[12:25] <mdz> aes: feedback on the menus would be good to have on the sounder list; I believe there's already a thread on it
[12:26] <seb128> mdz: no, the thread is on warty list
[12:26] <mdz> hmm, it should be on sounder
[12:26] <seb128> should probably be moved on sounder
[12:26] <seb128> yes ...
[12:31] <lamont> Kamion: I wonder if it would be possible to get the cd-creation code and an rsync-able directory of the ISO contents somewhere - that'd probably go faster than rsync'ing the iso, I think
[12:37] <aes> is warty list a list I should be on, or not?
[12:42] <mdz> no, that is the list where these discussions took place before we had the sounder list; it is obsolete for that purpose
[12:44] <aes> ok
[12:49] <Kamion> lamont: the CD-creation code is at http://ftp.no-name-yet.com/cdimage/code/ plus colin.watson@canonical.com--2004/cdimage--mainline--0, but I wouldn't really advise doing it that way at the moment ...
[12:50] <Kamion> lamont: you could also try netboot?
[12:58] <lamont> Kamion: yeah, I could at that.  What I'm actually going to do is hit an internet cafe in town tonight
[01:15] <Kamion> erk, has the terminal been removed from the menu entirely?
[01:16] <mdz> Kamion: Applications->System Tools->Terminal?
[01:16] <Kamion> not as of the latest gnome-panel
[01:16] <Kamion> (post-Sounder-7)
[01:19] <Keybuk> seb reverted that one, didn't he?
[01:19] <mdz> it's still there for me
[01:19] <Keybuk> system tools jumped menus briefly
[01:19] <mdz> gnome-panel 2.7.91-0ubuntu2
[01:19] <Kamion> you can get to it by right-click-on-desktop->Open Terminal, but that really sucks on PowerMacs which only have one mouse button
[01:19] <Kamion> ah
[01:19] <Kamion> I have 2.7.91.1-0ubuntu2 here
[01:21] <Kamion> gnome-vfs2 2.7.91-0ubuntu4 required apparently, ok
[01:21] <seb128> yes
[01:21] <seb128> accidently removed the system tools menu 
[01:21] <Kamion> good stuff, I was afraid it was deliberate :-)
[01:21] <Kamion> thanks for the fix
[01:22] <seb128> you're welcome :)
[01:22] <Kamion> mind you we should configure mouse button emulation on powermacs by default, I think ...
[01:26] <Oskuro> seb128: the entire menu? nice :D
[01:27] <seb128> Oskuro: that's a sub-menu of Applications dude :)
[01:27] <Oskuro> seb128: heh, if I didn't have enough with gtk and nautilus for "big uploads I need to do", I just did abiword to complete the set :)
[01:27] <Oskuro> seb128: yeah, still :)
[01:27] <seb128> Oskuro: yes, I've seen that
[01:27] <seb128> Oskuro: but I've not seen the nautilus one
[01:28] <seb128> Oskuro: do you need some help for the uploads ?
[01:30] <Keybuk> Kamion: warty base ... has that not been set with priorities?
[01:30] <Keybuk> stuff like discover and evms are "optional" according to aptitude
[01:31] <Oskuro> seb128: I guess I can manage, but if I haven't done them by tomorrow at 17:00 I guess you could do one of gtk or nautilus
[01:31] <Kamion> priorities haven't been changed AFAIK
[01:31] <Kamion> hell, they aren't in sync in Debian :-)
[01:31] <Keybuk> so what's the easiest way to install base from an existing install?
[01:31] <Oskuro> with gtk we need to be very careful first to check it really doesn't need shlib bumping
[01:31] <seb128> Oskuro: ok, just let me know
[01:31] <Oskuro> seb128: k
[01:31] <Kamion> Keybuk: debootstrap
[01:31] <Kamion> (hey, shouldn't debootstrap be in Ship?)
[01:31] <Keybuk> will that upgrade an existing machine?
[01:31] <Keybuk> I thought it just made new ones?
[01:32] <Kamion> confused, what are you trying to do?
[01:32] <Kamion> yes, debootstrap constructs new directory trees
[01:32] <Keybuk> change my debian/unstable machine into a warty one
[01:33] <Kamion> oh, I don't think we have a good way to identify "just base" in that situation ... although I'm not sure why you'd want that
[01:33] <Keybuk> because then it wouldn't be a warty machine?
[01:33] <Kamion> warty is more than just base
[01:33] <Keybuk> yeah, but isn't having base important?
[01:33] <Kamion> if you only have warty main in /etc/apt/sources.list then aptitude will tell you what's obsolete/local
[01:34] <Kamion> pkg-diff.py? :)
[01:34] <Keybuk> yeah the obsolete/local stuff I've dealt with
[01:34] <Keybuk> I've still got a huge bunch of "not installed" packages though
[01:34] <Keybuk> and it's not easy to work out which I'm supposed to have
[01:34] <Kamion> maybe we should have a "Task: base" so that you can do that easily
[01:35] <Kamion> sounds like a reasonable thing to add
[01:35] <Kamion> which reminds me, I need to fix aptitude to stop putting Desktop in "Unrecognized tasks"
[01:36] <Keybuk> heh
[01:39] <mdz> Keybuk: apt-get install `debootstrap --print-debs ...`
[01:41] <Kamion> mdz is more awake than I am
[01:41] <Kamion> I underestimated how many things were hard-coded into aptitude's source ... like /usr/share/tasksel/debian-tasks.desc
[01:42] <mdz> does it work to have a package which is part of multiple tasks?
[01:42] <mdz> yes, a base task would be nice
[01:44] <Kamion> lots of packages are part of multiple tasks in Debian
[01:45] <Kamion> chinese-s, chinese-t and cyrillic-desktop, russian-desktop seem pretty common at a glance
[01:45] <Keybuk> ok, so...
