[12:02] <jbailey> jdub: I'm on a ppc, need something done>
[12:02] <jdub> jbailey: http://people.ubuntu.com/~jdub/hoary/
[12:02] <jdub> dholbach: no
[12:03] <jbailey> jdub: How do I test it when I build it?
[12:04] <jdub> mplayer -ao esd http://mirror.fluendo.com:8800/
[12:04] <jbailey> We have mplayer?  Cool.
[12:04] <jdub> well
[12:04] <jdub> use totem :)
[12:05] <jbailey> Will totem-xine pick it up, or should I switch back to totem-gstreamer?
[12:06] <jdub> totem xine will be fine
[12:06] <jdub> meanwhile, i am sad
[12:06] <jdub> the latest gamin/inotify patch still does not work
[12:06] <mdz> jdub: by "should just work", did you mean "is implemented", or "ought to be implemented"?
[12:06] <jdub> but less spectacularly so this time
[12:07] <jdub> mdz: should be is, but ought to be if not
[12:07] <jdub> wouldn't be a hairy fix, given that it uses the libbacon stuff
[12:10] <jdub> WOO!
[12:10] <jdub> Setting up language-pack-zh-update (20050128) ...
[12:10] <jdub> Setting up language-pack-zh (20050128) ...
[12:10] <jdub> Generating locales...
[12:10] <jdub>   en_AU.UTF-8... done
[12:10] <jdub>   zh_CN.UTF-8... done
[12:10] <jdub>   zh_HK.UTF-8... done
[12:10] <jdub>   zh_TW.UTF-8... done
[12:10] <jdub> Generation complete.
[12:10] <jdub> 
[12:10] <jdub> :-)
[12:11] <jbailey> China gets Australian English? 
[12:12] <jbailey> jdub: checking for sdl-config... no
[12:12] <jbailey> jdub: Bad build-dep?
[12:13] <jdub> jbailey: should say it doesn't care about sdl too
[12:13] <jbailey> Yeah, just making sure you ahadn't missed something.
[12:13] <jbailey> FAils to link.  Want me to /msg you the failure?
[12:13] <jdub> nah
[12:13] <jdub> that's okay
[12:13] <jbailey> Something has a bunch of symbols with i386 hardcoded in them.
[12:14] <jdub> yeah
[12:22] <YokoZar> my0 kqweyboasrdasg0 juasdt0 gbrokqwe0 iasd0 thweqrqweg0 san 0on0 asdzxcerqweqwen0 kqweyboasrsda
[12:27] <bluefoxicy> ping
[12:47] <mdz> elmo: can you byhand the d-i build I'm about to do?
[01:01] <T-None> huh. FtC has gone down
[01:25] <lamont> mdz
[01:25] <lamont> around?
[01:28] <lamont> "1027x768 @ 61hz will make me go bonkers, and I'm sure my hardware supports better."
[01:28] <lamont> go warty!!
[01:30] <ogra> lamont: at least better the 800x600 @ 50Hz
[01:30] <lamont> yeah
[01:30] <mdz> lamont: yes
[01:37] <lamont> "Well, that was the least-effort Linux install I've ever done (even a base Debian install for requires a little bit more, like skipping past dsesect, etc)."
[01:37] <mdz> source?
[01:38] <lamont> former co-worker
[01:38] <lamont> he got a warty CD set today, I believe
[01:38] <lamont> he's also the source on that 1027x768 comment...
[01:38] <lamont> so I'm asking for xresprobe and ddcprobe output..
[01:39] <mdz> lamont: can you confirm that localechooser 0.04.0ubuntu2 made it into d-i- 20041227ubuntu7.0.200501281 ?
[01:39] <mdz> (initrd.list)
[01:39] <lamont> mdz: any particular arch, or do I need to check all 3 or 4?
[01:39] <mdz> the I'll mail elmo and Kamion so that they can finish up when they come back around
[01:39] <mdz> lamont: localechooser is arch: all, so if one has it, they all should
[01:40] <dholbach> ogra: :-)
[01:40] <ogra> dholbach: yay....
[01:41] <ogra> dholbach: suse finally goes for CMS after years of html css fiddling
[01:42] <lamont> mdz: yep.
[01:43] <lamont> Get:23 http://jackass.ubuntu.com hoary/main/debian-installer localechooser 0.04.0ubuntu2 [61.1kB] 
[01:43] <dholbach> ogra: i can fully understand that :-)
[01:43] <lamont> mdz: as it currently sits, the livecd rootfs and debian-installer daily builds happen at the same time on diff machines.  do we need to shuffle things around at all? (/me thinks not...)
[01:44] <mdz> lamont: thanks
[01:44] <mdz> rootfs. vs. d-i doesn't matter, as long as ISO builds happen after both
[01:44] <mdz> not that the d-i build goes into the archive right away anyway...
[01:44] <lamont> my builds happen at 0615, and take ~30-40 minutes.  the ISO builds happen (IIRC) at 0800, so we're golden
[01:45] <lamont> although given elmo's schedule, I rather expect that we're always building the ISO with the prior day's d-i build... :-)
[01:45] <mdz> yeah
[01:48] <lamont> does anyone know why my machine thinks it's fqdn is 'localhost.localdomain'?
[01:49] <ogra> lamont: sounds like debian...
[01:50] <dholbach> lamont: mine did too
[01:53] <lamont> hehehe
[01:53] <lamont> dholbach: what exactly do you have in /etc/hostname?
[01:54] <lamont> it doesn't like to have an FQDN in /etc/hostname
[01:54] <lamont> whic his a bug, in my opinion...
[01:54] <dholbach> lamont: now is says "bert" :-)
[01:55] <lamont> hrm..
[01:55] <lamont> iirc, the code says that when it can't figure out what to say
[01:55] <dholbach> lamont: i fixed it quite a while ago
[02:02] <jordi> Treenaks: heh, I just noticed you were the initial maintainer for gtranslaor
[02:02] <jordi> gtranslator even
[02:16] <dholbach> binary-or-shlib-defines-rpath  is finishing me off :-/
[02:20] <thom> jdub: claims on the dashboard list we need a newer gsf-sharp
[02:29] <dholbach> ok... 1:0 to lintian... but i'll re-appoint the match :-)
[02:30] <dholbach> good night everyone... i'm off
[02:35] <zul> hey
[03:14] <eruin> why oh why are we receiving a broken firefox with the GNOME icon theme in hoary?
[03:15] <thom> broken how?
[03:16] <eruin> a restart was in order
[03:16] <eruin> but are the gnome icons there to stay? they aren't exactly as snazzy as the default ones
[03:17] <thom> no, but they are consistent
[03:17] <thom> point being it actually looks and feels like it's meant to be on the desktop, rather than just bolted on after
[03:18] <bluefoxicy> has anyone seen pitti lately?
[03:18] <eruin> I thought the whole idea behind the new firefox theme was to make it look integrated in both windows, gnome and osx ?
[03:19] <schweeb> eruin: the icons match the ubuntu desktop
[03:19] <eruin> that theme sure didn't look bolted on, but I'm sure you guys have already had a long discussion about it
[03:19] <schweeb> eruin: if you don't like the theme, you're able to use one of your choosing
[03:19] <thom> it's not the same theme on windows and osx anyway
[03:19] <thom> but, shrug
[03:20] <thom> we'll let people say what they think; if there's consensus that we shouldn't do it, odds are we won't
[03:20] <thom> as it stands i think it looks a hell of a lot better
[03:21] <eruin> atleast provide the standard ff theme in addition
[03:21] <eruin> that theme isn't available for download anywhere, and the theme manager in ff still says firefox (default) by gerich
[03:21] <schweeb> thom: man, gnome file dialogues now, that's nice
[03:21] <eruin> yeah, that's great stuff :)
[03:22] <thom> eruin: patches cheerfully accepted
[03:25] <schweeb> thom: and complaints cheerfully ignored, eh :p
[03:28] <thom> schweeb: not at all
[03:28] <eruin> I doubt my php/html/css skills will be very useful in that regard
[03:29] <thom> schweeb: i have a list of bugs as long as my arm; if something is not a priority i'll do it if i can but if someone gets there first great
[03:29] <thom> complaints without patches just rank lower than those with :-)
[03:29] <eruin> :-)
[03:29] <elmo> mdz: done
[03:32] <bluefoxicy> !seen pitti
[03:32] <bluefoxicy> guess not
[03:38] <zul> thom: i thought that sudo patch wouldnt work properly
[03:39] <jbailey> Wow.  nfs root actually works.  Creepy.
[03:39] <thom> zul: yeah, it's unfortunate that it's not that easy
[03:39] <thom> but it's fixable for new installs; old ones will just have to cope
[03:40] <zul> yep
[03:41] <thom> g'night
[03:42] <zul> night
[03:44] <mdz> elmo: thanks
[03:56] <zul> goody...i have slmodem almost working in linux-restricted
[04:39] <mdz> lamont: have you tried hoary-live-ia64.iso yet?
[04:50] <zul> night..
[05:11] <thully> I just noticed a significant bug in the live CD's network configuration process
[05:12] <thully> I have a laptop w/ethernet and wireless - if i'm not connected to ethernet and not in wireless range, the CD keeps prompting me for a WEP key until I back out and skip the network step in debian-installer
[05:12] <thully> that causes the loopback interfance not to be configured, and things not to work right as I reported in a previous bug
[05:34] <thully> hi - one quick question - can I label a live CD bug "critical" or "blocker" if it is critical or blocker for the live CD release and not for the standard release?
[05:36] <thully> As this issue with wi-fi on the live CD seems critical for the live CD - we don't want to be shipping a live CD which doesn't work right without a network
[06:02] <farruinn> does this sound normal: I'm building gaim from ubuntu source package, I ran dpkg-buildpackage -rfakeroot, and it goes through the ./config process but when that finishes I get a message saying automake-1.8 is too old. It's not a fatal error though because ./configure runs again and apparently successfully because now it's compiling...
[06:03] <farruinn> I did sudo apt-get build-dep gaim before doing this then double checked that everything was up to date, and it is
[06:07] <syn-ack> farruinn: sounds like a bug in automake, or maybe a bug in the .config file...
[06:08] <farruinn> this is warty btw
[06:08] <farruinn> no backports or anything unusual
[06:08] <syn-ack> My question is this.... why build if its in the repo?
[06:09] <syn-ack> still it more than likely is a shoddy .config 
[06:09] <syn-ack> or makefile.
[06:10] <farruinn> guifications needs gaim.pc which debian/rules deletes after install
[06:10] <crimsun> hmm. Isn't that packaged in gaim-dev?
[06:11] <infinity> farruinn : Sounds like timestamp skew to me.
[06:11] <syn-ack> infinity: could be that too
[06:11] <infinity> farruinn : See /usr/share/doc/autotools-dev/README.Debian
[06:11] <farruinn> crimsun: I did apt-cache search gaim-dev but didn't find anything
[06:12] <infinity> That'd be because there isn't a gaim-dev.
[06:12] <infinity> Perhaps there should be. :)
[06:12] <syn-ack> aye, there should..
