds ? I reckon the g5 is for building PPC stuff right?
[01:41] <thom> daniels: compared to the xserves they're probably trivial
[01:42] <daniels> thom: really?
[01:42] <thom> sivang: the three g5s, 3 dual opterons, and 6 dual xeons are buildds
[01:42] <thom> three of the itaniums will be too
[01:42] <daniels> thom: i was under the impression they were space heaters, and the desktop g5s i've seen have run quite cool
[01:42] <thom> daniels: the xserves don't have fan/thermal support yet
[01:42] <thom> they run at full whack all the time
[01:43] <daniels> thom: oh dear.
[01:43] <thom> yeah
[01:43] <daniels> does this explain your need to evacuate the house during testing?
[01:43] <thom> yes
[01:43] <daniels> good god.
[01:44] <Kamion> hm, I thought thermal support did exist now
[01:44] <sivang> what's used for web serving and archive ?
[01:44] <thom> Kamion: maybe, but it's certainly not on the buildds yet
[01:44] <elmo> the PEs are 2650's
[01:45] <thom> hey elmo
[01:45] <daniels> elmo: 'morning
[01:46] <elmo> hey thom, daniels
[01:49] <elmo> daniels: Itanium's really aren't that bad
[01:49] <sivang> thom : you were fixing the network-admin crash stuff?
[01:50] <thom> sivang: no
[01:50] <pitti> sivang: does it _still_ crash?
[01:50] <thom> NetworkManager
[01:50] <pitti> sivang: this shouldn't happen any more with the latest gnome-system-tools
[01:50] <daniels> hm, I think we need some g5 xserves at trinity
[01:50] <thom> sivang: http://people.redhat.com/dcbw/NetworkManager/index.html
[01:50] <daniels> the server room (which has quite a few machines) was in a broom closet.  during summer, you'd just pick which services were needed (no labs at lincoln square, so there goes its nfs and pxe servers ...)
[01:51] <sivang> pitti : lemme check that now
[01:51] <Mithrandir> daniels: the opterons doesn't run too hot either.
[01:51] <daniels> yeah
[01:53] <sivang> pitti : shoot me up the bug#
[01:54] <sivang> thom : thanks. So that's the name for the little thingy on the panel..
[01:54] <thom> that and the bitching huge daemon underneath it, too
[01:54] <pitti> sivang: you already helped me to reproduce it, don't remember any more?
[01:54] <sivang> haha
[01:54] <pitti> sivang: #2177
[01:55] <sivang> pitti : ah no :)
[01:55] <daniels> sivang: at that price, I'll take two
[01:55] <pitti> sivang: boot memtest-human
[01:55] <sivang> I'm afraid it's gonna heat up too much :)
[01:56] <sivang> with memtest
[01:56] <pitti> well, you can't do anything for a few hours :-)
[01:57] <sivang> thom : how do I add it to my panel? I have to apt-get install network-manager first?
[01:58] <thom> i'll need to set up the apt repo first
[01:58] <thom> and it's still very flaky right now
[01:58] <sivang> pitti : yes, I won't be able to..unless memtest be rewritten to support time sharing :)
[01:59] <sivang> thom : well, just tell me where to get the .deb file, and no, I don't mind teashing my system in favor of testing some bleeding edge. (I know I might bleed heh)
[02:00] <thom> sivang: when i have it working right, i will
[02:01] <sivang> thom : thanks.
[02:02] <sivang> pitti : when I run network-admin from console, and I have already enabled my root account - trying to delete an interface just make it reappear the next time I fire up network-admin
[02:02] <pitti> sivang: can you please file a bug?
[02:06] <lamont> moo
[02:06] <pitti> lamont: bark
[02:09] <lamont> heh
[02:10] <pitti> lamont: Good morning!
[02:22] <lamont> hrm.. so how many times to we repeat '7' when counting the bugs, I wonder....
[02:22] <mjg59> Latest ibm-acpi lets you make LEDs flash
[02:23] <daniels> the atheros does the LED stuff on its own
[02:23] <daniels> actually, all my LEDs work just fine with the stock kernel
[02:24] <Keybuk> daniels: you can soft-control it if you like
[02:24] <Keybuk> and it has a flash-on-traffic mode as well
[02:24] <mjg59> daniels: It seems like the wireless LED isn't under software control
[02:24] <mjg59> You need to wave lines on the mini-PCI slot
[02:25] <mjg59> But I can make my power light flash
[02:25] <Keybuk> "hardware control" and "atheros" being somewhat a contradiction in terms
[02:28] <daniels> mjg59: oooer
[02:28] <daniels> Keybuk: heh
[02:29] <daniels> Keybuk: yeah, mine's in flash-on-traffic as per default
[02:29] <Keybuk> that just irritated me on the HP, the LED's a bright blue one and it really catches your eye :p
[02:30] <daniels> there's a solution to that ...
[02:32] <Keybuk> tap-tap-tap*OOH TRAFFIC*tap-tap*OOH TRAFFIC* etc.
[02:32] <rburton> haha
[02:32] <rburton> Keybuk: masking tape
[02:33] <rburton> Keybuk: or better yet, someone here has a blue led fetish and will buy it from you if you can get it out
[02:33] <daniels> my green LED is very unintrusive
[02:36] <fabbione> HaveLib64 Yes or No for x86_64 ?
[02:37] <fabbione> Mithrandir: ^^
[02:37] <daniels> no
[02:37] <daniels> because there is on /lib64
[02:37] <daniels> (iirc)
[02:37] <fabbione> can you see comments on debian/patches/600_amd64_support.diff
[02:37] <fabbione> ?