[01:45] <Keybuk> I have both warty and unstable in sources.list
[01:45] <Keybuk> warty is pinned at 750
[01:45] <Keybuk> apt wants to upgrade all the packages over to the unstable ones
[01:45] <Keybuk> (they're all the same version as the warty ones)
[01:46] <mdz> Keybuk: it doesn't work to have both warty and unstable in sources.list
[01:47] <mdz> warty and unstable have packages which have the same version number but are actually different
[01:47] <Keybuk> yeah, I'm noticing this
[01:47] <Kamion> --\ Tasks
[01:47] <Kamion>   --\ End-user
[01:47] <Kamion>     --\ Ubuntu desktop environment
[01:47] <Kamion> much better
[01:47] <Keybuk> Kamion++
[01:47] <Kamion> I wonder if we shouldn't be calling our tasks "ubuntu-base" and "ubuntu-desktop" though
[01:47] <Keybuk> it's kinda annoying ... there are some people who are going to want to cherry-pick from unstable still
[01:48] <Kamion> that way things won't have to break hideously if somebody decides to install tasksel from universe
[01:48] <mdz> Keybuk: that's what hoary is for
[01:48] <Keybuk> yeah, I guess our release process is fast enough that this isn't going to be a *real* problem
[01:48] <mdz> users should never need to go to Debian to get new stuff; we'll have it all in universe
[01:49] <Keybuk> have we fired up hoary yet?
[01:49] <Keybuk> or is that blocking on me not coding fast enough? :p
[01:49] <mdz> exactly
[01:50] <mdz> hct star-merge debian
[01:50] <Keybuk> heh
[01:50] <Keybuk> hct steal debian

[01:50] <Keybuk> grep -q off-line /proc/acpi/ac_adapter/*/state
[01:51] <Keybuk> ^ did thom not know about "on_ac_power" ?
[01:56] <sabdfl> npmccallum: see my bug on grub bootsplash?
[01:58] <Keybuk> once we get gold cvs & svn imports available, we'll then be able to put the upstream tarball imports on top, and then add debian and redhat
[01:58] <npmccallum> sabdfl: yes, I already responded
[01:58] <Keybuk> my plan for debian is to go back to the ubuntu split point using snapshot, and then import every release from there onwards
[01:58] <sabdfl> ok
[01:58] <npmccallum> sabdfl: Colin already posted a patch, I'll check it tomorrow
[01:58] <Keybuk> then we can make an ubuntu branch and import the warty changes with sourcerer onto it
[01:58] <npmccallum> sabdfl: I have to get a basic usplash in post-haste
[01:59] <Keybuk> and at *that* point, we can branch for hoary
[01:59] <Keybuk> there's a lot of cards there though :-/
[02:02] <Kamion> npmccallum: the patch as posted is wrong, but I've tested with a similar change and it's fine
[02:02] <Kamion>         if [ -n "$boot_device" ] ; then
[02:02] <Kamion>                 splash="${splash#/boot}"
[02:02] <Kamion>         fi
[02:02] <Kamion>         echo "splashimage=$grub_root_device$splash" >> $buffer
[02:03] <sabdfl> night all
[02:03] <npmccallum> Kamion: if it works, can you patch it and upload it?
[02:03] <Kamion> npmccallum: sure, let's make it tomorrow though :-)
[02:03] <npmccallum> Kamion: np
[06:49] <npmccallum> mdz: we have usplash packages :)
[07:11] <fabbione> morning guys
[07:14] <npmccallum> fabbione: morning?  I'm going to bed :)
[07:16] <fabbione> ehhe
[07:17] <fabbione> npmccallum: ain't my fault if you leave in the wrong tz ;)
[08:57] <Mithrandir> good morgning
[08:57] <Mithrandir> s/rg/r/;
[08:58] <fabbione> morning Mithrandir 
[08:58] <fabbione> Mithrandir: are you busy?
[08:59] <Mithrandir> for a little bit, yes.
[09:00] <fabbione> ok. did you have any time to check the changes to apache?
[09:00] <fabbione> we are really running out of time now
[09:03] <fabbione> Overfiend: never mind.. i get it
[09:03] <fabbione> ops
[11:23] <justdave> if anyone read the review I just posted on the sounder list, you know my X is dead. :)  Would anyone like to help me fix it?  I'm tired of working on console :)
[11:26] <justdave> X comes up, but the screen is scrambled
[11:27] <justdave> I'm out of ideas for what to try.  (this is on an iBook)
[11:27] <fabbione> justdave: sure..
[11:28] <fabbione> first.. apt-get --purge remove xserver-xfree86
[11:28] <fabbione> then be sure that there is no /etc/X11/XF86Config-4
[11:28] <fabbione> apt-get update
[11:28] <fabbione> apt-get upgrade
[11:29] <fabbione> apt-get install xresprobe discover1 mdetect laptop-detect
[11:29] <fabbione> apt-get install xserver-xfree86
[11:29] <fabbione> this should put all the correct things in the right place
[11:36] <fabbione> which kind of magic?
[11:36] <fabbione> is there any way to detect that we are working on a ibook2?
[11:36] <fabbione> if so what needs to be done to make it working properly?
[11:39] <justdave> ok, it's upgrading.
[11:40] <fabbione> justdave: than you will get asked for the resolution you want.
[11:40] <fabbione> and that's up to you to put reasonable values
[11:40] <justdave> ok
[11:43] <justdave> it's giving me 1152x864 as the default choice again.
[11:44] <fabbione> justdave: it's a hardencoded choise
[11:44] <fabbione> if the probe fails i have no way to know what is the resolution for your laptop
[11:44] <justdave> ok
[11:45] <justdave> ok, it started, and the video is still scrambled.
[11:45] <fabbione> otherwise i wouldn't bother to ask
[11:45] <justdave> so I guess we need to figure out what that ibook2 voodoo is :)
[11:46] <fabbione> justdave: can you send me /var/log/XFree86.0.log
[11:46] <fabbione> yes also
[11:46] <fabbione> justdave: but we need to be sure it is possible to recognize that we are working on a ibook2
[11:46] <mjg59> What's the name of the network applet you're using by default
[11:48] <justdave> the model number is available in /proc/(something)/serial-number
[11:48] <justdave> :)
[11:48] <justdave> what address do I mail it to?
[11:48] <fabbione> fabbione@fabbione.net 
[11:49] <justdave>  /proc/device-tree/serial-number
[11:50] <justdave> the first three characters are the last three of the serial number, then model number follows, followed by the first 8 of the serial number
[11:50] <fabbione> justdave: i have no ppc at home
[11:50] <justdave> no clue why it's wrapped around like that, but it is.