[06:13] <crimsun> right, it's in hoary/universe, and farruinn's using warty.
[06:15] <syn-ack> I think Im going to build a .deb of gaim 1.1.2 and check it out. heh
[06:16] <infinity> Ahh, I see.  gaim-dev is brand new.
[06:17] <farruinn> apt-get source gets exactly the same source that was used to build the binaries available through apt, right?
[06:17] <infinity> No.
[06:18] <infinity> apt-get source will get the most recent source listed in your deb-src lines.
[06:18] <crimsun> farruinn: correct [if you use strictly the warty repo] 
[06:18] <lamont> mdz: download speed has been rotten today - planning to make another run at it tomorrow.
[06:18] <crimsun> (btw, you can force apt-get source gaim=someversion)
[06:18] <lamont> I know that someone tried it (ia64live) today, and had no happiness on that particular machine - didn't detect the CDROM
[06:36] <lamont> firefox-bin(8312): unaligned access to 0x0000000042a86c74 at ip=0x0000000042a31b9b
[06:36] <lamont> firefox-bin(8312): unaligned access to 0x0000000042a86c7c at ip=0x0000000042a31bbb
[06:36] <lamont> firefox-bin(8312): unaligned access to 0x0000000042a86c84 at ip=0x0000000042a31bcf
[06:36] <lamont> bad firefox
[06:42] <lamont> Updating mozilla-firefox chrome registry...E: Registration process existed with status: 1
[06:42] <lamont> E: /usr/lib/mozilla-firefox/extensions/installed-extensions.txt still present. Registration might have gone wrong.
[06:46] <YokoZar> Would it be wrong to have a script in /usr/share/package and then symlink it in /usr/bin ?
[06:47] <syn-ack> that sounds wrong to me. What would be your purpose for that?
[06:49] <YokoZar> I'm trying to make a package around what is basically glorified Xdialog scripts, and I'm not sure where to put them
[06:49] <YokoZar> http://www.von-thadden.de/Joachim/WineTools/
[07:02] <YokoZar> syn-ack: Where should such scripts go?
[07:02] <lamont> --- SIGRTMIN (Unknown signal 37) @ 0 (0) ---
[07:02] <farruinn> would deb-make give me a debianized source suitable for ubuntu?
[07:02] <lamont> interesting
[07:03] <YokoZar> Oh I see share is only for read only stuff
[07:03] <syn-ack> YokoZar: I personally just place them in /bin
[07:03] <YokoZar> Yeah I think that's the right FHS way
[07:03] <syn-ack> or /usr/bin/
[07:03] <YokoZar> (reading it now)
[07:03] <YokoZar>  /usr/bin is right ;)
[07:04] <syn-ack> heh
[07:29] <farruinn> dpkg-buildpackage shouldn't touch anything outside the source dir, no?
[07:29] <farruinn> I was running it on guifications source and it tried this: "mkdir -p -- /usr/share/pixmaps/gaim/guifications/conf"
[07:31] <YokoZar> hmmm...sounds like a bad behaving package
[07:32] <farruinn> well, I debianized it myself so it's not surprising
[07:32] <farruinn> I've read the debian new maintainers guide, but I don't know what to do about this
[07:32] <YokoZar> Directories should only be made by dh_installdirs when it reads a file
[07:33] <Mithrandir> mdz: pong
[07:34] <YokoZar> farruinn: you need some files like (package.dirs) that have the directories you need made for you in them
[07:34] <YokoZar> then dh_installdirs will create them when it gets to that part in the rules file
[07:34] <farruinn> in debian/tmp, right?
[07:34] <YokoZar> Yeah
[07:34] <farruinn> excellent, thanks =)
[07:39] <syn-ack> hrm
[07:47] <YokoZar> Is there a standard for what goes in /usr/share/pixmaps?
[07:54] <sivang> morning all
[07:59] <Mithrandir> heh, I've pushed out about 40GB of live CDs since last afternoon.
[08:00] <zenrox> who here can help in #ubuntu have a prob with a user
[08:01] <zenrox> he pasted a long past and i had ask him to please use pastbin.com  and then he just put me on ignore
[08:02] <zenrox> witch is ok
[08:02] <zenrox> but the long past was wrong
[08:03] <YokoZar> What does dh_installexamples do?
[08:03] <Mithrandir> zenrox: if it was a one-time thing, ignore it.
[08:04] <crimsun> YokoZar: see the appropriate debhelper man page ;)
[08:04] <Mithrandir> zenrox: if he continues to be obnoxious, see if mdz is around.
[08:04] <crimsun> YokoZar: essentially, it installs [a hierarchy of]  example file
[08:04] <zenrox> Mithrandir,  ok
[08:04] <YokoZar> crimsun: I guess my question is why it's uncommented by default when I made my package
[08:05] <crimsun> YokoZar: a great many of the debhelper templates are uncommented by default :)
[08:06] <YokoZar> Also should I remove the changelog file from the docs file?  It was put in there automatically
[08:06] <YokoZar> As it's handled instead by dh_installchangelogs
[08:07] <crimsun> yeah, I never have changelog in .docs
[08:09] <YokoZar> Also if I remove a file from a source package will I have to remove it again when I do uupdate?
[08:10] <crimsun> when you run dpkg-buildpackage, it generates a .diff.gz, which is applied to the new upstream version when you run uupdate
[08:10] <Mithrandir> but it ignores deletions.
[08:10] <crimsun> right
[08:10] <Mithrandir> YokoZar: why do you remove the file from the package?
[08:11] <YokoZar> Mithrandir: It was a binary version of Xdialog (instead I depend on xdialog package and use that)
[08:11] <Mithrandir> YokoZar: ok.  You could ask upstream to remove it, though.
[08:12] <Mithrandir> and you can just remove it in the clean target and ignore the warning from dpkg-source
[08:12] <YokoZar> That could work but it would make the package needlessly larger, heh
[08:12] <YokoZar> Well, the source package
[08:13] <Mithrandir> you need to remove it from the original .tar.gz if you don't want the bloat there
[08:13] <fabbione> wow
[08:13] <YokoZar> Oh yeah I see
[08:13] <YokoZar> Ok, putting a thing into dh_clean
[08:13] <fabbione> i slept almost 24 hours
[08:13] <Mithrandir> hi fabbione
[08:13] <fabbione> hi Mith
[08:13] <daniels> fabbione: awesome :) 'morning
[08:13] <YokoZar> People thought you were dead fabbione
[08:13] <YokoZar> Although, in a way, you were.
[08:13] <crimsun> 24 is superhuman :)
[08:13] <fabbione> daniels: there is nothing awesome in having 39/40 of fever :(
[08:13] <Mithrandir> fabbione: _ew_
[08:14] <Mithrandir> YokoZar: just rm -f path/to/xdialog in the clean target should be ok
[08:14] <fabbione> i got a big flu i think
[08:14] <fabbione> anything important happened?
[08:14] <daniels> fabbione: ouch :( that's a horror
[08:15] <daniels> fabbione: fglrx is totally brken, but news at 11
[08:15] <fabbione> enocare
[08:15] <fabbione> :)
[08:15] <daniels> right :)
[08:15] <YokoZar> Thanks Mithrandir
[08:16] <fabbione> i think i am back to bed
[08:17] <crimsun> ciao
[08:17] <YokoZar> Sleep well.  We'll boil some chicken hearts and newt eyes to help you get well.
[08:17] <Mithrandir> fabbione: get well
[08:17] <fabbione> eheh
[08:17] <fabbione> thanks guys
[08:17] <fabbione> i might see you on monday
[08:19] <sivang> YokoZar: this is some kind of chineese tradisiont? ;-)
[08:19] <YokoZar> sivang: Voodoo, actually.
[08:22] <sivang> YokoZar: hehe , ok :)
[08:34] <dholbach> morning
[08:55] <sivang> dholbach: morning
[08:59] <YokoZar> Is there a package out there that doesn't actually compile anything (ie: just copies files) that I can look at?
[09:00] <sivang> YokoZar: if you want to achive that, you need to use dh_install
[09:00] <YokoZar> I'm trying to figure out a good rules file that doesn't confuse the hell out of debhelper
[09:00] <Mithrandir> YokoZar: don't use debhelper for the install part there, then.  Just cp stuff into the correct directories in debian.
[09:00] <sivang> YokoZar: and in the /debian folder, you need to have a file named <pkg-name>.install and list all the files you want to install and to where.
[09:01] <YokoZar> Well I have a bunch of cp stuff...the trick is figuring out the correct directories
[09:01] <sivang> YokoZar: well, what are you trying to copy?
[09:01] <YokoZar> At that part of the rules file it looks like I should be putting them into debian/tmp/usr/bin/ and such
[09:01] <YokoZar> but there is no tmp directory
[09:01] <sivang> tmp is created IIRC Only when dpkg attempts at building the sources
[09:02] <sivang> (or the stuff needs built in the pkg)
[09:02] <YokoZar> Or, rather, it isn't created by buildpackage then
[09:02] <YokoZar> Since I have the thing commented out I guess
[09:02] <Mithrandir> YokoZar: make it, either with mkdir or dh_installdirs.
[09:02] <YokoZar> And copy the files into the tmp folder then?
[09:02] <Mithrandir> yeah, or debian/$packagename/$wherever.
[09:03] <YokoZar> Can I have dh_installdirs do it and prepend tmp/ to the front?
[09:03] <dholbach> hai sivang :-)
[09:03] <sivang> YokoZar: no, you can leave the files inside the /debian folder, use dh_installdirs to create the dirs you need, and dh_install to read your pkgname.install file and copy files where needed.
[09:03] <sivang> dholbach: yo danile, 'sup?
[09:03] <Mithrandir> YokoZar: dh_installdirs knows where it should make the directories, so don't worry about debian/tmp and such.
[09:04] <YokoZar> Well what it's currently doing is making debian/package
[09:04] <YokoZar> And I can copy my files into package/usr/whatever
[09:04] <Mithrandir> sivang: dh_install installs the files from the tmp tree into the right directories, it doesn't copy them from the source package.
[09:04] <Mithrandir> YokoZar: yes
[09:04] <YokoZar> Then when it gets to dh_install -s I get this error:
[09:04] <YokoZar> cp: cannot stat `./usr/bin': No such file or directory
[09:04] <dholbach> sivang: i'm trying to package a more recent version of glibmm, but i stumble over binary-or-shlib-defines-rpath
[09:05] <Mithrandir> YokoZar: don't use dh_install for what you are trying to do, it's the wrong tool.
[09:05] <YokoZar> Oh ok then
[09:05] <dholbach> so is anyone of you familiar with  binary-or-shlib-defines-rpath  and  -rpath  being used by Makefiles? there's a huge thread (at http://lists.debian.org/debian-mentors/2002/02/msg00042.html) and people seem to me not to be sure if such a lintian warning was worth messing around with a proper compilation/installation of a library
[09:06] <Mithrandir> YokoZar: this is fairly basic stuff -- have you read debhelper(7) and the debian new maintainer's guide?
[09:06] <YokoZar> Yeah I have I've even made the wine packages...I guess by not knowing exactly when dh_install was getting called
[09:07] <Mithrandir> dholbach: are you using DESTDIR when installing?