[02:37] <fabbione> (xfree86)
[02:38] <daniels> AMD64 is not a "bi-width" architecture in Debian at present; instead it is
[02:38] <daniels> a pure 64-bit environment, so do not define "HaveLib64" as "YES".
[02:38] <fabbione> daniels: there is also that AMD64/x86_64 renaming that i am not completely sure about
[02:38] <daniels> *just opened it up then, and it seems to agree with what I think)
[02:38] <daniels> yeah, that's pretty bong, but I think upstream's right
[02:38] <fabbione> upstream has AMD64
[02:39] <daniels> bongtastic
[02:39] <daniels> i'll defer to mithrandir on this one
[02:40] <Mithrandir> what does HaveLib64 mean?
[02:40] <fabbione> # if defined (AMD64Architecture) || defined (s390xArchitecture) || defined (Ppc64Architecture)
[02:40] <Mithrandir> we have a /lib64, but it's a symlink to /lib
[02:40] <fabbione> Mithrandir: it means: stick the lib in /usr/lib64 
[02:40] <fabbione> instead of /usr/lib
[02:40] <fabbione> ok
[02:41] <fabbione> so it's NO
[02:41] <fabbione> thanks :-)
[02:41] <Mithrandir> yeah, you should never reference lib64 (except if you're ld)
[02:41] <fabbione> i guess that will be the same for PPC64
[02:41] <fabbione> the only one that requires YES is s390
[02:41] <fabbione> so we are all happy :-)
[02:42] <daniels> huzzah
[02:43] <fabbione> this night i will dream about dbs-edit-patch :-)
[02:43] <fabbione> daniels: not too bad.. already 1500 lines of patches just for xc/config/ 
[02:44] <daniels> mmm
[02:44] <fabbione> we only miss 298500 lines to check :P
[02:45] <daniels> ehm
[02:45] <daniels> all of the stolen from HEAD stuff is slightly lower-priority
[02:47] <Mithrandir> fabbione: so you will love me when I start whining about multiarch?
[02:48] <fabbione> Mithrandir: oh yes.. 
[02:48] <fabbione> i can give you all the a**l love you want :P
[02:49] <Mithrandir> ;)
[02:51] <fabbione> Mithrandir: otherwise you can start whining about it in februrary
[02:51] <fabbione> you will have almost 16 days of X all for yourself :)
[02:54] <mjg59> Mm. Multiarch.
[02:55] <Mithrandir> fabbione: I need to talk to matt about it, I think we want it for ubuntu, but else, it goes in my free time, which there's not too much of atm. :/
[03:25] <sivang> pitti : is it ok for network-admin to ask my _root_ password when executed from terminal as the regular user?
[03:25] <pitti> no
[03:25] <pitti> it should be _your_ password
[03:25] <sivang> pitti : I have enabled the root account
[03:26] <pitti> sivang: hmm, but still it uses gksudo...
[03:27] <sivang> pitti : yes but it asks for my root password for some strange reason.
[03:27] <pitti> what did you enter exactly?
[03:27] <Keybuk> no, that's right -- network-admin asks for *root*'s password
[03:27] <Keybuk> all of g-s-t do
[03:28] <sivang> pitti : Well, I fired up a gnome-termina.
[03:28] <Keybuk> we get around that by putting "gksudo" into the launcher so they think they're being run as root, so don't ask for a password
[03:28] <Keybuk> sivang: sudo network-admin
[03:28] <pitti> Objection, if I start network-admin from the Menu, it asks me for _my_ password
[03:28] <Keybuk> or just use the launcher
[03:28] <sivang> pitti : yes, because it uses gksudo
[03:28] <Keybuk> pitti: yes, because the *launcher* contains "gksudo"
[03:28] <pitti> ah
[03:29] <pitti> but I saw the gksudo integrated into the C source somewhere...
[03:29] <Keybuk> not in g-s-t
[03:29] <pitti> probably
[03:29] <Keybuk> that was gnome-cups-manager, wasn't it?
[03:29] <pitti> I hacked in so many programs, I probably mixed that up
[03:29] <pitti> could very well be
[03:29] <pitti> well, but I think this is not a critical bug
[03:30] <pitti> sivang: still, can you please file it in bz? it should be fixed for Hoary, should be relatively easy
[03:31] <sivang> strange.
[03:32] <sivang> yes I assume so, but I can't reproduce it anymore..
[03:32] <pitti> sivang: I can
[03:33] <sivang> pitti : I mean for the part where it doesn't delete the interface you try to delete.
[03:33] <sivang> pitti : not for the missing gksudo thingy
[03:33] <pitti> ah, that one
[03:35] <pitti> Bug report: n-a does not apply my changes if I click the Cancel button :-P
[03:35] <sivang> :))
[03:37] <sivang> pitti : you have and idea why that code has so _few_ remarks? I mean, HACKING tells you about the connection between the frontend and backend, but that's about it :)
[03:37] <sivang> remarks = comments
[03:38] <pitti> maybe the fundamental programmer paradigma:
[03:38] <pitti> "It was hard to write, it should be hard to read"
[03:39] <Keybuk> actually, bug-report n-a doesn't follow GNOME HIG
[03:39] <Keybuk> thom: hrm?