[11:51] <fabbione> i have no clue of ibook ser numbers but i would like to have something more robust than that
[11:54] <justdave> getting a nameserver timeout trying to find your mail server :(
[11:55] <fabbione> uh?
[11:56] <justdave> my dns is working, I can look up other domain names, but it says "no servers could be reached" when I try fabbione.net
[11:56] <fabbione> yes i am checking
[11:57] <justdave> the mail is going to bounce anyway, postfix seems to think my domain name is ibook.rieo.kni
[11:57] <fabbione> the isp where my dns's are hosted is down
[11:57] <fabbione> crap
[12:00] <justdave> I was on the hotel's LAN when it was installed, so that's what it stuck in the main.cf file. heh
[12:01] <fabbione> justdave
[12:01] <justdave> ok, postfix config is fixed.  I mailed it again, but it's queued since your dns it down.  you'll get it eventually :)
[12:01] <fabbione> justdave: did X work while we were in Oxford?
[12:02] <justdave> yes.
[12:02] <fabbione> without any magic?
[12:02] <justdave> I had sid that I overwrote with warty via apt-get dist-upgrade
[12:02] <justdave> which was kind of flaky in other places, but X worked.
[12:03] <justdave> on Friday I backed up my home directory, wiped it, and did a clean install of warty 6
[12:03] <fabbione> justdave: in that case X config was not overwritten
[12:04] <justdave> so my X config from sid would probably work, except that I didn't save it.
[12:04] <justdave> I did get it to work *once* after installing warty 6, but I haven't been able to repeat it again since.
[12:05] <justdave> that was after a remove --purge and reinstalling the package off the CD.
[12:05] <fabbione> justdave: please try this: edit your X config file
[12:05] <justdave> but after getting the newer one from the online repository, I couldn't get it to work again, even after reverting to the one on the CD
[12:05] <fabbione> go to Section "Monitor"
[12:05] <fabbione> and add these 2 lines:
[12:06] <fabbione> HorizSync "28-49"
[12:06] <fabbione> VertRefresh "43-72"
[12:07] <fabbione> restart X
[12:08] <justdave> ok, it won't even start now.
[12:08] <justdave> just get the error dialog from GDM saying it's going to disable it
[12:08] <fabbione> justdave: i need to see the log
[12:08] <fabbione> at least the last part of it with the error
[12:08] <fabbione> ops
[12:09] <justdave> I'm on console, not sure how to do that
[12:09] <fabbione> justdave: ssh ?
[12:09] <fabbione> put it on a web somewhere?
[12:09] <fabbione> chinstrap?
[12:09] <fabbione> anywhere it's fine
[12:09] <justdave> oh, that's a thought.
[12:09] <justdave> hang on
[12:12] <justdave> http://www.justdave.net/XFree86.0.log
[12:14] <justdave> hmm, that's obvious in that log...
[12:14] <justdave> hmm, it *is* followed by a range of numbers.
 HorizSync "28-49"
[12:15] <fabbione> try to remove the ""
[12:15] <fabbione> yeah it's without ""
[12:15] <fabbione> also for the Vertical
[12:16] <justdave> done, works :) :)
[12:16] <fabbione> is X scrambled now?
[12:16] <justdave> nope, it's perfect
[12:16] <fabbione> justdave: thanks
[12:18] <justdave> fabbione: thank you much for the assistance. :)
[12:18] <hrdwrbob_> hm
[12:18] <hrdwrbob_> time for ubuntu instlal on this box
[12:19] <justdave> irssi is actually quite nice for a text client.
[01:17] <fabbione> justdave: you around?
[01:18] <fabbione> anybody with a ppc?
[01:22] <hrdwrbob_> no sorry
[01:22] <hrdwrbob_> four intels no ppc :/
[01:22] <fabbione> hrdwrbob_: thanks
[01:22] <thom> I DID HAVE, until some fascist stole it
[01:23] <Mithrandir> thom: how do you know he wasn't a liberal, or a labourer?
[01:23] <thom> royal% uname -m
[01:23] <thom> ppc
[01:23] <hrdwrbob_> thom: have you located, beaqten, subdued, hung, drawn, quarted, and smacked the suspected perpetrator??
[01:23] <Mithrandir> fabbione: I think it might return ppc64 as well
[01:23] <fabbione> also on custom kernel?
[01:23] <thom> that's a patched but basically standard debian kernel
[01:24] <fabbione> xresprobe uses $ARCH to detect if we are on ppc...
[01:24] <fabbione> but $ARCH isn't set anywhere
[01:24] <fabbione> so i need to find a way other than dpkg-architecture to detect it
[01:25] <fabbione> perhaps /proc/cpuinfo ?
[01:25] <thom> there are ppc boxes that don't say 'ppc' to uname -m?
[01:26] <fabbione> thom: i would love to know that...
[01:27] <thom> i doubt there are
[01:27] <fabbione> ok...
[01:28] <fabbione> thom: can you do a test for me if you can? but you will have to exit X
[01:28] <fabbione> ah no
[01:28] <fabbione> never mind..
[01:29] <fabbione> i need justdave to do it
[01:31] <Kamion> I can't imagine any answers other than ppc or ppc64 ever
[01:33] <justdave> dave@ibook [12:32 malone 43]  tcsh> uname -m
[01:33] <justdave> ppc
[01:34] <fabbione> justdave: please test this for me:
[01:34] <fabbione> export ARCH=powerpc
[01:34] <fabbione> xresprobe
[01:35] <justdave> Driver must be specified.
[01:35] <fabbione> justdave: well specify a driver please
[01:35] <justdave> dave@ibook [12:34 malone 48]  tcsh> xresprobe ati
[01:35] <justdave> grep: /tmp/xprobe.7357/xfree86.log: No such file or directory
[01:35] <justdave> grep: /tmp/xprobe.7357/xfree86.log: No such file or directory
[01:35] <justdave> id: Generic Laptop LCD
[01:35] <justdave> res:
[01:35] <justdave> dave@ibook [12:35 malone 49]  tcsh> xresprobe r128
[01:35] <justdave> id: Generic Laptop LCD
[01:35] <justdave> res:
[01:35] <fabbione> justdave: which version?