[09:07] <sivang> Mithrandir: it can be used for a couple of file installment no?
[09:08] <sivang> Mithrandir: (would like to know if I am using the wrong tool ;-)
[09:08] <dholbach> Mithrandir: yes, it's in most of the Makefiles
[09:09] <YokoZar> ok I commented out dh_install
[09:09] <sivang> Mithrandir: it allows me to add how many files that I would like on the .install file, and relieves me from acutally changing the make rules in debian/rules...
[09:09] <Mithrandir> sivang: it seems you can use it both ways.
[09:10] <sivang> Mithrandir: could you elobrate? I am not sure I have followed you...
[09:10] <sivang> *elaborate
[09:10] <YokoZar> That's odd it didn't ask for my key or anything but spat out a .deb file
[09:10] <Mithrandir> sivang: read the man page rather than listening to me. :)
[09:10] <sivang> Mithrandir: okok, mang pages were cool with dh_install so, nm ;-))
[09:12] <Mithrandir> dholbach: hmm, I seem to have participated in that thread. :)
[09:13] <dholbach> Mithrandir: yes... and you (as infitiny@#debian-devel) told me something about chrpath, but i'm not sure, if that's of anyuse
[09:15] <Mithrandir> dholbach: you could just do chrpath -d $binaryname
[09:15] <Mithrandir> but I wonder why it gets set; can you put your package online somehwere?
[09:15] <Mithrandir> s/somehw/somewh/
[09:16] <dholbach> Mithrandir: seems to me that a standard Makefile (for a library) spitted out by auto* uses -rpath
[09:17] <dholbach> at least it also did in a library, i coded myself
[09:17] <Mithrandir> not necessarily.
[09:18] <dholbach> Mithrandir: in the current state the package doesnt build because the linker is confused :-)
[09:19] <Mithrandir> ok, tell me when you have unconfused it and I can take a look, then
[09:19] <dholbach> Mithrandir: brilliant :-)
[09:22] <sivang> Mithrandir: is there a counterpart for dh_install ? I want to be able to remove the files my package put so I won't get a dpkg error when it tries to remove the dir
[09:22] <dholbach> erm... cool :-)
[09:23] <dholbach> :-D
[09:23] <dholbach> libtoolize --force --copy   and the    debian/rules-libtool-is-a-fool--hack   did the trick
[09:24] <Mithrandir> dholbach: libtool-is-a-fool shouldn't be needed any more -- try without it and see if it works.
[09:24] <dholbach> Mithrandir: ok
[09:24] <Mithrandir> sivang: uhm?  I'm not sure I understand what you are talking about
[09:26] <sivang> Mithrandir: I have used dh_install to install a couple of files into /etc/pkgname
[09:27] <sivang> Mithrandir: when I remmove the package (purge) I get an error that dpkg couldn't remve the dir because it is not empy, which is right :)
[09:29] <dholbach> Mithrandir: now i get those lintian-warnings again
[09:29] <Mithrandir> sivang: uhm, if you install stuff to /etc/pkgname, it's marked as a conffile and dpkg should DTRT when purging the package.  If you can provide the package in question, it's probably easier to see what's going on
[09:30] <YokoZar> Would it be ok to have symbolic links in /usr/share/package to /usr/bin ?  The scripts that I'm packaging are kinda screwy that way
[09:30] <Mithrandir> dholbach: hm.  If I may, I'd like to take a look at your package.
[09:30] <dholbach> right
[09:30] <Mithrandir> YokoZar: sure, why shouldn't it be ok?
[09:30] <YokoZar> Because /usr/share is for read only non-executables
[09:31] <smurfix> YokoZar: a symlink is not an executable. ;-)
[09:31] <YokoZar> Yeah but what if someone had /usr/share mounted with no execute
[09:31] <YokoZar> As he is supposed to be able to do
[09:31] <smurfix> YokoZar: ... though I might suspect that the stuff belongs in /usr/lib if it wants to exec from there
[09:31] <YokoZar> Wouldn't trying to run a symlink there barf?
[09:31] <smurfix> YokoZar: I don't know offhand if symlinks are followed before that test
[09:31] <smurfix> YokoZar: let me check
[09:32] <Treenaks> jordi: yeah, a long long time ago :)
[09:32] <smurfix> YokoZar: no, that works
[09:32] <YokoZar> So I can have the symlink?
[09:32] <dholbach> Mithrandir: http://moz.gotdns.org/ubuntu/
[09:33] <smurfix> YokoZar: as I said, it might want to live in /usr/lib instead, but in principle, yes
[09:34] <sivang> smurfix: Moins :)
[09:36] <sivang> Mithrandir: on my way, will send you a link in a momnet.
[09:40] <sivang> Mithrandir: http://sivan.workaround.org/gnome-system-tools_1.1.90-0ubuntu1_i386
[09:55] <sivang> Mithrandir: did you get my last msg to you?
[09:57] <dholbach> Mithrandir: wb
[10:04] <kagou> hi
[10:04] <dholbach> hi kagou
[10:15] <abelli> smurfix: ping
[10:15] <smurfix> abelli: 
[10:15] <abelli> smurfix: what was it?! ;) i can't read
[10:16] <smurfix> that's a chinese character. "pong".
[10:16] <abelli> smurfix: wow ;)
[10:16] <abelli> smurfix: did you receive the email?
[10:18] <dholbach> Mithrandir: did you receive the link before you flew out?
[10:18] <smurfix> abelli: seems not
[10:19] <dholbach> Mithrandir: http://moz.gotdns.org/ubuntu/glibmm2.4-2.4.5/ (to make sure ;-))
[10:19] <smurfix> abelli: .. which could mean that I deleted it accidentally :-/
[10:20] <abelli> smurfix: im forwarding it..
[10:27] <sivang> Mithrandir: are you back?
[10:27] <dholbach> sivang: idle:32 minutes :-)
[10:29] <sivang> dholbach: yeah :) I wonder where he vainshed ;-)
[10:30] <dholbach> sivang: gives me time to get some appropriate webspace :-)
[10:30] <dholbach> sivang: i only have 16K/s upstream :-)
[10:30] <abelli> smurfix: you, should probably, have mail ;)
[10:32] <sivang> dholbach: hehe, I have 128Kb upstream now, had 96 in the past..
[10:32] <dholbach> sivang: bit or byte? :-)
[10:36] <dholbach> Mithrandir: ok... once the upload finished (some minutes), it's all on http://ubuntu.stufenseite.de
[10:40] <sivang> dholbach: errr, bits :)
[10:41] <dholbach> sivang: yes... me too :-)
[10:42] <sivang> dholbach: we'd wish we had Mithrandir *fast* net access :)
[10:43] <YokoZar> Can someone test a package for me?  It creates a working build, but it never asks me for my password to sign the key file (make has a weird error at the end)
[10:44] <YokoZar> deb-src http://tuzakey.com/~scott/apt/ source/
[10:44] <YokoZar> package name winetools
[10:44] <dholbach> sivang: i rather wish, i'd finished the exam next week and my thesis :-)
[10:46] <YokoZar> Mithrandir: ping
[10:47] <dholbach> poor mithrandir... 3 people hopping around him :-)
[10:48] <YokoZar> Heh, it's the price you get for being helpful and knowledgable :)
[10:48] <YokoZar> On a Friday night, no less
[10:48] <dholbach> it's saturday morning ;-)
[10:48] <dholbach> and i must be off - revision for an exam :-/
[10:48] <dholbach> so have a nice day, you all
[10:49] <dholbach> see you later
[10:51] <Kaloz> morning
[11:10] <sivang> Kaloz: morning 
[11:46] <jdub> thom: we *have* gsf-sharp? :)
[11:54] <thom> apparently
[11:57] <jdub> $ apt-cache search gsf cil | wc -l
[11:57] <jdub> 0
[11:57] <jdub> $ apt-cache search gsf sharp | wc -l
[11:57] <jdub> 0
[11:58] <sivang> seb128: Morning :)
[12:00] <seb128> hello
[12:02] <sivang> seb128: you know anything about a g-s-t bug which display the postfix, had, dbus usernames in the user editor?
[12:02] <Hwolf> Is there any bug known that can cause I/O errors when doing a dist-upgrade?
[12:03] <seb128> sivang: no idea on what you are talking about
[12:03] <Mithrandir> Hwolf: hardware errors.
[12:04] <Mithrandir> T-Bone: what's the problem?
[12:04] <Hwolf> Mithrandir: I doubt it. I got it dist-upgrading 2 ubuntu-installations yesterday. On 2 different hdd's
[12:04] <T-Bone> Mithrandir: well, I'm having some issues with zx6000: the MPT fusion drivers isn't properly loaded:
[12:04] <Mithrandir> Hwolf: same machine or different machines?
[12:04] <T-Bone> Mithrandir: mptbase gets loaded, but not mptscsih, so that d-i doesn't see any SCSI host, and thus no disk on the box
[12:05] <Mithrandir> T-Bone: loading it by hand works?
[12:05] <T-Bone> yes
[12:05] <sivang> Mithrandir: I don't reckon you recived my link before you were diconnected?
[12:05] <Hwolf> Mithrandir: same machine, different hdd's
[12:06] <Mithrandir> sivang: can't see any, no.
[12:06] <Mithrandir> Hwolf: well, bad memory, bad motherboard, stuff like that.
[12:06] <Mithrandir> T-Bone: sounds like the discover data files might need updating.  Do you know if it's autodetected by hotplug?
[12:07] <jdub> hrm, you guys know of any p2p vpn technologies?
[12:07] <jdub> (stressing the P in VPN, not necessarily security)
[12:07] <T-Bone> jdub: freenet?
[12:07] <Mithrandir> jdub: OE+IPSec?
[12:07] <Hwolf> Mithandir: Isn't it odd that it happens only on a hoary-dist-upgrade?
[12:07] <T-Bone> Mithrandir: that info can be found in the installer logs?
[12:08] <Mithrandir> Hwolf: wrong, you have only seen it happen in that particular case.  :)
[12:08] <Mithrandir> T-Bone: I'm not sure if the installer uses hotplug, which is why I'm asking.
[12:08] <T-Bone> yeah iirc it uses discover
[12:08] <Hwolf> Mithrandir: Compiled gentoo on this box without faults, I think it'd have shown up...
[12:08] <Mithrandir> you can use both, I know it uses discover.
[12:09] <Mithrandir> T-Bone: I was the guy developing d-i for a long time. :)
[12:09] <Mithrandir> Hwolf: true.
[12:09] <T-Bone> Mithrandir: since i have not installed that box yet (and not yet willing to do so) it's gonna be hard to tell
[12:09] <T-Bone> Mithrandir: hehe, just noticed who you are, thanks to /whois ;)
[12:10] <Mithrandir> T-Bone: what's the PCI ID of the card?
[12:10] <jdub> Mithrandir: needs to be a private network, not very interested in encryption; OE ends up being a bit useless there ;)
[12:11] <T-Bone> Mithrandir: hold on, the box is booting.