[03:39] <thom> Keybuk: it stores network details in gconf
[03:39] <thom> with the essid as the key
[03:39] <thom> essids can have spaces
[03:40] <Keybuk> indeed they can
[03:40] <daniels> bongtasmic
[03:40] <thom> gconf keys can't
[03:40] <sivang> thom : I hope your relationship with it won't go where you had already gone with mozilla ;)
[03:40] <Keybuk> they can have any utf-8 or non-NULL $charset character too
[03:40] <mjg59> essids can?
[03:41] <Keybuk> I suspect they can have \0 in them as well, theoretically
[03:41] <daniels> mjg59: yes
[03:41] <mjg59> Are they defined as utf-8, then?
[03:41] <Keybuk> but a lot of code (including iwlib) kinda assumes they don't
[03:41] <thom> mjg59: my essid is "Bitch Whore"
[03:41] <sivang> Keybuk : is there anything they *can't* have?
[03:41] <Keybuk> mjg59: up to 32-octets specified as octets plus length
[03:41] <mjg59> thom: Oh, sure, spaces are good :)
[03:41] <thom> (don't ask why, hysterical raisins)
[03:41] <mjg59> thom: Have a second gconf key that defines the location of any spaces
[03:42] <Keybuk> sivang: well, there seems to be a general assumption that \0 is bad
[03:42] <Keybuk> mjg59: aren't gconf keys ASCII only?
[03:42] <Keybuk> "Cyber Caf" is a valid ESSID
[03:42] <mjg59> Keybuk: Shit, I hope not
[03:42] <mjg59> I'm fairly sure you can store utf-8
[03:43] <Keybuk> mjg59: sure, in the *value* ... but what about the key?
[03:43] <mjg59> Well, you /can/ store utf-8, because dasher does it. Whether or not it's valid is a separate issue, I guess :)
[03:43] <sivang> Keybuk : well, you should allocate space before hand and make sure you know the length I guess instead of realying on someone putting \0 to end a buffer.
[03:43] <mjg59> Oh, I see what you mean
[03:43] <mjg59> Hrm
[03:43] <daniels> thom: heh
[03:43] <Keybuk> sivang: yeah, the actual packet is length + octets
[03:44] <daniels> how about if you had essid0/wep0, essid1/wep1, or whatever, i nseparate keys
[03:44] <sivang> Keybuk : oh, I overlooked that when you mentioned that few lines ago
[03:44] <daniels> (kconfig is worse in this regard: your key can't have = in it)
[03:46] <thom> the gconf manual says: "Characters in a path should be alphanumeric or underscore"
[03:47] <thom> but doesn't appear to define charset
[03:47] <Keybuk> heh, assume ASCII :)
[03:49] <thom> yeah
[03:52] <thom> aaargh, IEEE web pages use blink
[03:52] <thom> or a blinking jpeg
[03:52] <thom> *yar*
[03:53] <daniels> thom: !!
[04:17] <amu> moind
[04:20] <mjg59> Much development in acpi-land this week
[04:20] <mjg59> If I produce kernels this evening, can people test them?
[04:28] <thom> mjg59: yep
[04:29] <thom> mjg59: include inotify as well, for much beagle love? :-)
[04:29] <daniels> does anyone here use nvidia on amd64?
[04:30] <thom> fuck no
[04:31] <Keybuk> inotify rocks
[04:31] <thom> Keybuk: yeah. dbus hates me right now though, otherwise i'd be basking in the glow of my indexed home directory
[04:31] <jdub> mjg59: you adding that crackrock DSDT-on-initrd patch?
[04:32] <bob2_> this thomboy thing looks cool
[04:32] <bob2_> I actually get to have thom sort and index all my notes?
[04:32] <tseng> hah thomboy.
[04:32] <bob2_> rock!
[04:32] <thom> tomboy is ++good
[04:33] <thom> bob2_: i'll give you a guided tour too
[04:33] <bob2_> hahahaha
[04:33] <Keybuk> rofl @ thomboy
[04:33] <thom> ... "and these are your porn links, and this is your suckful laptop ... "
[04:33] <thom> "and this here is the url to buy the X40 you know you want"
[04:34] <bob2_> don't tempt me, it was bad enough when a friend told me I could get 10% off
[04:36] <jdub> we should totally have the X40 on our 'book list' on the website
[04:36] <jdub> "this is what the developers use"
[04:36] <daniels> heh
[04:36] <jdub> "support monoculture!"
[04:36] <daniels> thom: that's what Jim wanted me for
[04:36] <bob2_> especially now I see it supports suspend-to-ram
[04:36] <bob2_> which00:05 < bob2_> hm, apache segfaults when I hit a php page
[04:36] <bob2_> 00:05 < elmo> bob2: it's trying to tell you something - take the hint ;-)
[04:36] <daniels> 'someone I know is using the nVidia driver on AMD64 on Ubuntu and it's arse'
[04:36] <bob2_> oops
[04:36] <bob2_> and has a fucking proper middle mouse button
[04:37] <daniels> thom: dude, the tour would totally kick arse
[04:37] <seb128> jdub,mdz: I need an approval for #1221 #2357 and #2477 ...
[04:37] <bob2_> so I don't have to use a key on my keyboard tha I hit randomly
[04:37] <daniels> thom: 'thom, what directory am I looking at?'
[04:37] <bob2_> "thom, what's the name of the building in this picture?"
[04:38] <thom> daniels: oh lovely
[04:39] <thom> daniels: i think Mithrandir has one
[04:39] <daniels> Mithrandir: you win!  any luck with the pos binary driver?
[04:40] <bob2_> Mithrandir: btw, your p150 works as usb-storage, right?