[01:36] <fabbione> ok
[01:36] <fabbione> thanks
[01:36] <justdave> ii  xresprobe      0.2-0ubuntu2   X Resolution Probe
[02:25] <Kamion>     - Give the user a choice to leave the network unconfigured if nothing is
[02:25] <Kamion>       plugged in. (Closes: #264476)
[02:25] <Kamion> I think I'll do a merge
[02:26] <sladen> ...or even if it is     eg, it's detected the hotel wireless but using that (because it doesn't work) will fsck the rest of the install process over
[02:27] <justdave> yeah, I had that problem this morning trying to mail logs to fabbione...
[02:27] <justdave> the laptop wound up with the hotel LAN's domain name in the postfix config
[02:27] <Kamion> wireless detection is at a separate layer in netcfg, will look at it separately
[02:43] <lamont> justdave: dhcp is not always your friend.
[02:47] <Kamion> Mithrandir: for future reference, it's "#, fuzzy", not "# , fuzzy"
[03:01] <Mithrandir> Kamion: ahkay, I copied stuff off somewhere else.
[03:27] <fabbione> can anybody kindly confirm me if xresprobe has been accepted?
[03:27] <fabbione> (i am without mails)
[03:37] <mdz> *yawn* morning
[03:38] <mdz> fabbione: version?
[03:38] <fabbione> 0.3-0ubuntu1
[03:38] <mdz> -rw-r--r--          637  2004-08-25 14:25  xresprobe_0.3-0ubuntu1.dsc
[03:38] <mdz> source is in the archive
[03:38] <mdz> no binaries yet
[03:39] <fabbione> mdz: thanks
[03:39] <fabbione> no problem about the binaries... they can wait
[03:44] <Kamion> mdz: working on upgrading our netcfg to be based on Debian 1.01, which has several fixes we want; hope that's ok
[03:46] <mdz> Kamion: if you're ok, I'm ok
[04:05] <thom> Kamion: remind me how to get d-i to select the "copy all debs to disk" udeb?
[04:07] <Kamion> thom: boot with anna/choose_modules=archive-copier
[04:07] <Kamion> thom: and remember to boot with KEEP_DEBS=yes on the first reboot
[04:08] <thom> right
[04:08] <thom> thanks
[04:08] <thom> booting with APIC *utterly* fries my desktop - the installer just randomly hangs :/
[04:10] <mjg59> Which network applet are you guys using by default?
[04:15] <lamont> mdz: tetex-base appears to have a bug related to dh_fixperms :-(
[04:15] <mdz> lamont: _another_ bug? grr...
[04:16] <lamont> that's the /var/cache/fonts/* perms issue
[04:16] <mdz> this package has problems
[04:16] <lamont> it sets them to 1777, and then (I think - verifying now) dh_fixperms "fixes" them to 755
[04:16] <lamont> s/has/is/ :(
[04:16] <mdz> Kamion: oh, forgot to mention...I did another round of sounder 7 testing using archive-copier and KEEP_DEBS=yes; seemed to work perfectly
[04:18] <lamont> verified broken in current-sid...  time to fix and fiel
[04:20] <Kamion> mdz: kewl
[04:22] <rburton> hm, print in xpdf doesn't work for me
[04:23] <rburton> looks like it neglects a formfeed to get the printer to actually print
[04:24] <mdz> printing works in other apps?
[04:24] <rburton> not all
[04:25] <Kamion> mdz: that iftab thing in netcfg could go upstream, couldn't it?
[04:26] <mdz> Kamion: I assume so
[04:31] <thom> Kamion: um, current sounder daily seems well and truly unhappy
[04:32] <thom> didn't start base-config after the reboot, no /etc/network/interfaces, no entry in sudoers
[04:34] <thom> on i386, this is
[04:36] <Kamion> no idea how that could happen
[04:36] <Kamion> it was fine yesterday
[04:36] <thom> trying it again
[04:37] <thom> this system throws up some entertainments
[04:45] <thom> ooh, that's a fun one
[04:46] <thom> my dvd-rom drive only admits to being a dvd drive, not a cdrom drive. so /dev/dvd -> /dev/hdc and /dev/cdrom -> /dev/hdd
[04:46] <thom> (hdd is the rewriter)
[04:51] <edd> oh, ubuntu's metacity has some composite smarts, it seems. nice.
[04:53] <mdz> daniels: ping
[04:53] <daniels> mdz: pong
[04:53] <mdz> daniels: are you logged into jabber from someplace else?  tried to reach you there first
[04:54] <daniels> mdz: er, afaik I haven't been logged into Jabber for the last 48h. what's up?
[04:54] <fabbione> was xresprobe 0.3-0ubuntu2 accepted?
[04:55] <mdz> fabbione: not yet
[04:55] <fabbione> hmm i uploaded it 15 minutes ago....
[04:55] <fabbione> elmo: is katie running?
[05:00] <whiprush> will any mono apps be bundled with core? Or are they just going to remain in "universe"?
[05:01] <Keybuk> whiprush: hoary probably; we froze warty before mono 1.0
[05:01] <whiprush> ah, k.
[05:01] <Keybuk> and even the Ximian guys aren't hugely pushing the mono apps yet, so hoary is probably the right timeframe for it anyway
[05:02] <whiprush> I think I'm already addicted to f-spot is all.
[05:52] <mjg59> thom: What happened to the dreadful laptops?
[05:53] <thom> kamion has the C3 and the amd, iirc
[05:53] <thom> seb got the tosh
[05:54] <mjg59> Right, I need to harass Colin then
[05:55] <mjg59> thom: Ultra-rad laptop support should include netapplet
[05:55] <thom> mjg59: i was thinking that
[05:55] <mjg59> Haha
[05:55] <mjg59> Have you tested it?