[12:11] <jdub> T-Bone: hrm
[12:11] <jdub> T-Bone: where would i find out more?
[12:11] <T-Bone> www.freenet.org iirc
[12:11] <jdub> ta
[12:11] <Mithrandir> jdub: you could just use ip tunneling.
[12:11] <jdub> Mithrandir: mmm, considering it
[12:12] <jdub> Mithrandir: sorry for doubling up your sounder mail, btw, was replying as you8rs arrived
[12:12] <Mithrandir> jdub: ip tunneling and possibly do RIP or OSPF on top of it.  I don't know what complexity you are talking about.
[12:12] <T-Bone> jdub: it has moved to freenet.sf.net
[12:12] <Mithrandir> jdub: ah, I guess he's just happy about two answers.
[12:12] <Mithrandir> :)
[12:13] <jdub> T-Bone: hrm, doesn't sound appropriate ;)
[12:13] <jdub> so the idea is
[12:14] <jdub> to have a vpn across the 'net that our xboxes are attached to
[12:14] <jdub> and possibly dual-use it for private sharing fun
[12:14] <T-Bone> tunneling is what you want then
[12:14] <jdub> seems to be the simplest approach
[12:14] <abelli> whata about tunneling..yeah..
[12:15] <T-Bone> add a bit of IPSEC for security and you're there
[12:15] <jdub> security is for people who have something to hide
[12:15] <Mithrandir> jdub: normal static ip tunneling is dog simple to set up, at least as long as you don't want this to suddenly become a mesh or something.
[12:15] <jdub> ;-)
[12:15] <Mithrandir> I should get ipsec up and running on my home net
[12:15] <jdub> Mithrandir: that's where it breaks down; i predict that being the next step
[12:15] <T-Bone> hehe
[12:15] <jdub> and i don't really want all the data going through a central host
[12:16] <Mithrandir> jdub: daisy-chain?
[12:16] <jdub> pretty breakable ;)
[12:16] <Mithrandir> that's also fairly simple, as long as the chain's not too long.
[12:16] <T-Bone> Mithrandir: come to think of it: there's no way hotplug would _need_ to detect mptfusionh:
[12:16] <Mithrandir> T-Bone: mhm?
[12:16] <T-Bone> Mithrandir: the box _can't boot_ if that driver isn't preloaded by the initrd
[12:16] <T-Bone> (or builtin the kernel)
[12:17] <Mithrandir> T-Bone: ah, makes sense.  Kindof.  Discover's data files should possibly just be updated, then.
[12:18] <T-Bone> Mithrandir: PCI ID: 20:01.[01]  SCSI storage controller: LSI Logic / Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI (rev 07)
[12:18] <T-Bone> (there are two of these)
[12:18] <T-Bone> Mithrandir: if you tell me how to do that, I'm ok to do it :)
[12:19] <Mithrandir> T-Bone: look at the /usr/share/discover/pci-26.lst file; The format is fairly self-explanatory.  That is generated from somewhere, but I don't remember where offhand.
[12:20] <Mithrandir> jdub: look at mobilemesh?
[12:21] <T-Bone> k. lemme boot a ubuntu box :)
[12:21] <jdub> Mithrandir: looking...
[12:22] <Mithrandir> jdub: I think it might be appropriate, but I've never used it myself.
[12:22] <Mithrandir> seems like it was written for 2.2, so it might need a bit of brushing-up.
[12:22] <ogra> mrning
[12:23] <Mithrandir> jdub: if you could rely on a single host as a coordinator (not traffic point), you could probably whip something up.
[12:23] <Mithrandir> hiya ogra 
[12:23] <Kaloz> hi ogra 
[12:24] <ogra> Kaloz: hey mr MOTU :)
[12:24] <ogra> Mithrandir: hi
[12:24] <Kaloz> :)
[12:24] <T-Bone> Mithrandir: there is alas one big issue on ia64: 2.6.10 seems to hate udev/hotplug :P
[12:25] <T-Bone> (or the other way around)
[12:25] <Mithrandir> T-Bone: that sucks. :)
[12:25] <T-Bone> indeed :)
[12:26] <ogra> does anybody know if gparted is making it into sid soon ? i'm just updating the UniverseCandidates page to give some MOTU suggestions...
[12:28] <sivang> Mithrandir: would you be in for checking my pakcage? I ahve it online now
[12:28] <Josephus> hey
[12:28] <sivang> ogra: hi 
[12:28] <ogra> sivang: hey ;)
[12:29] <sivang> ogra: 'sup?
[12:29] <Mithrandir> sivang: url?
[12:29] <sivang> Mithrandir: sec
[12:31] <sivang> Mithrandir: http://sivan.workaround.org/gnome-system-tools_1.1.90-0ubuntu1_i386.deb
[12:31] <Mithrandir> I need the source package -- the binary package is uninteresting. :)
[12:32] <sivang> Mithrandir: ah ok :) How do I create a source package from my modified source tree?
[12:32] <Mithrandir> debuild -S
[12:32] <ogra> sivang: debuils -S
[12:32] <ogra> oops
[12:32] <sivang> Mithrandir: ok , tnx
[12:32] <sivang> ogra: tnx you too
[12:32] <sivang> :)
[12:35] <sivang> Mithrandir: building..
[12:35] <T-Bone> Mithrandir: who should I send the patch to? :)
[12:36] <sivang> Mithrandir: how do I tell it to not try sign the pkg? -ns ?
[12:36] <sivang> Mithrandir: (I don't have the key at hand)
[12:36] <Mithrandir> T-Bone: have you tried it yet?
[12:36] <Mithrandir> sivang: -us
[12:36] <sivang> Mithrandir: tnx
[12:37] <T-Bone> Mithrandir: that's the other question: how should i test it?
[12:37] <seb128> sivang: 1.1.90-0ubuntu1 is already in the archive, better to change the number if you make changes
[12:37] <sivang> seb128: yes I know, I am just experimenting without dch'ing
[12:38] <Mithrandir> T-Bone: build a new initrd with the fixed list.
[12:38] <sivang> seb128: btw, the "bug" I was referring was a strange behavior as I noted, however it may be specific to my system, I hope Mithrandir has a free box to try my crack on see if it also shows him some of the "system" users rather then only the interactive ones.
[12:39] <seb128> sivang: if you don't describe the bug in a better way ..
[12:42] <T-Bone> Mithrandir: the easier way is to look at debian's discover data and notices that it has much more stuff than we have, including the patch i was considering
[12:42] <Mithrandir> T-Bone: hm, ok.  Prod Kamion about it, then?
[12:42] <T-Bone> Mithrandir: seems we need to update our data
[12:42] <thom> seby seb! nautilus is doing odd stuff
[12:42] <T-Bone> Mithrandir: yup, will do
[12:44] <Mithrandir> thom: of course, it's _nautilus_.
[12:44] <thom> Mithrandir: weirder than usual
[12:46] <Mithrandir> thom: it's saturday.  And it's before noon (for you).  Obvious you'll have it do weirder odd stuff than usual.
[12:46] <Mithrandir> s/Obvious/&ly/
[12:46] <T-Bone> mail sent to Kamion
[12:46] <Mithrandir> it was probably out drinking last night
[12:48] <sivang> seb128: ok, when you just fire up users-admin you get to see only interactive users right, it's not showing you all the system internal use users for all the services and daemons.
[12:49] <seb128> sivang: correct
[12:51] <thom> Mithrandir: true enough
[12:51] <thom> Mithrandir: i think seb128 is ignoring me tho ;P
[12:52] <sivang> seb128: ok, then I get to see to additional following: postfix, messagebus, hal and 2 more...
[12:52] <seb128> thom: what did you do to nautilus again ? :p
[12:53] <seb128> sivang: works here ...
[12:53] <sivang> seb128: ok, I hope it not something my package breaks...
[12:53] <sivang> Mithrandir: any luck? ;-)
[12:53] <thom> seb128: open up an sftp connection, open my home folder, drag and drop from the remote to the local, and the local window gets closed every time
[12:53] <Mithrandir> sivang: I haven't seen an url from you yet?
[12:54] <sivang> Mithrandir: http://sivan.workaround.org/gnome-system-tools_1.1.90-0ubuntu1_i386.deb
[12:54] <sivang> sorry
[12:55] <seb128> thom: that's weird ... anything in .xsession-errors ?
[12:55] <seb128> sivang: better to give the package sources than a deb
[12:55] <ogra> is there been any progress on #5870 (evo on amd64) ?
[12:57] <sivang> seb128: eer ops
[12:57] <sivang> Mithrandir: sorry, oops
[12:57] <Mithrandir> ogra: I told seb what was wrong, he bounced it back to me.
[12:58] <ogra> hmm
[12:58] <Mithrandir> so I need to whack it a bit.
[12:58] <ogra> heh
[01:00] <Mithrandir> it's a shame that the distributed nature of the project means incitations in the form of alcohol doesn't really work.
[01:16] <T-Bone> feels like OSX' Update Manager :^)
[02:01] <kent> wow, the new update-manager in Hoary is very nice :)
[02:23] <Josephus> indeed
[03:04] <trulux> bluefoxicy: pitti online
[03:11] <dholbach> re
[03:13] <rouven> is anyone in here taking part in the website look and feel competition?
[03:17] <dholbach> rouven: how long will it be until submission?
[03:18] <sivang> pitti: ping
[04:30] <dholbach> re ogra
[04:30] <ogra> hey
[04:30] <robertj> flash crashed firefox and it took me 5 minutes before I could kill the thing with the gui, where is the appropritae place to file a bug?
[04:31] <Treenaks> robertj: yes, at Macromedia
[04:31] <Treenaks> robertj: (flash is not supported)
[04:31] <robertj> Treenaks: why should a browser crash bring my whole desktop experience to a grind?
[04:32] <Treenaks> robertj: no, the flash player is buggy and brought your browser down, which brought your desktop down
[04:32] <Hwolf> Treenaks: Why is a browser able to bring a desktop down?
[04:32] <robertj> Treenaks: it's a bug, and not a flash problem, deal with it
[04:32] <Treenaks> Hwolf: well, it could make it slow... add more memory
[04:33] <robertj> if someone goes and uses your file format to do something it wasn't intended to do, and it crashes, it shouldn't grind the whole desktop to a halt
[04:33] <Treenaks> robertj: ask on #ubuntu - this is the development channel. not the supoprt channel.
[04:33] <Treenaks> robertj: so file a bug on the C compiler.. it allows you to write programs that bring down the machine
[04:34] <Hwolf> Treenaks: isn't there a system to kill rampant memory bloat?
[04:34] <Treenaks> Hwolf: yes. in the kernel: the oom killer (out-of-memory)
[04:34] <trulux> robertj: what happens exactly?
[04:34] <trulux> robertj: anything back from the kernel?
[04:35] <robertj> trulux: surely not
[04:35] <robertj> trulux: planet.gnome.org has a link to an swf file which brings down flash because excessive # of objects created by a bug in the vnc2swf socftware bringi t down
[04:36] <Hwolf> guys: Is there any chance that a software bug starts I/O errors?