[05:06] <Mithrandir> bob2_: yes.
[05:06] <daniels> Mithrandir: have you used the nvidia driver at all?
[05:07] <Mithrandir> daniels: worked when I tested it, yes.
[05:07] <Mithrandir> on an amd64 system
[05:07] <Mithrandir> with a tnt2 ultra card
[05:07] <daniels> no mysterious segfaults?
[05:08] <Mithrandir> none I saw, no, but I really just used it enough to get the package working.
[05:08] <bob2_> Mithrandir: hm, how happy are you with it?
[05:08] <bob2_> I'm looking at cameras, but I know jack about them
[05:08] <Mithrandir> bob2_: unlike most other digital cameras, it feels like a real camera, speed-wise.
[05:08] <bob2_> and the p150 seems to be getting quite good reviews
[05:09] <Mithrandir> it's a bit too hard on the image enhancement/noise reduction part, which means you lose some details, but I'm willing to live with that.
[05:10] <bob2_> when do you notice the loss of detail? 800x600 on a crt or 1600x1200 and a magnifying glass?
[05:12] <Mithrandir> just a sec, I'll pull some pics off the camera
[05:14] <bob2_> ah, thanks
[05:17] <Mithrandir> bob2_: http://raw.no/tmp/dsc00071.jpg ~3.2MB, max resolution.
[05:17] <Mithrandir> I can go outside and take some pictures of trees and leaves, you'll probably see it better there.
[05:19] <bob2_> hmm, I see
[05:19] <bob2_> is that a bloody x40?
[05:19] <Mithrandir> no, r32.
[05:20] <bob2_> ah, good ;)
[05:20] <bob2_> the keys seems to have a bluish tinge
[05:20] <Mithrandir> they're well-worn so they reflect a bit of light.
[05:20] <bob2_> ah
[05:20] <Mithrandir> the keyboard really is a bit bluish in this light.
[05:21] <Mithrandir> but give me five minutes and I'll find you some trees. :)
[05:21] <bob2_> heh, I'm just not sure where I should look for loss of detail :)
[05:25] <Mithrandir> http://raw.no/tmp/dsc00076.jpg ; you'll see it a bit if you look at the leaves.
[05:26] <Mithrandir> but again, it's not really a big deal; the camera is the same size as the ixus and is just as good quality-wise and a whole lot faster.
[05:26] <Mithrandir> I recommend it a lot
[05:29] <bob2_> hm, that's not too bad
[05:32] <Kamion> elmo: if you want to kill daily-installer-*, now's a good time
[05:34] <elmo> kill on the buildds, or ?
[05:35] <sivang> Mithrandir : hey I saw my nick on the photo :)
[05:35] <sivang> Mithrandir : do you have any photos of youerself?
[05:35] <bob2_> Mithrandir: thanks for that, now I'll have to go play with one and then justify the purchase :)
[05:37] <Mithrandir> sivang: http://planet.debian.org/heads/tollef.png is the one I have on planet debian
[05:39] <Keybuk> heh, that was from Mlaga iirc.
[05:40] <Mithrandir> yeah, I think so.  it was you who made it, I think?
[05:40] <Keybuk> yup
[05:41] <bob2_> ph33r.
[06:00] <fabbione> daniels: do you have any hard feeling for ProjectManSuffix?
[06:02] <sivang> pitti : are you familiar with dnsmasq ?
[06:03] <sivang> opos his not here , anybody else knows what I'm talking about?
[06:12] <pitti> sivang: no idea
[06:17] <sivang> pitti : It's a nice small DNS hack which provides basic DHCP services to an internal LAN, as well as DNS caching.
[06:18] <sivang> pitti : but only some of my machines on the internal lan can be accessed using their hostname..I couldn't find a way to make those who aren't to be recognized besides their IP address.
[06:26] <lamont> Mithrandir: did the patch for warty#2495 ever get pushed up to debian?
[06:27] <Mithrandir> evidently not; I don't remember if I did or not.
[06:33] <pitti> mdz: Morning :-)
[06:33] <mdz> morning
[06:33] <pitti> mdz: you missed the funny flame^Wdiscussion
[06:34] <mdz> I know, I couldn't spare the sleep
[06:34] <mdz> reading it now
[06:34] <pitti> mdz: well, I don't think you missed sth truly critical :-)
[06:46] <alextreme> gday
[06:53] <seb128> mdz: hello
[06:53] <mdz> seb128: hi
[06:54] <seb128> mdz: need some approvals: #1221 #2357 and #2477
[06:56] <mdz> seb128: annotated the bugs
[06:56] <seb128> thanks
[06:57] <lucas_> I'm trying to install Ubuntu from a old Preview CD. During the "load installer components from CD-ROM part", I get the following error in the 4th console :
[06:57] <lucas_> "anna" process says "grep: /cdrom/dists/stable/Release : Not a directory
[06:58] <lucas_> is this problem known ?
[06:59] <mdz> lucas_: this is not a support channel, /join #ubuntu
[07:01] <seb128> mdz: are we supposed to get a gnome2-user-guide modified for warty before the release ? IIRC some guys were working on it. If not I would like to upload 2.8.1 (we have 2.8.0) which add the documentation about desktop sharing with vino
[07:02] <mdz> seb128: please go ahead with 2.8.1; we can always change it
[07:02] <seb128> ok
[07:11] <sivang> seb128 : I am working on it
[07:12] <sivang> seb128 : I have trouble reviewing it with yelp,
[07:12] <seb128> sivang: how long before getting it ready to upload ?