[05:56] <edd> it's very nice
[05:56] <thom> gonna talk to jdub when he gets back
[05:56] <mjg59> Cool
[05:56] <thom> i'm amd64 monkey currently
[05:56] <mjg59> I've just fixed a bug that meant it picked up multiple copies of the same interface if ipv6 was being used
[05:56] <mjg59> It still doesn't deal with mapping stanzas, but other than that it seems pretty solid
[05:56] <thom> cool
[06:02] <edd> mjg59: looks like the netapplet has the same bug about the notification area disappearing that many apps have, sigh
[06:02] <daniels> edd: i suspect that's the one psi has
[06:02] <mjg59> edd: Which one is that?
[06:02] <mjg59> I had someone complaining that they couldn't see the icon, but could click on it
[06:03] <daniels> edd: half the time, psi just kicks up a new window with its notification icon, instead of using the area
[06:03] <edd> it's crash-on-re-add notif area
[06:03] <daniels> sometimes that window is blank, and the icon is in the panel's area
[06:03] <edd> daniels: sounds like a different bug
[06:03] <Kamion> mjg59: what about them?
[06:03] <edd> mjg59: this one's caused by c&p from broken code in libegg example. rhythmbox and gnome-obex-server get it right.
[06:04] <edd> mjg59: if i get the energy i'll make a patch
[06:04] <daniels> edd: rad
[06:04] <mjg59> edd: Thanks - that sounds good
[06:04] <mjg59> Kamion: Getting hold of the dsdt from the C3 would be good, if possible
[06:04] <edd> and meanwhile i spit flaming plastic death at libegg
[06:04] <Kamion> mjg59: what's a dsdt?
[06:04] <mjg59> Kamion: What you get when you cat /proc/acpi/dsdt :)
[06:05] <mjg59> It's the table that tells the ACPI interpreter how to manage the hardware
[06:05] <Kamion> ok, I haven't taken the C3 out of its box yet actually, but remind me tomorrow
[06:05] <Kamion> or if it's more urgent I'll make time this evening
[06:06] <daniels> it's not urgent. he's been waiting weeks already, he can deal with some more ;)
[06:06] <mjg59> Heh
[06:06] <mjg59> Yeah, it's not urgent
[06:06] <Kamion> it's getting quite scary
[06:07] <thom> gaim is broken wrt notification area, too
[06:07] <mjg59> Why did you end up with the C3? Just for C3 testing, or because you have sinned?
[06:08] <daniels> Kamion: well, clearly you have no need for that grey pos and can send it to me ;)
[06:09] <Kamion> mjg59: I suspect a bit of both :-)
[06:09] <Kamion> mjg59: Mark wanted to arrange for me to have more scary installer-testing facilities, I think
[06:10] <Kamion> the amd is being incredibly useful as a source of entirely trashable disk
[06:11] <Kamion> so I can actually answer yes to "Erase entire disk?" without being terrified
[06:11] <thom> yay, xscreensaver segfaults on amd64
[06:11] <thom> how lovely
[06:12] <Mithrandir> thom: it does?
[06:12] <thom> it just did, but i have no idea what module it was trying to run
[06:12] <thom> most modules seem fine
[06:12] <Mithrandir> blank works very stable. :)
[06:13] <thom> heh
[06:13] <mjg59> thom: 3D failure?
[06:15] <thom> 0: child pid 5066 (noof) exited abnormally
[06:15] <thom> wee!
[06:16] <thom> i wish xscreensaver didn't suck so hard
[06:16] <thom> noof is one of the nice ones, too
[06:16] <daniels> Kamion: you have an amd64?
[06:19] <Mithrandir> daniels: all cool people have amd64s
[06:19] <daniels> Mithrandir: bah
[06:19] <thom> Mithrandir: so just you and me then? ;-)
[06:20] <Mithrandir> thom: yup ;)
[06:25] <Kamion> daniels: not quite yet
[06:27] <daniels> Kamion: rad
[06:42] <Kamion> I think netcfg 1.01ubuntu1 should be a considerable improvement; it doesn't ask me those IP address questions when I'm without a network card, now.
[06:42] <Kamion> I might do an ubuntu2 to make it ask you if you forgot to plug the network cable in
[06:44] <Md> there should be a way to check if there is no link, but I can't remember it...
[06:45] <Kamion> netcfg's already doing that
[06:45] <Kamion> it uses mii-diag
[06:45] <Kamion> right, add "... if mii-diag said the card was unlinked" to the end of what I said
[06:46] <Kamion> I keep being surprised every time another Debian person shows up here :-)
[06:47] <Kamion> "eth0 does not seem to be connected to any sort of network. This could be due to a missing network cable or network adapter malfunction. ..."
[06:47] <Kamion> bingo
[06:48] <daniels> word
[06:53] <mdz> npmccallum: after setting up vesafb, I get: Starting Ubuntu..., delay while initrd runs, nice animation, then some kernel messages overwrite parts of it, then a black screen with scrolling text while the animation continues to play only in the center of the screen
[06:53] <mjg59> Kamion: What does mii-diag output if you're using coax rather than rj45?
[06:53] <mdz> perhaps console= is also needed?
[06:53] <Mithrandir> mjg59: "crack"? :)
[06:53] <mjg59> Actually, I can test this...
[06:54] <mdz> mjg59: you can?  ewww
[06:55] <mjg59> mdz: My network setup partially dates back to the time when I couldn't afford a hub
[06:55] <mjg59> SIOCGMIIPHY on eth0 failed: Operation not supported
[06:55] <mjg59> Haha
[06:55] <Kamion> mjg59: haven't the foggiest idea
[06:55] <npmccallum> mdz: try sticking -- /bin/echo "0 0 0 0" > /proc/sys/kernel/printk -- in mountvirtfs just after the -- domount proc "" /proc -- line
[06:56] <mjg59> Oops - that's entirely the wrong machine
[06:56] <Kamion> new netcfg gives you the option to configure the network statically, anyway (defaulting to "no, just go ahead and ignore the network please")
[06:56] <npmccallum> mdz: also, does your kernel cmdline have console=tty1?
[06:56] <Kamion> if mii-diag fails
[06:57] <mjg59> Ok, it gives me SIOCGMIIPHY on eth0 failed: Operation not supported on an ne2000
[06:57] <npmccallum> mdz: actually, with the console=tty1 option, you may not need the echo line
[06:57] <mjg59> It also gives me that if I run it on a wireless card
[07:04] <npmccallum> mdz: you *do* still need the echo line
[07:07] <mdz> npmccallum: I tried with console=tty1; no difference
[07:08] <npmccallum> mdz: are you getting initscript messages? or kernel messages?