[04:36] <robertj> which is fine, I expected "Flash has crashed. Would you like to restart it?"
[04:36] <trulux> robertj: OK
[04:36] <trulux> Hwolf: pretty difficult in my opinion
[04:36] <trulux> Hwolf: what software are you talking about?
[04:37] <robertj> I thought things were supposed to be sufficiently niced these days to stop that from occuring
[04:37] <Hwolf> trulux: i dist-upgraded my hoary installation last night, and I dpkg/apt died on me with I/O errors
[04:38] <Hwolf> trulux: Then I installed warty on the same machine, other hdd and dist-upgraded it fresh. Again those errors.
[04:38] <trulux> strange
[04:38] <trulux> anything runnind under root privileges has at least CAP_SYS_RAWIO capability
[04:38] <trulux> so, if something running under such privileges is doing weird stuff... a bad thing
[04:38] <tseng> whats rawio have to do with apt
[04:38] <Hwolf> trulux: that is alien language to me.
[04:38] <trulux> tseng: nothing
[04:39] <robertj> Hwolf: come to #ubuntu-sed
[04:39] <trulux> tseng: it should be a problem not related with userland stuff
[04:44] <mjt> hmm. after installing xorg 6.8.1-1ubuntu10, xv stopped working - the picture looks like some green garbage (any app using xv - xine, tvtime, mplayer, ...).  Any pointers?
[04:45] <mjt> s/6.8.1-1ubuntu10/6.8.1-1ubuntu11/
[05:08] <ogra> trulux: how about graveman 0.3.4 ?
[05:12] <ogra> trulux: oops, pardon, i meanr trukolo.... damned autocompletion...
[05:22] <bluefoxicy> trulux: what pitti where?
[05:23] <bluefoxicy> trulux:  I made a mistake on those exec-shield tests
[05:23] <bluefoxicy> I forgot to execstack -c them all, and some had PT_GNU_STACK
[05:23] <bluefoxicy> so it turs out execshield didn't fail ALL the tests
[05:23] <bluefoxicy> it passed ONE
[05:23] <bluefoxicy> (non-executable stack)
[05:24] <trulux> grrrreat
[05:24] <trulux> one test of....
[05:24] <trulux> around 20?
[05:24] <bluefoxicy> 19
[05:24] <bluefoxicy> I counted.
[05:24] <bluefoxicy> also
[05:24] <trulux> that's a great result man
[05:24] <trulux> it could even worst
[05:24] <bluefoxicy> well anything that can evade propolice can evade their NX stack
[05:25] <bluefoxicy> so a propolice deployment would probably be "genuinely better"  :P
[05:25] <bluefoxicy> http://lists.debian.org/debian-devel/2003/11/msg00227.html
[05:25] <bluefoxicy> and Ingo calls foul on paxtest 0.9.5
[05:25] <trulux> bluefoxicy: we can write an es regression test , specific for it
[05:25] <bluefoxicy> so I audited and anylized 0.9.6 and find no such mmap() or mprotect() calls that would disable exec shield, it just sucks
[05:25] <bluefoxicy> trulux:  I swear they're just going to kill me
[05:26] <bluefoxicy> trulux: es regression test is easy
[05:26] <bluefoxicy> int main() {
[05:26] <bluefoxicy>   printf("Passed.\n");
[05:26] <bluefoxicy> }
[05:26] <tseng> funny
[05:26] <trulux> XD
[05:26] <trulux> this intentionally calls mprotect(PROT_EXEC) for the highest possible
[05:26] <trulux> address one can think of. This call has no useful purpose at all. In other
[05:26] <trulux> words, this is a specific, underhand cheat to trigger 'Vulnerable'
[05:26] <trulux> messages for all items when running paxtest on exec-shield kernels.  
[05:26] <trulux> Bravo!
[05:26] <trulux> Mingo Dixit
[05:26] <bluefoxicy> trulux:  yeah, I think Ingo had a point about 0.9.5
[05:26] <sivang> Mithrandir: ping, I have a source pkg
[05:26] <sivang> Mithrandir: is it still relavent? ;-)
[05:27] <bluefoxicy> but I used 0.9.6, and it looks OK (even an strace shows nothing nasty)
[05:27] <trulux> ok
[05:27] <trulux> listen
[05:27] <trulux> maybe Ingo is not so wrong
[05:28] <trulux> paxtest DOES specifical stuff for pax testing
[05:28] <bluefoxicy> doesn't matter.  I used real tests and got real results.
[05:28] <trulux> and PaX has a REALLY different mem mapping style
[05:28] <trulux> so
[05:28] <tseng> are you guys making a point, or is this #debian-hardend stuff?
[05:28] <bluefoxicy> uh
[05:28] <bluefoxicy> tseng has a point
[05:28] <bluefoxicy> I just woke up
[05:28] <bluefoxicy> and I was looking for pitti
[05:28] <trulux> using a mprotect call against a top mem area gets killed on PaX
[05:28] <trulux> but not on ES
[05:29] <trulux> it uses a different manner for that, even unsecure when doing so
[05:29] <trulux> but different in the end
[05:29] <trulux> anyway this is a bit funny:
[05:29] <trulux> here are the paxtest-0.9.5 results with that single purpose-less line
[05:29] <trulux> removed, for the categories that matter to me:
[05:29] <trulux>  Executable anonymous mapping             : Killed
[05:29] <trulux>  Executable bss                           : Killed
[05:29] <trulux>  Executable data                          : Killed
[05:29] <trulux>  Executable heap                          : Killed
[05:29] <trulux>  Executable stack                         : Killed
[05:30] <trulux> paxtest gives Killed if the called test returns 1) exit(0) 2) breaks 3) gets SIGKILL
[05:30] <trulux> it's just like when you catch up the elf_map() call in the kernel
[05:30] <trulux> with a file_mmap() syscall returning something but not 0 or 1
[05:30] <trulux> (0 )
[05:30] <trulux> so
[05:31] <trulux> elf_map()->file_mmap()->...
[05:31] <trulux> if file_mmap() returns -EPERM, -EINVAL, whateverelse rather than 0
[05:31] <trulux> then it breaks
[05:31] <trulux> abd gets "Killed"
[05:31] <trulux> that's what vSecurity does
[05:31] <trulux> in the TPE engine
[05:32] <bluefoxicy> now i know how tseng feels when I do it.
[05:32] <bluefoxicy> :)
[05:32] <trulux> so
[05:32] <bluefoxicy> but heh
[05:32] <trulux> Ingo is a bit wrong if he thinks that killed means explicitly not vulnerable
[05:32] <trulux> collision-rts handles the output in a different manner, so, most Killed calls will be because of dirty exits and such
[05:33] <trulux> caught?
[05:33] <trulux> :)
[05:33] <bluefoxicy> heh
[05:33] <trulux> Please, guys, don't have your discussion here. I don't think we really
[05:33] <trulux> care about the differences between PaX and exec-shield. Debian is not,
[05:33] <trulux> and, to the best of my knowledge, will not, choose one for its kernels,
[05:33] <trulux> so there is no need to prove that one or the other is better.
[05:33] <trulux> -- 
[05:33] <trulux> gram
[05:33] <trulux> lol
[05:34] <trulux> ;P
[05:34] <bluefoxicy> trulux:  so pitti isn't online anymore?
[05:35] <trulux> no idea
[05:35] <trulux> anyways, I'm starting to loss the faith on Hardened Debian
[05:35] <bluefoxicy> why, you're doing a good job
[05:35] <trulux> and if it doesn't give something new out I will continue doing upstream work
[05:36] <bluefoxicy> upstream work is part of hardening isn't it?
[05:36] <trulux> bluefoxicy: i dislike being pissed off with stupid things like "trademark uses, etc"
[05:36] <sivang> Mithrandir: http://muse.19inch.net/~sivan/g-s-t/
[05:36] <trulux> sure
[05:36] <robertj> trulux: really the only thing I think really needs to get done is to have pam check for weak passwords on outward-facing daemons
[05:36] <trulux> robertj: that's cracklib man
[05:36] <robertj> yeah
[05:36] <trulux> and it's inside PAM already
[05:36] <trulux> and not
[05:37] <trulux> passwords are weak
[05:37] <azeem> sivang: there's no .orig.tar.gz
[05:37] <robertj> it's not installed by default thoug hright?
[05:37] <trulux> passwords are obscurity for security by meaning
[05:37] <trulux> passwords are pure crap
[05:37] <robertj> trulux: and they are GoodEnough(TM) for most people
[05:37] <robertj> krb5 is where it's at though
[05:37] <trulux> somebody relying on them for system security is really a man in a *nut*shell
[05:37] <trulux> in a big clue ;)
[05:37] <bluefoxicy> robertj: http://ubuntulinux.org/wiki/USNAnalysis
[05:38] <trulux> robertj: krb5 has demonstrated to be not as secure as many people commented in the past
[05:38] <trulux> or at least it's implmentations
[05:38] <trulux> but dunno
[05:38] <sivang> azeem: eh yes, I guess it'd be the same as apt-get source gnome-system-tools
[05:38] <azeem> sivang: so then the .diff.gz is missing
[05:39] <sivang> azeem: ok, checking
[05:39] <trulux> bluefoxicy: I'm going to a start a flame at the LKML
[05:39] <bluefoxicy> robertj:  security is more of keeping an eye on how many ways someone can log into your system, and getting rid of the non-authenticated ones
[05:39] <bluefoxicy> trulux:  don't do that :p
[05:39] <trulux> bluefoxicy: "sys_chroot() broken by design, enhance or die"
[05:39] <sivang> azeem: strange, that's waht I got debuilding -S
[05:39] <bluefoxicy> haha
[05:39] <trulux> "use fscking jail()"
[05:40] <trulux> "stop arguing that i do shit, I know it already
[05:40] <trulux> "I don't have pains, I just clear up my terminal with 2.6 oops"
[05:41] <trulux> bluefoxicy: the other point is that I CAN'T add a new LSM hook for sys:chroot, they WILL NOT accept it
[05:41] <trulux> so
[05:41] <trulux> I use my own BSD Jails code base from Serge
[05:41] <trulux> and send chroot jails out
[05:41] <trulux> bluefoxicy: we have an userland tool for chroot namespace changing
[05:41] <azeem> sivang: then you uploaded the wrong files :)
[05:41] <robertj> blufox: my big security question is, drumroll...
[05:41] <trulux> as a replacement of chroot
[05:42] <bluefoxicy> robertj:  no it won't be harder to use.
[05:42] <mjt> anyone expirienced with X internals around?  Or, maybe, what to do with a particular driver problem I have (trident_drv in 6.8.1ubuntu11 have issues with xv)?
[05:42] <sivang> azeem: I'll send you my ls output
[05:42] <trulux> bluefoxicy: anyways, I must finish up the vsec stuff
[05:42] <robertj> blufox: ready for this...is it possible to secure mod_dav?
[05:42] <trulux> ajmitch: ping
[05:42] <bluefoxicy> robertj:  what the heck is mod_dav?  :P  apache module
[05:42] <robertj> yeah
[05:42] <bluefoxicy> don't know, i'm not an apache administrator guy
[05:43] <trulux> robertj: DAV? PUT shit /boo/ :)
[05:43] <trulux> ajmitch: there?