[07:12] <sivang> seb128 : lemme estimate
[07:12] <sivang> seb128 : thought : we could always provide an update after release, right?
[07:14] <sabdfl> jdub: around still?
[07:14] <thom> sabdfl: he went to bed ages ago
[07:14] <sabdfl> ok
[07:15] <seb128> sivang: more useful before the release :)
[07:16] <mjg59> Any volunteers for acpi testing?
[07:16] <mjg59> I need someone with a machine that currently works, and someone with one that doesn't
[07:17] <Keybuk> testing what, in particular?
[07:17] <thom> mjg59: my x40 is available
[07:17] <thom> on the understanding that it's available to me again afterwards ;-)
[07:18] <Keybuk> thom: dude, it's an X40 ... you are available to *it*, not the other way around
[07:18] <mjg59> Keybuk: People have found a couple of reasons for resume not working 
[07:18] <thom> Keybuk: chuckle
[07:19] <Keybuk> sure, I'll have a test
[07:19] <sivang> seb128 : with yelp crashing on me here and then :) and well, taking in consideration I've got about 70% more to go , I'd say about 4-5 hours..:(
[07:20] <seb128> sivang: ok, no problem
[07:20] <Keybuk> seb128: fix rhythmbox *puppy dog eyes*
[07:20] <mjg59> Might fix the HP weirdness
[07:20] <Keybuk> the HP weirdness is just weird :-/
[07:21] <mdz> pitti: here?
[07:21] <pitti> mdz: yes
[07:21] <mdz> pitti: looking at #1499
[07:21] <sivang> seb128 : thanks
[07:21] <mdz> pitti: I understand the simple change of the hal property name
[07:21] <pitti> mdz: :-/ Sometimes I hate my timing
[07:21] <mdz> pitti: but why the hal_device_get_property_string -> strcmp?
[07:21] <pitti> mdz: I think the change itself is not the problem
[07:21] <mjg59> Keybuk: The two issues here are:
[07:22] <pitti> mdz: it's not a simple yes/no boolean, but the string can have various values
[07:22] <mjg59> 1) The wakeup address being stored as a virutal address rather than a physical one
[07:22] <mdz> ah, I see
[07:22] <mdz> so not only the name, but the semantics changed
[07:22] <mjg59> 2) The GDT not being addressable from real mode
[07:22] <sivang> seb128 : you know anyway to make yelp like, "refresh" the xml page it's looking at?
[07:22] <Keybuk> those both sound likely candidates
[07:22] <sivang> (instead of killing it and starting it with the updated versino)
[07:22] <pitti> mdz: I think the patch itself is uncritical, but the new feature has not been tested
[07:23] <pitti> mdz: it works fine for me, but if 100 other people use it, it might reveal serious problems
[07:23] <mdz> pitti: ok, understood
[07:23] <seb128> sivang: no
[07:23] <mdz> I thought from your comments that it was only a name change
[07:23] <pitti> mdz: e. g. Debian package has another patch that disables mounting at startup if the user disabled automounting at all
[07:24] <pitti> mdz: it's a change like filesystem=true to usage="filesystem"
[07:24] <mdz> right
[07:24] <pitti> mdz: people who really want it can still get it from my unofficial repo...
[07:26] <mjg59> Keybuk: thom: Can you grab http://www.codon.org.uk/~mjg59/tmp/acpikernel ?
[07:26] <mjg59> It's just a kernel, no modules
[07:26] <Keybuk> boot with init=/bin/sh ?
[07:27] <mjg59> Then boot it from grub with init=/bin/bash and mount proc, then echo -n 3 >/proc/acpi/sleep
[07:27] <mjg59> Yeah
[07:27] <mjg59> thom: You'll want acpi_sleep=s3_bios, too
[07:27] <Keybuk> does that include the radeon wakeup patch, btw?
[07:28] <Keybuk> vgapost, or whatever it does
[07:28] <mjg59> Nope
[07:28] <mjg59> We'll worry about that later
[07:30] <thom> right, rebooting
[07:31] <sivang> seb128 : send you the bt
[07:31] <sivang> seb128 : for yelp :)
[07:32] <seb128> ok
[07:32] <lamont> could some bandwidth-possessing person please grab http://people.ubuntu.com/~lamont/testing/LiveCD/current/warty.iso and tell me if it boots?
[07:32] <lamont> whatever that means. :-)
[07:33] <sivang> lamont : downloading...
[07:33] <lamont> thanks.  If it boots, I want to ask #ubuntu to test it, etc.
[07:34] <sivang> lamont : ofcourse
[07:34] <lamont> this is more of a process check to make sure I didn't screwup
[07:34] <thom> mjg59: it suspended ok
[07:34] <Mithrandir> ETA ~30 minutes
[07:34] <mjg59> thom: And resumed?
[07:34] <sivang> lamont : no prob
[07:34] <sivang> ah, Mithrandir is way ahead of me
[07:34] <sivang> i'm at eta 1:53
[07:35] <thom> mjg59: blank screen when it came out of suspend (ie, the suspend light went off, nothing else changed)
[07:35] <mjg59> thom: Was it alive?
[07:36] <thom> mjg59: appeared to be
[07:37] <mjg59> thom: Had you passed acpi_sleep=s3_bios?
[07:37] <thom> oh, arse
[07:37] <mjg59> Heh
[07:37] <mjg59> Ok, sounds like it works
[07:37] <mjg59> Keybuk: Ping?