[07:08] <npmccallum> mdz: if kernel, you need to add the echo line I mentioned
[07:08] <mdz> initially, only kernel messages
[07:08] <mdz> then init script messages
[07:09] <npmccallum> mdz: you get initscript messages after init switches to runlevel 2?
[07:09] <mdz> I'm retrying after setting kernel.printk; I'll let you know
[07:10] <mdz> ok, that silenced the messages
[07:10] <mdz> however, the screen is still cleared to a black background
[07:11] <npmccallum> at what point?
[07:11] <mdz> I have no point of reference
[07:11] <mdz> several seconds before gdm starts
[07:11] <npmccallum> this is on an ubuntu system?
[07:11] <mdz> of course
[07:12] <npmccallum> the only thing that can cause that is echoing a clear screen escape command to that terminal
[07:13] <npmccallum> you get white first, then black?
[07:13] <daniels> mdz: this could be to do with vt switching being pure crap on your chipset maybe
[07:14] <mdz> npmccallum: correct
[07:14] <npmccallum> mdz: at what point does gdm load on your system? Sxx in runlevel 2?
[07:14] <mdz> npmccallum: ubuntu default
[07:14] <mdz> this is a fresh install as of yesterday
[07:14] <npmccallum> mdz: which is what?
[07:14] <mdz> sounder 7 + upgrades
[07:15] <mdz> npmccallum: rc2.d/S99
[07:15] <npmccallum> mdz: what happens when you move gdm forward to S01 (that is where it runs on my system)?
[07:16] <npmccallum> mdz: It doesn't use the logger or anything like that, so it doens't actually depend on anything in runlevel2
[07:17] <mdz> npmccallum: no change
[07:17] <npmccallum> mdz: still blanks to black?
[07:18] <mdz> the change to black background still happens before gdm starts
[07:18] <npmccallum> can I get a list of what you have in rcS.d?
[07:18] <mdz> npmccallum: it's a stock ubuntu desktop install
[07:19] <mdz> if I had to guess, I'd say it's probably when the console font is being set
[07:19] <mdz> npmccallum: you have the same laptop that I do; you don't see this?
[07:19] <npmccallum> mdz: I don't
[07:20] <npmccallum> My transitions nicely into gdm
[07:20] <mdz> I think there are just too many variables here for Warty
[07:20] <mdz> let's get it into hoary as soon as it opens up, and do it right from there
[07:20] <npmccallum> ok
[07:21] <mdz> thanks for all of the work you've put into it; I think it's going to be great, but we just need more time than we have available for Warty
[07:23] <npmccallum> no problem
[07:23] <npmccallum> things have to get cut
[07:24] <mdz> hoary should give us plenty of time to sort out choosing modes and initializing the framebuffer, the initrd stuff, etc.
[07:24] <npmccallum> mdz: I was actually looking at rhgb
[07:24] <npmccallum> just to see what they've done
[07:25] <mdz> it's certainly  pretty
[07:25] <npmccallum> they just assume that /usr is mounted
[07:26] <npmccallum> then they just load standard X with -nolock
[07:27] <npmccallum> I was talking with daniels about that we could possibly embed kdrive and statically link gtk, which would give us a lot of freedom
[07:27] <Kamion> would be fun with gtk security updates ...
[07:27] <npmccallum> yeah
[07:28] <npmccallum> or
[07:28] <npmccallum> we can just check for /usr and don't run if its not mounted
[07:28] <Kamion> ok, time to get rid of that ssh-askpass-gnome.desktop
[07:28] <thom> mdz: is totally rad laptop support a feature goal?
[07:28] <npmccallum> we don't have to embed or statically link anything in that case
[07:29] <npmccallum> having /usr on the root fs is a fairly safe requirement for desktop installs
[07:29] <mdz> thom: depends on your next question :-)
[07:29] <Kamion> or run in some degraded mode that doesn't include gtk
[07:29] <Kamion> dunno, I think you might be surprised at how much even some desktop users like to customize their partition layout
[07:29] <Kamion> I know we don't default to that, but ...
[07:29] <mdz> couldn't we use something like the tinyx server if we used an rhgb approach?
[07:29] <thom> mdz: netapplet is utterly rad...
[07:30] <npmccallum> mdz: yes, we could... does tinyx run on vesa?
[07:30] <daniels> mdz: uhm, tinyx is pre-kdrive
[07:31] <daniels> kdrive is a renamed, and slightly forked, tinyx
[07:31] <daniels> by forcing several features out, you can make kdrive really small
[07:31] <mdz> npmccallum: it uses fbdev, at least the way I've built it in the past
[07:31] <daniels> so we could have a kdrive-tiny, or whatever
[07:31] <mdz> it doesn't need to be particularly tiny, as long as it doesn't require stuff like /usr
[07:31] <npmccallum> if we do use x/gtk we could have status icons pop up, and if an error occurs, click on the icon for text :)
[07:31] <mdz> tinyx could be built to run basically standalone, with no config file, modules, etc.
[07:31] <mdz> which is what we would want
[07:32] <npmccallum> right
[07:32] <npmccallum> mdz: why are we making all kinds of exceptions for non root /usr?  just make it a requirement...
[07:32] <npmccallum> if /usr isn't on the root fs, don't run
[07:33] <mdz> npmccallum: well, we also want it to be simple and start up very quickly
[07:33] <npmccallum> mdz: agreed
[07:33] <npmccallum> that would be one of my feature goals
[07:33] <mdz> tinyx starts up in about 1 second on my Zaurus (200MHz ARM)
[07:33] <thom> mdz: http://tech9.net/rml/log/2004082401
[07:34] <mdz> thom: looks like a totally rad hoary feature :-)
[07:34] <npmccallum> daniels: how does kdrive rate against a one second startup?
[07:34] <thom> mdz: and indeed http://www.nat.org/2004/august/
[07:35] <thom> mdz: i *knew* you were gonna say that
[07:35] <thom> :-)
[07:35] <mdz> f-spot is one of those scary mono things, right?