[05:43] <robertj> I've been meaning to look over the  OS 10.3 server at work and see how it handles the situation
[05:43] <trulux> bluefoxicy: back to work, see you later
[05:44] <robertj> bluefoxicy: you don't like krb5 though?
[05:45] <bluefoxicy> I dunno..  I don't use kerberos, I don't have a kerberos network
[05:46] <robertj> it's nice but the docs suck and the gui tools are non-existant
[05:46] <robertj> and security is very gui-centric
[05:46] <robertj> at least gnome keyring is getting together
[05:52] <robertj> sometimes I wonder if the needed info could be stored somewhere in case krb5d was brought online later
[06:27] <dholbach> can I specify files somewhere in  ./debian/  that are to be ignored?
[06:28] <T-Bone> dholbach: define 'ignored'
[06:28] <dholbach> doxygen spits out an empty doxyfile-warnings.txt, which i dont want to remove by some shell hack
[06:28] <dholbach> T-Bone: where can i specify that?
[06:29] <T-Bone> that happens during the binary build, i suppose?
[06:29] <dholbach> T-Bone: yes
[06:29] <T-Bone> and you don't want to have it in the binary deb?
[06:30] <T-Bone> if it's not part of the install target, it won't get in. And if it is, then remove it before it actually makes it in  (edit the rules file)
[06:30] <dholbach> and  WARNINGS = no   doesnt keep it from dropping that stupid file
[06:30] <T-Bone> dholbach: as long as it is not part of the install target, it's not a problem, it won't be included in the final package
[06:31] <dholbach> T-Bone: so you suggest some shell-hack
[06:31] <T-Bone> dholbach: no
[06:31] <dholbach> T-Bone: i want to get rid of the lintian-warning
[06:31] <T-Bone> dholbach: i don't actually see why it would be part of the install target. Either the Makefile is wrong, or you need some modifications to the rules file
[06:31] <T-Bone> what does lintian says?
[06:31] <dholbach> T-Bone: i guess the latter
[06:32] <dholbach> T-Bone: W: libgtkmm-2.4-doc: zero-byte-file-in-doc-directory usr/share/doc/libgtkmm-2.4-doc/reference/html/doxygen-warnings.txt
[06:32] <T-Bone> ah!
[06:32] <T-Bone> it gets in docs
[06:32] <dholbach> i guess everyone could live with an empty file, but... :-)
[06:33] <T-Bone> no
[06:33] <T-Bone> hmm
[06:33] <dholbach> but i want lintian-warning-free packages :-)
[06:33] <T-Bone> sure
[06:33] <T-Bone> that's a good thing
[06:33] <dholbach> i could try to turn doxygen to be QUIET
[06:33] <dholbach> but it'd take me 20 minutes to find out
[06:33] <dholbach> so i thought some  ignore = ... thingie would be cool
[06:34] <T-Bone> well i don't know how your package gets build, so i can't tell off hand what to do. In such a case, i'd suggest using the '-X' flag of dh_installdocs
[06:35] <dholbach> oh yes cool *has a look*
[06:35] <dholbach> T-Bone: YOU rock
[06:39] <Hwolf> I just installed a hoary array, and it did not pick up my sound. a warty-hoary upgrade did.
[06:46] <sivang> is there anybody interested in testing my bin package? now install and removal works, and you should be getting two profiles for adding users. Also, config data files are now expected to be in /etc/gnome-system-tools
[06:46] <sivang> http://muse.19inch.net/~sivan/gnome-system-tools_1.1.90-0ubuntu3_i386.deb
[06:47] <dholbach> sivang: if you'd provide a source "package", i'd do a test on amd64 :-)
[06:47] <sivang> dholbach: ok, sure :)
[06:47] <sivang> dholbach: if you could, please test netowrk config also to see if ti stores antying in /etc/g-s-t
[06:52] <dholbach> sivang: you're doing a   debuild -S  ?
[06:52] <sivang> dholbach: already did, uploading..
[06:52] <dholbach> sivang: cool :-)
[06:53] <T-Bone> Kamion: ping?
[06:54] <sivang> dholbach: takinhg ages to upload..
[07:13] <marcin_ant> hello - any zope/plone guru/user available here?
[07:21] <Nonphasis> I wonder if this is the place to whine about Hoary breakage?
[07:21] <usual> whats broken
[07:22] <Nonphasis> sound
[07:23] <Nonphasis> spdif goes "off"
[07:23] <Nonphasis> it works ok in the beginning with esd (nhythmbox etc), breaks in beep-media player
[07:24] <Nonphasis> I wonder if alsa-utils update today broke it
[07:48] <thully> hi - is the lack of localization on the live CD a known issue - I'm learning Spanish and decided to try using Spanihs from the live CD, but the booted system was still in English.
[07:53] <mdz> seb128, thom: firefox is stealing my window manager shortcuts now; is this a firefox change or a metacity change?
[07:54] <seb128> since when ?
[07:54] <seb128> I doubt that's a change on the GNOME side
[07:55] <mdz> since I restarted firefox, approximately
[07:56] <mdz> I use alt+1, alt+2, alt+3 etc. for switching desktops
[07:56] <mdz> and now they switch tabs in firefox
[07:56] <seb128> ok, so firefox
[07:56] <mdz> so when I am switching between desktops, I get to firefox and I can't escape :-)
[07:56] <seb128> ah ah
[07:56] <thully> is the lack of localization on the live CD a known issue at this point?  I tested it just for kicks and noticed this, and I think I remember someone saying something about this
[07:56] <seb128> hum, I've to go now, later guys
[07:58] <dholbach_dogwalk> bye seb128
[08:01] <mdz> thully: it is a known issue in array3.5; it should be fixed in the latest daily
[08:07] <marcin_ant> I cannot login to zope manage application - could someone tell me what is default password and login on ubuntu?
[08:07] <T-Bone> has someone tried 'man <anything>' on hoary lately?
[08:07] <T-Bone> i'm getting square chars in the output
[08:07] <T-Bone> that makes the manpages hard to decypher...
[08:08] <rubenv> T-Bone: in what page specifically?
[08:08] <rubenv> and what term?
[08:09] <T-Bone> rubenv: VT terminal
[08:09] <T-Bone> man hdparm, man lspci for instance
[08:09] <rubenv> both gnome term and (u)xterm do great here
[08:10] <T-Bone> it doesn't do that for all manpages tho
[08:10] <T-Bone> but it does so with man lspci and man hdparm
[08:10] <rubenv> aha
[08:10] <T-Bone> got it?
[08:10] <rubenv> getting it too on the tty's
[08:10] <T-Bone> does the same on remote ssh term
[08:11] <rubenv> that's rather a matter of your local term
[08:11] <dholbach> T-Bone: btw it worked ;-)
[08:11] <rubenv> but i guess it's a font issue with the tty
[08:11] <rubenv> (didn't it change lately?)
[08:12] <T-Bone> rubenv: i don't see why it'd be a matter of my local term, which works fine with a good 50 remote machines i connect to, including funny OSes such as IRIX, Solaris and HPUX, among a few :P
[08:12] <T-Bone> and my 'local' term isn't an Ubuntu one, fwiw (running debian here)
[08:13] <rubenv> T-Bone: it's your local term that displays the weird chars
[08:13] <T-Bone> rubenv: it's the remote manpage that sends them
[08:13] <T-Bone> rubenv: it does the same on a VT100 console I have hooked to the serial port
[08:13] <rubenv> T-Bone: man works great on xterm or gnome term
[08:14] <T-Bone> which i'm using
[08:14] <T-Bone> so i'm correcting: it works (presumably, can't test that yet) great with hoary's gnome term or xterm
[08:14] <rubenv> then you're not getting the same i'm getting :)
[08:14] <thully_> hi - I lost my connection - so (sorry if you've already seen this) - is the lack of localization on the live CD a known issue?  I tested it for kicks and it didn't work (GNOME was still in english)
[08:16] <T-Bone> rubenv: you're getting a weird accentuated char followed by two squares in particular for each "-" option listed on the manpage
[08:16] <T-Bone> right?
[08:17] <rubenv> yes, but only on my tty, not in an X terminal
[08:17] <T-Bone> right, tested just now
[08:17] <T-Bone> it works fine with hoary's xterm
[08:17] <T-Bone> not with debian xterm
[08:17] <rubenv> hmmm
[08:17] <T-Bone> that's definitely a bug
[08:18] <rubenv> anyone has a warty to test with?
[08:18] <T-Bone> xterm Xt error: Can't open display: %s
[08:18] <rubenv> xorgs xterm is quite different
[08:18] <T-Bone> that's another bug AFAICT
[08:18] <T-Bone> (localization bug?)
[08:18] <sladen> UTF-8 is enabled in each case?
[08:18] <rubenv> localization?
[08:18] <rubenv> full utf8 here
[08:19] <T-Bone> rubenv: the "%s"
[08:19] <T-Bone> it should be the name of the host i'm trying to connect to
[08:19] <rubenv> T-Bone: that's not necessairly localization :)
[08:20] <T-Bone> right, just a guess ;)
[08:20] <T-Bone> man mkfs does the same but not man mkinitrd (both in section 8)
[08:21] <T-Bone> man less is also fux0red
[08:21] <T-Bone> well i'm not gonna list them all, i think there's enough for a testcase already :P
[08:36] <Hwolf> b
[09:07] <thully> I wanted to test installer clock handling in the case that the clock is set to local time - how could I force the installer to do this (it always goes to UTC without asking, and I've already tested that case to be fine)
[09:12] <abelli> my gnome-volume-manager just crashed, and it even didnt told me why!?!? 8)
[09:12] <andrewski> can anyone answer this question: why are the packages in the universe repository not quite current?  can users contribute?
[09:12] <andrewski> abelli: #ubuntu ?
[09:14] <rubenv> andrewski: check the wiki if you want to become a maintainer
[09:15] <andrewski> rubenv: that sounds a bit too extensive; i just want to contribute packages. :)
[09:16] <crimsun> abelli: hal+udev? ;)
[09:16] <crimsun> abelli: fwiw, a reboot resolved that for me
[09:16] <abelli> crimsun: the system is still alive,
[09:17] <abelli> gnome just told me that gnome-volume-manager went out for a walk
[09:17] <crimsun> abelli: yes, mine was, too, but the load avg was horrendous. ~18
[09:17] <abelli> 26%
[09:24] <thom> mdz: as far as I'm aware, alt+<num> has always changed tabs in ffox; if metacity isn't grabbing those keys first, i guess that's a change in metacity's behavious
[09:25] <schweeb> o_O
[09:25] <schweeb> I've never had ffox change tabs w/ alt+num
[09:26] <schweeb> always used ctrl+pgup/dn
[09:26] <schweeb> and it suddenly works for me as well
[09:26] <schweeb> and I'm using fluxbox
[09:27] <farruinn> erm, in ff it changes tabs with ctrl+kpnum for me
[09:27] <farruinn> (in warty)
[09:27] <farruinn> maybe firefox has gone from ctrl to alt?