[07:37] <lamont> sivang/Mithrandir: when it appears hung, you have to give it a minute or 3... :-)  gnome startup isn't fast reading from a compressed disk image on cdrom
[07:38] <pitti> mdz: permission to upload #2496?
[07:38] <mdz> pitti: yes
[07:40] <thom> mjg59: works fine with acpi_sleep=s3_bio
[07:40] <thom> s
[07:40] <seb128> sivang: turn gail off for the crasher
[07:41] <Mitario> heya
[07:41] <mjg59> thom: Rock
[07:41] <mjg59> Now, where's Scott gone?
[07:41] <mjg59> I need someone with utterly broken ACPI resume...
[07:42] <mdz> lamont: please name the live CD .isos something more descriptive than "warty.iso"
[07:42] <lamont> ok
[07:42] <lamont> you want a date, or just 'warty-live.iso'
[07:42] <lamont> ?
[07:43] <Kamion> lamont: they'll be warty-live-i386.iso in the final published CD images
[07:44] <Kamion> or warty-final-live-i386.iso, depending
[07:44] <Kamion> mdz: OK to upload this to debian-installer?
[07:44] <Kamion>   * Manual changes:
[07:44] <Kamion>     - Refer to 'Ubuntu 4.10 "Warty Warthog"' rather than just 'Ubuntu 4.10'.
[07:44] <Kamion>     - howto/installation-howto.xml: Ubuntu branding. Fix up for Ubuntu
[07:44] <Kamion>       installer modifications.
[07:44] <Kamion>     - install-methods/official-cdrom.xml: Link to releases.ubuntulinux.org.
[07:44] <Kamion>   * Install Ubuntu splash image, so that pxeboot.tar.gz now has the correct
[07:44] <Kamion>     image too (closes: Ubuntu #2343).
[07:44] <mdz> Kamion: yes
[07:44] <sivang> seb128 : how do I turn it off?
[07:44] <Kamion> good-oh, should hopefully be the last for Warty
[07:45] <Kamion> I'm trusting nobody has any urgent initrd-udeb changes ...
[07:45] <mjg59> Anyone else with fucked ACPI able to test something?
[07:45] <lamont> mdz: renamed to warty-live-i386.iso
[07:46] <Keybuk> mjg59: back.
[07:47] <Keybuk> so after I got a mini root on /boot ... :p  no, didn't seem to work
[07:47] <mjg59> Bah.
[07:47] <mjg59> Identical failure to before?
[07:47] <Keybuk> not quite
[07:47] <Keybuk> I think the keyboard came back for a few seconds
[07:48] <mjg59> Hmm. Interesting.
[07:48] <mjg59> Can you give it a go with acpi_sleep=s3_bios and acpi_sleep=s3_mode ?
[07:48] <mjg59> One at a time to start with, and finally both together
[07:49] <sivang> seb128 : it only happens when I enable "assitive technologies "
[07:49] <sivang> seb128 : I tuned it off globally
[07:53] <mdz> lamont: ETA 20m
[07:56] <Keybuk> mjg59: I'm assuming that 'echo 3 > /proc/acpi/sleep' doesn't return until the machine is awake?
[07:56] <mjg59> Keybuk: Correct
[07:57] <mjg59> Unless the sleep fails
[07:57] <Keybuk> ok, I have a little shell script that does that, and immediately after writes a file and calls sync a few times
[07:57] <Keybuk> the sleep works fine, it's resume that's unwilly
[07:57] <Keybuk> uh, unwilling
[07:57] <mjg59> Yeah. It's almost certainly the suspend code that's broken, though.
[07:58] <mjg59> Hmm. So no significant difference between that code and previous attempts?
[08:00] <lamont> mdz: thoughts on 2482?
[08:01] <mdz> lamont: investigate and fix if necessary
[08:01] <lamont> duh.
[08:01] <lamont> was just reading the changelog... looks like sync'ing would be risky
[08:06] <Keybuk> mjg59: just testing every combination of things I can think of
[08:07] <mjg59> Keybuk: Ok
[08:07] <Keybuk> (and pausing in between because you don't have thermal zone support in that kernel, so my laptop sulks a little :p)
[08:08] <mjg59> Heh
[08:17] <Keybuk> nope, doesn't look like any combination of options produces an "I'm awake" in the log
[08:18] <mjg59> Right. And this is the same behaviour as older kernels?
[08:18] <Keybuk> yup
[08:19] <mjg59> Hmm.
[08:19] <mjg59> Tracking this down is likely to involve a moderate amount of pain.
[08:22] <mjg59> Keybuk: Next step is probably putting beep statements into various parts of the wakeup code
[08:22] <Keybuk> heh
[08:22] <Keybuk> beep & sleep ? :p
[08:22] <Keybuk> and count how many beeps I hear?
[08:22] <mjg59> Yeah
[08:22] <Keybuk> I'll give it a go
[08:23] <mjg59> Pavel's got an x86 assembly beep implementation on acpi-devel at the moment
[08:23] <Keybuk> though that'll involve finding the wakeup code and beep code
[08:23] <Keybuk> lol
[08:23] <mjg59> I'll head home and then start stuffing stuff into a kernel
[08:23] <Keybuk> ok, cool
[08:23] <mjg59> Back soon
[08:24] <Keybuk> okies
[08:32] <lamont> mdz: CD status?