[07:35] <Kamion> yes
[07:36] <thom> yep, ignoring f-spot, even though it is rad with a k
[07:36] <Kamion> is mono still using binfmt-support?
[07:36] <Kamion> I'll be unhappy if not, since I implemented some quite complicated stuff just for them :)
[07:36] <mjg59> It doesn't have a lot of choice
[07:36] <Kamion> maybe I should make that bit be shell rather than perl, though
[07:36] <daniels> npmccallum: not quite sure, but it's pretty damn quick. most of it is, of course, in the mode switch :\
[07:37] <mjg59> Kamion: Yeah, mono-common depends on binfmt-support
[07:37] <npmccallum> daniels: can we make kdrive use vesa without a mode switch (ie. if use the current mode)?
[07:37] <Kamion> excellent
[07:37] <npmccallum> daniels: that is what xdirectfb does and its fast
[07:38] <mdz> daniels: well, if the framebuffer is already initialized, then there should be no mode switch
[07:38] <daniels> npmccallum: maybe it already does that
[07:38] <daniels> i'll check it out when i get back
[07:38] <daniels> gotta run for a bit now
[07:44] <Keybuk> thom: netapplet disappointed me somewhat
[07:45] <sladen> ?
[07:45] <thom> Keybuk: how so?
[07:46] <thom> it's pretty frickin' awesome
[07:46] <Keybuk> well, it's just a not-applet for switching the active interface
[07:46] <Keybuk> the way it was hyped to me I was expecting HAL integration of local wireless networks and other fun things
[07:47] <thom> i'm guessing that's the end game
[07:47] <thom> hal doesn't do wireless at all right now iirc
[07:47] <Keybuk> indeed not
[07:48] <Keybuk> I've been kinda thinking about the non-wireless side of wifid as a result though
[07:48] <Keybuk> would be cool to use netapplet as a UI for wifid
[07:50] <daniels> last i checked, hal not doing wireless was NOTABUG
[07:50] <daniels> there was a large discussion about it on hal@ a while back, during which it was concluded that most wireless stuff was out of scope, iirc
[07:51] <SteveA> how do I get flash plugins working in warty?
[07:51] <mdz> SteveA: apt-get install swf-player
[07:52] <Keybuk> don't see what's wrong with it, I kinda like the idea of seeing link strength of adjacent networks through hal
[07:52] <mdz> in firefox anyway
[07:52] <Keybuk> makes writing applets so much easier :p
[07:52] <thom> gar! i wish xscreensaver didn't suck quite so hard
[07:52] <daniels> Keybuk: heh
[07:52] <daniels> thom: still?
[07:52] <Kamion> SteveA: (that's in universe)
[07:52] <SteveA> mdz: thanks.  I was searching in synaptic for names with "flash" in, not "swf"
[07:52] <mjg59> swf-player is barely a flash plugin...
[07:53] <thom> or, in this case, test that fortune exists before trying to run it, and if not, move on
[07:53] <Kamion> or we could install fortune ...
[07:53] <Kamion> xscreensaver is sitting permuting "No such file or directory" around the amd's screen at the moment
[07:53] <thom> Kamion: well, that would work, but a bit of robustness is never a bad thing :-)
[07:53] <daniels> Keybuk: maybe this thread: http://freedesktop.org/pipermail/hal/2004-May/000170.html
[07:54] <daniels> thom: HEH
[07:54] <mdz> Kamion: yes
[07:54] <mdz> in fact, let's take care of that now, because it keeps coming up and I keep forgetting about it
[07:54] <mdz> my only issue with fortune is that it has a lot of shitty fortunes
[07:55] <mdz> shall we just clean it out and add it to desktop?
[07:55] <Kamion> ubuntu-fortunes? :-)
[07:55] <thom> as long as there is NO mention of ssl certs :P
[07:55] <Kamion> we can have fortune-mod without all the actual data, can't we?
[07:55] <sladen> Kamion: is that the xscreensaver-fortunes when fortune isnt installed?
[07:55] <Kamion> sladen: yes
[07:55] <Mithrandir> thom: please, just a small ssl cert fortune?
[07:55] <Mithrandir> ;)
[07:55] <thom> Mithrandir: NO!
[07:55] <Keybuk> daniels: yeah, read that one
[07:55] <thom> ;-P
[07:56] <thom> yay, bumbps terminated with SIGSEGV
[07:56] <Keybuk> setting the essid through HAL is silly because should HAL then kill dhclient and restore the static IP settings needed for *that* essid?
[07:56] <Keybuk> besideswhich, iwlib sucks
[07:56] <mdz> Kamion: why is it named 'fortune-mod' anyway?
[07:56] <Kamion> mdz: I've often wondered that
[07:57] <Kamion> "modified" from BSD?
[07:57] <mdz> Kamion: come to look, fortunes-min is actually pretty reasonable
[07:57] <mdz> it's 'fortunes' that has all the crap
[07:57] <Kamion> wow, the jumping cow screensaver is stunning
[07:57] <Kamion> we should so do a jumping warthog
[07:58] <thom> bouncing cow is genius
[07:59] <mdz> fortune-mod added to desktop
[08:00] <Keybuk> talking of which ... hal doesn't work on warty
[08:01] <mdz> considering that's entirely untrue, perhaps you can be a bit more specific?
[08:01] <Keybuk> starts then exits
[08:01] <Keybuk> trying to figure out why atm
[08:02] <npmccallum> Keybuk: I have the same problem
[08:02] <npmccallum> Keybuk: though, I installed debian's kernel-image-2.6.8 and it seems to work, go figure
[08:02] <Kamion> seems to be working for me
[08:03] <Kamion> are we going to take 2.6.8?
[08:03] <mdz> works for me
[08:03] <mdz> Kamion: quite likely
[08:03] <mdz> considering 2.6.7 has security issues
[08:03] <Keybuk> aha
[08:03] <Keybuk> it's core dumping
[08:04] <Kamion> mdz: s'what I thought
[08:04] <mdz> hal logs a lot of garbage to syslog, though
[08:04] <mdz> Aug 25 11:03:52 localhost hald: [W]  linux/osspec.c:700 handle_udev_node_created_found_device() : No HAL device corresponding to device file /dev/vcs2 
[08:04] <Keybuk> mdz: what kernel are you running?