[09:28] <thom> i've always used alt
[09:29] <mdz> thom: seb128 already blamed it on firefox :-)
[09:29] <mdz> thom: it's always been alt, but if they were bound in the window manager, it wouldn't eat them
[09:29] <mdz> thom: also, alt+tab used to work
[09:29] <mdz> and continues to work in other apps
[09:30] <thom> mdz: i can't reproduce alt+tab not working
[09:30] <thom> nor ctrl+alt+left/right
[09:30] <thom> which focus mode are you using?
[09:31] <mdz> mako: ping?
[09:31] <mdz> thom: defaults
[09:32] <thom> mdz: click to focus, then?
[09:32] <mdz> thom: yes
[09:32] <mdz> hmmm
[09:33] <mdz> they don't work on that desktop, even when firefox doesn't have focus
[09:33] <mdz> lemme try closing firefox
[09:33] <thom> even with click to focus (i use sloppy), i can't reproduce
[09:33] <mdz> what the hell
[09:33] <mdz> ok, the problem is redefined as "window manager shortcuts no longer work on desktop #5"
[09:33] <thom> whacky
[09:33] <mdz> but continue to work everywhere else
[09:33] <thom> that's, um, impressive
[09:34] <mdz> ok, closing all windows on that desktop and diddling about a bit seems to have gotten it back
[09:34] <mdz> even after reopening firefox
[09:34] <mdz> so RESOLVED/WTF
[09:35] <thom> i'm so closing that bug with this chunk of irc log ;-)
[09:35] <mdz> probably a metacity bug in there somewhere
[09:35] <mdz> but there is no hope now
[09:36] <thom> yeah, that's truely a bizarre one
[09:36] <thom> GAH, HAL is spinning my cdrom drive
[09:38] <dholbach> thom: and gamin keeps sitting on the cd
[09:38] <dholbach> i "kill -s SIGHUP"ed it 5 times today
[09:39] <fabbione> hi guys
[09:39] <dholbach> hi fabbione
[09:40] <schweeb> thom: since you're here, I'll consult you before bugzilla, to see if it's fixed... when I click on an XPI in the newest ffox in hoary, it doesn't open it w/ ffox... I have to save as, then go to file->open file to install an XPI extension
[09:40] <thom> schweeb: is that a change in behaviour recently?
[09:41] <schweeb> afaik, happened with the last update... been a while since I've installed an extension, I guess
[09:41] <thom> schweeb: (I have not extensions or translations installed, and have never done so that i can recall. so i'm not familiar with the behaviour)
[09:41] <schweeb> but usually, I clicked on the link, it'd download it, then I'd get a box that would install it
[09:42] <schweeb> in older versions
[09:42] <schweeb> whiprush noticed the same thing
[09:42] <thom> very strange
[09:42] <schweeb> thom: yea, I don't usually use extensions, but I tried Jybe last night
[09:42] <schweeb> which is pretty neat.
[09:42] <thom> file a minor bug, with a url to an extension and i'll take a look
[09:43] <schweeb> alrighty
[09:43] <thom> well, i'll have dinner and then maybe ;-)
[09:52] <schweeb> thom: hrm, now that I look, it may be a MIME thing... cause extensions from texturizer will install
[09:53] <andrewski> ogra or haggai... hello?
[09:54] <Kamion> T-Gone: pong. with regard to your man problem, check your locale on either side; I'll be willing to bet that there's a mismatch between the locale that the terminal's running and the locale environment variables set on the remote end of the ssh connection.
[09:54] <Kamion> Mithrandir: I've already mailed T-Bone about it, but FYI, Hoary's installer does not use discover any more.
[10:06] <farruinn> would much come of reporting a bug on warty?
[10:12] <farruinn> I'm just wondering if there's much point in doing so since most of your efforts are probably focused on hoary
[10:14] <dholbach> what did you want to report? maybe it's still an issue for hoary
[10:14] <Hwolf> farruin: ask someone to see if the bug is still in hoary, then report it.
[10:15] <dholbach> farruinn: warty is closed, just security fixes are pushed back in
[10:15] <farruinn> Hwolf: would this be an appropriate channel to ask in?
[10:16] <farruinn> or maybe I'll just post to the mailing lists...
[10:16] <Hwolf> farruin: ask in #ubuntu. Some people here can be picky.
[10:17] <farruinn> ok, thanks all =)
[10:17] <dholbach> farruinn: just tell what the bug was about
[10:18] <farruinn> the stickies applet doesn't remember the widths I set across logins
[10:18] <farruinn> I'll see if anyone else has noticed this
[10:19] <dholbach> farruinn: have a look at bugs.gnome.org too
[10:19] <farruinn> ah, right, I hadnt' thought of that
[10:21] <jdub> yo Kamion 
[10:24] <dholbach> who wants to test some g*mm-packages? :-)
[10:25] <Hwolf> dholbach: I've had 3 hoary installs give out on me in 48 hours. I'll wait.
[10:25] <seb128> dear thom, stop reassigning your firefox bugs on epiphany, kthxbye
[10:25] <jdub> hrm, something wrong with uploads?
[10:25] <jdub> morning seb128!
[10:25] <seb128> evening jdub :)
[10:25] <jdub> i woke up this morning
[10:26] <jdub> and saw that seb128 had fixed a gtk bug!
[10:26] <Kamion> jdub: yo, but I have to go again :)
[10:26] <jdub> hrm
[10:26] <jdub> do i accept a webcomp entry sent to ubuntu-announce? :)
[10:27] <seb128> jdub: stop trolling you know that gtk has no bug :p
[10:28] <jdub> Connection failed, aborting. Check your network (110, 'Connection timed out')
[10:28] <marcin_ant> jdub: Hi! I have a short question about website competition: are you going to publish previews of already submitted projects?
[10:28] <jdub> bong!
[10:28] <jdub> marcin_ant: only once it's closed
[10:30] <marcin_ant> jdub: hmmm ok. So, I have to submit my project anyway...
[10:32] <herzi> jdub: ping
[10:32] <jdub> yo herzi 
[10:32] <marcin_ant> jdub: BTW could you give me an answer on short and propably naive question about zope and plone?
[10:33] <herzi> the link to jimmacs page on planet.g.o directs one to his rss feed (i don't think it's supposed to do that)
[10:33] <jdz_> marcin_ant: shoot :)
[10:33] <jdub> marcin_ant: might be able to ;) i'm pretty short and naive about plone ;)
[10:33] <jdub> herzi: his rss feed possibly provides that link
[10:33] <marcin_ant> ok - then - what is default username and password to login to manage application?
[10:34] <jdub> no idea
[10:34] <jdz_> marcin_ant: there is no default.  you set one when you create a zope instance.
[10:34] <marcin_ant> first thing is that currently (probably to make website competition more interesting) zope and plone packages for hoary are broken
[10:35] <jdub> marcin_ant: might want to ask doko about that when he's aroudn
[10:35] <jdz_> marcin_ant: The zope package in hoary works fine, I'm using it right now.  I don't use ubuntu's plone though.
[10:35] <marcin_ant> jdub: I have packages from http://nathan.faho.rwth-aachen.de/debian/zope/
[10:36] <marcin_ant> jdub: so no problem
[10:36] <jdub> "no problem"?
[10:36] <jdub> dude, you just said plone was broken in hoary
[10:36] <jdub> that is a problem! :)
[10:36] <jdz_> marcin_ant: how are they broken?
[10:36] <marcin_ant> jdub: no problem for me because I have packages but not from hoary repo
[10:37] <marcin_ant> jdz_: no python2.3-xml
[10:37] <marcin_ant> jdz_: just a moment
[10:37] <marcin_ant> jdz_: I'll remove, refresh and try to install again
[10:39] <thom> dear seb, when you fix gtk i'll buy you dinner, love thom
[10:40] <seb128> ah ah
[10:40] <thom> (and it _is_ an epiphany bug, since the crash doesn't happen in firefox but does in ephy ;-) )
[10:40] <seb128> have you read my comments ?
[10:40] <seb128> happen in firefox here
[10:40] <seb128> and the bt come from firefox !
[10:41] <thom> i could not get firefox to break at all
[10:41] <thom> but ephy did it first time
[10:41] <seb128> firefox/mozilla/galeon/epiphany crash here
[10:41] <seb128> every single time
[10:41] <seb128> hello world
[10:41] <seb128> bottom
[10:41] <seb128> right
[10:41] <seb128> BOUM
[10:42] <thom> it did it that time
[10:43] <thom> i tried it like 10 times early and it didn't break once
[10:43] <thom> fucken odd
[10:43] <thom> oh well, my apologies
[10:43] <seb128> :)
[10:43] <seb128> that's fine don't worry :)
[10:44] <sjoerd> thom: he still gets the dinner right ?
[10:44] <seb128> but I don't know why the bt has no details, I've rebuilt firefox in nostrip nopopt
[10:44] <seb128> noopt
[10:44] <seb128> sjoerd: good point :)
[10:44] <jdub> seb128: 'BOUM' is a french explosion?
[10:45] <seb128> jdub: correct
[10:45] <seb128> what's the english version ?
[10:45] <jdub> "BOOM"
[10:45] <seb128> BOOM ?
[10:45] <seb128> right
[10:45] <marcin_ant> jdz_: I can confirm
[10:45] <tseng> jdub!
[10:45] <jdub> probably because most of our explosions are conventional, not nuclear. :-)
[10:45] <thom> sjoerd: no, he's not fixed gtk yet
[10:46] <marcin_ant> jdz_: on my system zope requires python2.2-xml but this package is not installable
[10:46] <jdub> sjoerd: would you like to make soft, tender debian love to my gamin packages?
[10:47] <sjoerd> jdub: if they are sweet and nice i will
[10:47] <jdub> they're tender and juicy!
[10:47] <sjoerd> bring them on then!
[10:48] <thom> jdub: are they nicely rounded?
[10:48] <thom> in all the right places?
[10:49] <thom> seb128: i guess i'm going bugzilla trawling on monday, then
[10:49] <dilinger> clearly i'm maintaining the wrong packages
[10:49] <seb128> ok
[10:49] <dilinger> mine are all old and crusty, w/ sharp edges that cut me
[10:49] <dholbach> dilinger: so everyone has different dispositions, it seems ;-)
[10:49] <tseng> dilinger: meh, -as2 is pretty smooth
[10:49] <thom> dilinger: linux, php and mod_perl. what more needs saying
[10:50] <dilinger> tseng: working on -as3 now :)
[10:50] <marcin_ant> jdz_: how it is available that you have zope installable?
[10:50] <thom> dilinger: you didn't exactly pick loving upstreams ;-)
[10:50] <dilinger> heh
[10:52] <jdz_> marcin_ant: I apt-get'ed it a few months ago, and have been using it daily ever since.  I'm kind of suprised it would be broken.  Do you get an error of some sort I could look into?
[10:53] <jdub> great film!