[08:33] <mdz> lamont: burning
[08:34] <mdz> yay, it's labeled Ubuntu
[08:36] <lamont> mdz: Yeah, that's a simple post-install of /usr/bin/make-iso :(
[08:42] <mdz> lamont: sound comes up muted
[08:42] <mdz> but GNOME starts and I'm logged in
[08:42] <lamont> interesting - sound isn't muted for me...
[08:42] <lamont> (home edition)
[08:42] <lamont> if you unmute it, does it work?
[08:43] <mdz> network came up
[08:43] <mdz> firefox works
[08:44] <mdz> doesn't seem to work even after unmuting
[08:44] <mdz> modules are loaded
[08:45] <mdz> ah, I think the sound devices are coming up in reverse order
[08:45] <mdz> compared to real Warty
[08:46] <mdz> yep
[08:46] <mdz> my sound device is /dev/dsp1 now
[08:48] <seb128> mdz: #2141
[08:50] <mdz> seb128: go ahead
[08:50] <seb128> thanks
[08:50] <lamont> alsa sound is what we want?
[08:50] <lamont> (because that's what I get...)
[08:52] <mdz> yes
[08:52] <mdz> alextreme: how does the loading of ALSA modules work in morphix?
[08:52] <mdz> apparently, it's different from what we do with hotplug
[08:55] <alextreme> mdz: uses modules.dep and a bunch of sed & awk, is a modified version of a script in alsa afaik
[08:55] <mdz> eek
[08:55] <mdz> it happens to load the drivers in reverse order from hotplug on my laptop
[08:55] <mdz> so my modem gets /dev/dsp rather than the useful sound device
[08:56] <alextreme> heh, we've been working on a new one now we've moved to 2.6, makes things a lot easier
[08:59] <lamont> bbiab
[08:59] <alextreme> how're wartys RC bugs hanging in?
[09:13] <lucas_> 1
[09:13] <lucas_> oops
[09:27] <thom> mjg59: i think you broke him
[09:30] <mdz> alextreme: no RC bugs remain
[09:32] <alextreme> cool
[09:37] <mdz> did the new X make it onto the daily CD?
[09:47] <alextreme> not sure exactly if it made mine, CDT is... -6?
[09:52] <mdz> yes
[09:52] <mdz> er
[09:52] <mdz> no, CDT is -5
[09:52] <mdz> but why CDT?
[09:56] <mdz> lamont: the CD you pointed me to seems to work well
[09:57] <elmo> why does postfix not log to it's own logfile?
[10:00] <elmo> lamont: why does postfix not log to it's own logfile?
[10:00] <elmo> thanks mdz ;-P
[10:01] <mdz> anytime
[10:09] <alextreme> was looking at warty-changes to figure out when the new X entered the archive
[10:21] <amu> testing/LiveCD/current is the correct folder ? 
[10:21] <mjg59> Keybuk: Got another kernel for you
[10:22] <mjg59> Keybuk: http://www.codon.org.uk/~mjg59/tmp/acpikernel
[10:22] <mjg59> Should produce a continuous tone on resume. If so, we start a binary search. If not, things are very fucked.
[10:23] <Keybuk> okie dokie
[10:23] <Keybuk> try with/without acpi_sleep=* ?
[10:24] <mjg59> Without for now
[10:31] <Keybuk> no beeping
[10:32] <mjg59> None at all?
[10:32] <Keybuk> nope
[10:32] <mjg59> Oh dear
[10:33] <Keybuk> isn't there any earlier in the wakeup you can put that then? :)
[10:33] <mjg59> Not trivially...
[10:33] <mjg59> That's right at the start of wakeup_start
[10:35] <Keybuk> yeah, it's always felt like that whatever code the kernel has to wake up isn't getting run
[10:36] <mjg59> So there's something wrong in the suspend code
[10:36] <Keybuk> it's like the kernel suspends it ok, but doesn't put the right address in the right place or something to be woken up -- the hardware comes up by there's no kernel waiting for it
[10:36] <mjg59> It's an nc4010, right?
[10:36] <Keybuk> yes
[10:38] <mjg59> And the speaker is switched on?
[10:38] <Keybuk> I assume so; I assume you compiled in pcspkr or whatever
[10:39] <mjg59> No, but it's writing directly to the chip
[10:39] <Keybuk> should be fine then
[10:39] <Keybuk> grub can make it beep, anyway
[10:40] <lamont> alextreme: CST is -6, CDT is -5
[10:40] <lamont> elmo: because Wietse didn't want to reinvent syslog when there was already a nice facility there.
[10:40] <lamont> mdz: OK.  I'll ask for testers on #ubuntu then
[10:45] <elmo> lamont: meh, that's lame
[10:56] <alextreme> lamont: k, thanks :)
[11:00] <mjg59> Argh. We're getting the entire Smoothwall development team?
[11:01] <mdz> lamont: syslog is not "nice" :-)
[11:03] <lamont> mdz: well, yeah
[11:03] <lamont> so why did we not hit the gzip futex bug until recently?
[11:06] <plovs> any of the documentation people awake?
[11:11] <mdz> lamont: what's the latest mplayer story, now that xvidcore was synched?
[11:12] <mdz> lamont: I guess that relatively few folks are using futexes on SMP systems with apt; it seems to be difficult to trigger
[11:12] <lamont> unknown/mplayer_1:1.0-pre5-0.4: Dep-Wait by buildd+macaroni [-:uncompiled] 
[11:12] <lamont>   Dependencies: libavcodec1-dev
[11:12] <lamont> grumble
[11:15] <lamont> ffmpeg delivers libavcodec2-dev, but it's ftbfs
[11:15] <lamont> i386/dsputil_mmx.c: In function `h263_h_loop_filter_mmx':
[11:15] <lamont> i386/dsputil_mmx.c:634: error: can't find a register in class `INDEX_REGS' while reloading `asm'
[11:17] <lamont> mdz: in other words, we're still short a build-dep or 6
[11:18] <mdz> lamont: are you overriding its optimization flags?