[08:04] <doko> why isn't 2.6.8.1 not packaged yet?
[08:04] <mdz> Keybuk: stock Warty
[08:05] <npmccallum> mdz: Is there a reason we haven't sync'd to 2.6.8 yet?
[08:05] <mdz> npmccallum: yes, we're expecting a custom package
[08:05] <npmccallum> ooh :)
[08:07] <Keybuk> syndicate hal# hald --daemon=no
[08:07] <Keybuk> (process:5222): GLib-GObject-WARNING **: instance with invalid (NULL) class pointer
[08:07] <Keybuk> (process:5222): GLib-GObject-CRITICAL **: file gsignal.c: line 2121 (g_signal_emit_valist): assertion `G_TYPE_CHECK_INSTANCE (instance)' failed
[08:07] <Keybuk>   *more splurge*
[08:07] <Keybuk> zsh: segmentation fault (core dumped)  hald --daemon=no
[08:15] <Keybuk> npmccallum: what wifi card have you got in that thinkpad of yours?
[08:15] <sladen> Keybuk: atheros
[08:15] <Keybuk> heh, snap
[08:16] <Keybuk> if I tear down the card and modules hald starts
[08:16] <Keybuk> as soon as I bring it back up, hald core dumps
[08:16] <sladen> whereas /me has an orinoco :)
[08:16] <npmccallum> Keybuk: madwifi
[08:16] <Keybuk> Aug 25 19:16:44 syndicate hald: [E]  linux/net_class_device.c:138 mdio_read() : SIOCGMIIREG on ath0 failed: Cannot allocate memory
[08:16] <npmccallum> Keybuk: oh, I know this bug
[08:16] <Keybuk> didn't I see that in scrollback?
[08:16] <npmccallum> Keybuk: its fixed now upstream
[08:17] <npmccallum> Keybuk: I made a patch for it one time a while back
[08:17] <npmccallum> Keybuk: let me get the patch
[08:20] <npmccallum> Keybuk: http://bugs.breakmygentoo.net/attachment.cgi?id=327&action=view
[08:20] <Keybuk> that's somewhat cheating patchwise, isn't it? <g>
[08:20] <Keybuk> "uncomment the code that broke"
[08:20] <Keybuk> uh, comment
[08:21] <npmccallum> Keybuk: no, the code was removed upstream a few days after my patch
[08:22] <npmccallum> Keybuk: they decided to roll back the wifi stuff
[08:22] <Kamion> patches should generally remove code rather than comment it out, I feel - history is what revision control systems are for :)
[08:22] <npmccallum> fine, then find the line and remove it :)
[08:22] <Kamion> not that I can talk considering the state of /var/lib/dpkg/info/passwd.config at the moment
[08:23] <Keybuk> ah, the debian code has if (!is_80211) around that
[08:23] <Kamion> heaving mass of comments
[08:25] <Keybuk> if we remove that line, then normal network cards won't have link state in hal?
[08:36] <Keybuk> hm, ok; hal won't build
[08:40] <npmccallum> Keybuk: we could always get the newest dbus and hal (which may not actually be a bad idea)
[08:42] <Keybuk> yeah, the new stuff seems to work; but that's up to mdz
[08:43] <mdz> which actual bugs does it fix?
[08:44] <npmccallum> mdz: most of the wifi stuff was taken out of the newer hals
[08:44] <Keybuk> can't start hald (and thus g-v-m) on machines with atheros cards
[08:44] <npmccallum> mdz: the wifi stuff had lots of problems with lots of wifi cards (not just atheros)
[08:45] <mdz> before the conference, I was using an atheros card, and hal was starting OK
[08:45] <npmccallum> mdz: we seem to be using an older version which still has the old wifi stuff in hal
[08:45] <npmccallum> mdz: yeah, its an intermittant problem
[08:45] <npmccallum> mdz: I was fine for a while too
[08:47] <npmccallum> if we do upgrade hal, I believe we have to upgrade dbus as well
[08:48] <thom> yep
[08:56] <Keybuk> thom: didn't you remove "Suspend" from the battery applet?
[08:57] <thom> yeah, i've not uploaded it yet
[08:57] <Keybuk> ah, ok
[08:57] <Keybuk> thought seb had fumbled a patch there for a second :p
[09:15] <edd> seb does not fumble :)
[09:16] <mdz> thom: how does the mozilla situation look?
[09:17] <Oskuro> seb128: what's wrong with the suspend bit?
[09:17] <seb128> Oskuro: what ?
[09:17] <Oskuro> 20:56 < Keybuk> thom: didn't you remove "Suspend" from the battery applet? 20:57 < thom> yeah, i've not uploaded it yet
[09:18] <Oskuro> I wish gnome-terminal didn't select an arbitrary number of lines when I triple click
[09:18] <edd> it calls apm, iirc
[09:18] <edd> last i knew, ubuntu doesn't have apm
[09:19] <Oskuro> oh
[09:23] <fabbione> lamont: can you give it a kick later tonight_?
[09:47] <thom> mdz: i got caught up in ddcprobe/xresprobe on amd64; just cooking dinner then will look
[10:13] <mdz> Oskuro: not only that, suspending with apm requires root and there was no provision for that
[10:13] <mdz> we'll do something useful with it in the future when it can be done correctly
[10:17] <mdz> thom: I don't seem to have notes about your schedule; what hours are generally good for you as far as meetings and the like?
[10:18] <thom> mdz: i'm usually awake and working 10am-6pm (currently GMT+1)
[10:18] <mdz> ok, thanks
[10:18] <thom> but the latter tends to stretch later most of the time
[10:25] <mdz> npmccallum: how about you?  what are your normal waking/working hours?
[10:28] <thom> mdz/Kamion: thoughts on tmpfs for /tmp ?
[10:28] <mdz> thom: I use it everywhere
[10:28] <thom> mdz: should we do it by default?
[10:29] <mdz> possibly; float the idea on the sounder list?
[10:29] <thom> yeah, will do. was just wondering if there was a previously discussed reason we don't that i'd forgotten
[10:29] <mdz> I don't think so; I'd been meaning to bring it up myself