[10:53] <jdub> thom: btw, greeno says thanks very much
[10:53] <thom> and, aaaargh, I just tried to put Michael Banck into IMDB when I meant ben affleck. CURSE YOU ELMO
[10:54] <jdub> thom: he mailed, but wasn't sure if he had the right address
[10:54] <thom> jdub: i need to reply to his email
[10:54] <thom> yes
[10:54] <jdub> ah
[10:54] <jdub> we visited yesterday
[10:54] <jdub> the sprog seems to be human
[10:54] <thom> cool! how are they doing?
[10:54] <thom> heh
[10:54] <marcin_ant> jdz_: what error do you want?
[10:54] <jdub> good, "life is substantially different all of a sudden"
[10:54] <jdub> :)
[10:54] <dilinger> john cusack can do no wrong
[10:54] <jdz_> marcin_ant: how is it broken?  maybe you should just file a bug ;)
[10:54] <marcin_ant> jdz_: it's like I said - apt-get install zope and you simple cannot install this
[10:55] <marcin_ant> jdz_: because there is no python2.2-xml
[10:55] <marcin_ant> jdz_: just problem with dependencies
[10:55] <marcin_ant> jdz_: of course I can paste error message for you but I have system with polish locales 
[10:56] <marcin_ant> jdz_: so you propably don't want error message in polish?
[10:57] <thom> marcin_ant: ok, so we should change the defaul to point at zope2.7, which installs fine
[10:58] <thom> please file a bug
[10:58] <thom> jdub: heh, surprise
[10:58] <thom> dilinger: *nod*
[10:59] <jdz_> ah, yes!  right.  I have zope2.7 intsalled.
[11:00] <jdz_> marcin_ant: yes, the zope package seems to be broken ;(  zope2.7 works :)
[11:00] <marcin_ant> jdz_: sure - but then remember about plone
[11:00] <sjoerd> jdub: just lemme know where they are when your done
[11:01] <marcin_ant> jdz_: because I can currently install zope-2.7 but then plone is not installable
[11:01] <jdz_> marcin_ant: right.  sounds like plone's dependencies need to be fixed up to use zope2.7.  file a bug like thom sugested :D
[11:04] <marcin_ant> jdz_: ok - but then... how to create skin for ubuntu website contest?
[11:05] <jdub> marcin_ant: see the announcement, you don't have to create a skin
[11:05] <jdz_> marcin_ant: go to http://plone.org/downloads/ and grab the Plone Core tarball :D  - untar it into your zope instance's products directory, and bingo!
[11:07] <marcin_ant> jdz_:  so I should install zope-2.7 and unpack plone core, right?
[11:07] <jdz_> marcin_ant: thats what I do :D
[11:07] <marcin_ant> jdub: yes I know but I prefer working on something "real"
[11:07] <dholbach> cooool: i never saw this one: We trust you have received the usual lecture from the local System Administrator. It usually boils down to these three things:     #1) Respect the privacy of others.       #2) Think before you type.      #3) With great power comes great responsibility.    (sudo on debian)
[11:08] <jdub> marcin_ant: puts you at a serious disadvantage :)
[11:09] <jdub> sjoerd: http://www.gnome.org/~jdub/debian/
[11:10] <marcin_ant> jdub: :)
[11:13] <mdz> jbailey: ping?
[11:18] <thom> dilinger: http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=110683815626149&w=2
[11:18] <thom> dilinger: seriously, please get involved on that thread :-)
[11:24] <jbailey> mdz: Here.
[11:28] <mdz> jbailey: since you happened to be working on hotpluggish stuff, I was wondering if you had an opinion on #1763
[11:29] <mdz> I just always have bugzilla tabs open, so the upper-right box on most bugzilla pages serves the purpose for me ;-)
[11:30] <thom> by the way, bugzilla.ubuntu.com/1763 DTRT
[11:32] <jbailey> mdz: Yup, this is hotplug suckage.  It doesn't actually probe the IDE bus, which is pretty much what you'd need.
[11:33] <jbailey> And whoosh, I see that further down.
[11:33] <jdub> thom: see the mails re: default firefox theme on ubuntu-user?
[11:34] <jbailey> mdz: ide-generic is necessary still.
[11:34] <jdub> thom: gnomestripe is nowhere near as good as industrial, either :|
[11:35] <dholbach> brb
[11:36] <jbailey> mdz: I'd far prefer to see Marco's patch included in hotplug.  It would be correct, I think, to just probe the bus that way.  SCSI might need the same thing for tape devices, too.  I haven't run a pure udev hotplug system with a scsi tape.
[11:36] <sjoerd> jdub: did you try the gamin packages on a debian box ? my nautilus hangs on startup with them..
[11:37] <sjoerd> jdub: still checking what the problem is.....
[11:37] <mdz> jbailey: SCSI dtrt already
[11:37] <mdz> jbailey: I guess there are two  issues
[11:37] <mdz> jbailey: one is that the ide-* modules aren't loaded based on devices found on the bus (that bit is addressed by marco's patch)
[11:37] <mdz> jbailey: the other bit is ide-generic needing to be loaded at the right time
[11:38] <jbailey> Lemme do a quick check, but I don't think ide-generic needs to be loaded in any particular order, I think it just needs to be loaded.
[11:38] <mdz> I'm fairly certain it needs to be loaded after the chipset-specific driver
[11:38] <mdz> after _all_ chipset-specific drivers, even
[11:38] <jbailey> Right now, my initramfs setup is loading piix, ide-disk, then ide-generic.  Lemme drop in ide-generic first and see what happens.
[11:40] <jbailey> mdz: While that's rebooting, I got a full nfs boot setup working yesterday and confirmed today on #ltsp that there isn't a 2.6 swap over NFS patch around.
[11:41] <T-Gone> Kamion: i'm sorry but my locale is C in both place, and I don't see how there would be a 'locale mismatch' on the own machine's tty
[11:41] <T-Bone> Kamion: last but not least, manpages are in english, and the troubles doesn't happen everywhere
[11:41] <T-Bone> and f*ck i can't write proper english anymore :P
[11:42] <mdz> jbailey: what about nbd?
[11:43] <jbailey> mdz: Haven't tried it yet.  The ltsp pages I had found first mentioned swap over NFS.  Since I was already using NFS, it made sense to try and minimise the number of protocols I was working with.
[11:43] <T-Bone> s/everywhere/in every manpage/ that is
[11:43] <T-Bone> +s
[11:43] <T-Bone> damn i got a problem with plural forms :P
[11:46] <mdz> T-Bone: it sounds an awful lot like UTF-8/iso-8859-1 mismatch
[11:46] <dholbach> can any one give me some gdb-superpowers to get coaster going in ubuntu? :-)
[11:47] <T-Bone> mdz: well i tend to think about that, but then I'm pretty puzzled by the fact that it only affects a few manpages
[11:47] <mdz> T-Bone: which ones?
[11:47] <T-Bone> mdz: besides, i don't see how/why that would mess with things both on local ttys and on remote logins
[11:47] <T-Bone> mdz: a few examples are: lspci, less, mkfs...
[11:48] <mdz> which part of the lspci man page?
[11:48] <T-Bone> mdz: i've tried a few at random and it happened several times
[11:48] <T-Bone> all parts
[11:48] <T-Bone> mdz: it doesn't affect gnome terminal or xterm on hoary
[11:48] <jbailey> mdz: At a quick glance it appears that ide-generic can be loaded before the piix driver.  My other machine is an sis5513.  I'll try that too.
[11:48] <T-Bone> mdz: try them on your local VTs
[11:48] <T-Bone> mdz: you'll immediately understand what i'm talking about
[11:49] <mdz> T-Bone: ok, I didn't see that you were talking about the console
[11:49] <mdz> that's an entirely different set of issues
[11:49] <mdz> probably the unicode hyphen character is missing from the font, or some such
[11:49] <T-Bone> mdz: but wth would that affect only a few manpages?
[11:50] <mdz> T-Bone: it will affect any man page which uses '-'
[11:51] <mdz> jbailey: hmm...you're using an Ubuntu kernel?
[11:51] <T-Bone> mdz: no
[11:51] <T-Bone> mdz: try man ls
[11:51] <T-Bone> man man
[11:51] <jbailey> mdz: Yes.
[11:51] <T-Bone> or whatsoever
[11:51] <mdz> T-Bone: try looking at the source for those man pages
[11:51] <mdz> T-Bone: they use '\-' not '-'
[11:51] <T-Bone> o_O
[11:52] <mdz> they are two different characters
[11:52] <T-Bone> /o\
[11:52] <T-Bone> mdz: well it's sort of a bug anyway you put it
[11:52] <T-Bone> mdz: some one remotely logged from a non hoary machine will get all messed up
[11:53] <mdz> T-Bone: yes, it certainly is a bug.  if you can confirm the cause and help to fix it, that would be fabulous
[11:53] <T-Bone> and not being able to use local VTs properly is definitely not a good thing imho (consider the server-side of things, no pretty-GUI thingy there)
[11:53] <T-Bone> err
[11:53] <mdz> my guess is that it's a font problem
[11:54] <T-Bone> can I behave like a normal user every once in a while? :)
[11:54] <T-Bone> like "i report a bug, not my problem" :P
[11:54] <mdz> not in this channel :-P
[11:54] <T-Bone> *sigh*
[11:55] <mdz> jbailey: hmm, let me try this
[11:57] <mdz> hmm, there's no way to unload the chipset-specific driver, is there
[11:57] <T-Bone> mdz: the only thing i can tell so far is that it's something new, and definitely a regression from warty (just checked, it doesn't happen)
[11:57] <mdz> T-Bone: hoary uses utf-8, warty doesn't
[11:57] <T-Bone> that's a hint
[11:57] <jdz_> T-Bone: I can confirm in, I have the same issue :)
[11:57] <mdz> I expect the same bug would exist in Warty if you use a UTF-8 locale
[11:57] <T-Bone> jdz_: ok then i pass it to you! Find and fix the bug :)
[11:58] <jbailey> mdz: You mean forcably unload?
[11:58] <jbailey> mdz: otherwise, they can be rmmod'd fine.
[11:58] <thom> jdub: i don't agree, actually. I much prefer gnomestripe
[11:58] <mdz> jbailey: hmm, something is holding a reference to mine
[11:58] <mdz> and I can't see what
[11:58] <mdz> via82cxxx              13852  1
[11:59] <jbailey> mdz: That's what the current initrd does at startup.  It loads all the ide drivers, mounts the filesystem, and tries to unload them all.  Whatever is left is yours. =(
[11:59] <mdz> I've unloaded ide-* (except ide-core, which is used by via82cxxx)
[12:00] <jbailey> IF your partition is mounted on it, it'll hold the reference internally.
[12:00] <thom> jdub: also, someone (schweeb or eruin?) asked about getting winstripe as an option, so i'll look at doing that later
[12:01] <jbailey> Interesting. It looks like ide-generic itself is finding the ide bus on my boxes, and that the chipsets are then saying no bus'.
[12:01] <mdz> jbailey: yeah, that's the thing
[12:01] <jbailey> Feh.
[12:01] <mdz> ide-generic needs to be loaded last
[12:02] <jbailey> Nasty.
[12:02] <mdz> we could load it at the end of hotplug startup or something awful like that