[11:18] <lamont> mdz: we force optimization to 0..3, you can't say -O6, or you get -O3.
[11:18] <lamont> that's all on that front.
[11:18] <lamont> but we do force the architecture level...
[11:18] <lamont> hrm.
[11:19] <lamont> or rather, ffmpeg, eh>
[11:21] <lamont> build running
[11:31] <amu> lamont: morphix grubsplash, booting takes hours ;)  gnomepanel is working, sound same problems like before
[11:33] <lamont> amu: which image are you booting?
[11:33] <amu> testing/LiveCD/current  
[11:34] <amu>  688748544 2004-10-18 19:28 warty-live-i386.iso
[11:35] <lamont> amu: process issue resolved. thanks
[11:36] <amu> i didn't took the time, guess I waited for 5min. ;) 
[11:36] <sivang> jdub : around ?
[11:38] <sivang> seb128 : ping
[11:38] <seb128> pong
[11:38] <amu> lamont: bootspash looks empty, you should write something nice "Take your Ubunto to everywhere" *ducks*  
[11:38] <sivang> seb128 : going through the manual, I see we don't have "Browse Folder" option on the default nautilus layout
[11:38] <Kamion> "Ubuntu"
[11:39] <seb128> sivang: ?
[11:39] <lamont> alextreme: thoughts on the sound-isms?
[11:39] <lamont> amu: all the splashes are just me throwing something in.  Not to be confused with any kind of blessed artwork.
[11:39] <sivang> seb128 : the manual describes how to use the "open each fodler in a new window" nautilus window
[11:40] <sivang> seb128 : and choos "File"->"Browse Folder" to make it show in the file browser (it'
[11:40] <sivang> seb128 : mode of nautilus. But we don't have this on the default layout..DO you know if it's something that was chagned in Ubuntu compared to upstream 2.8 ?
[11:40] <amu> hmm alsa, there was a cannot cp file errormessi while starting 
[11:40] <seb128> sivang: I've it here
[11:41] <seb128> can you make a screenshot of a window with the file menu open ?
[11:41] <sivang> seb128 : yes
[11:41] <alextreme> lamont: what soundisms?
[11:41] <amu> lamont: cp /etc/devfs/conf.d/Alsa no such file or dir 
[11:42] <lamont> alextreme: devices getting found in the (apparently) reverse order, leading to the sound system running on the modem, and not the sound card.
[11:42] <lamont> that is /dev/dsp pointing to the wrong device, because the two "sound cards" were found in the reverse order.
[11:43] <alextreme> ahh, what matt said earlier. i'll take a peek, shouldn't be hard to reverse
[11:43] <lamont> yeah - that
[11:43] <lamont> with luck, that's the whole sound issue
[11:44] <mdz> lamont: how did ffmpeg go the second time around?
[11:44] <mdz> it has lots of inline assembly, so it tends to be sensitive to optimisation
[11:44] <amu> alextreme: btw. hi 
[11:45] <lamont> mdz: uploaded
[11:45] <mdz> cool
[11:45] <alextreme> amu: gday :)
[11:46] <lamont> mdz: but that doesn't fix the problem, you know.
[11:46] <mdz> lamont: oh?
[11:46] <lamont> ffmpeg delivers libavcodec___2___
[11:46] <lamont> and mplayer b-d's libavcodec1-dev
[11:47] <mdz> http://www.las.ic.unicamp.br/pub/debian-marillat/dists/unstable/main/source/mplayer_1.0-pre5-0.6.dsc
[11:48] <mdz> build-depends: libavcodec2-dev
[11:48] <mdz> please request a sync
[11:48] <lamont> from that url?
[11:49] <lamont> sent
[11:50] <elmo> uh
[11:50] <elmo> ftp://ftp.nerim.net/debian-marillat/
[11:50] <elmo> is what I've been using?
[11:51] <amu> :)
[11:51] <lamont> but I bet elmo has a good site for it.
[11:51] <lamont> esp if it has 1.0-pre5-0.6 :-)
[11:51] <sivang> this is also what I am using
[11:52] <elmo> oh, never mind it's one of the stupid, "I have two versions in the Sources file" victims
[11:52] <mdz> elmo: that's a mirror of the same stuff
[11:52] <elmo> I didn't sync in case it had actual changes
[11:52] <mdz> which google found first
[11:53] <lamont> mdz: btw, doing a blind bootstrap attempt on everything currently dep-waiting, just to be thurough and tie up a buildd for a while.
[11:55] <lamont> OT: hppa is now 30.5 hours into building the latest gcc-3.4
[11:55] <doko> lamont: wrong channel ;)
[11:56] <doko> should I start another one, maybe I get it finished earlier ...
[11:57] <lamont> doko: well, yes. hence the "OT"
[11:57] <lamont> actually, I killed the 8 runaway processes from a previous build
[11:57] <lamont> unless you want to claim lt-gij or SyncTest.exe
[11:58] <doko> smp kernel?
[11:58] <lamont> once it gets more than 12% of the CPU, it should build faster.
[11:58] <lamont> UP kernel
[11:58] <doko> strange, I've seen this with smp kernel only.