IRC chat logs for #ltsp on irc.libera.chat (webchat)


Channel log from 10 April 2008   (all times are UTC)

00:07asac has quit IRC
00:07asac_ is now known as asac
00:12mccann_ has quit IRC
00:13Patina_ has joined #ltsp
00:33
<dberkholz>
johnny: i've seen you having fun with the fetches
00:33
<johnny>
:(
00:33* johnny is suprised the perl herd was so out of date
00:34
<johnny>
2 or 3 versions behind or something
00:34
on two seperate packages.. THAT WERE REQUIRED BY GIT
00:34
<dberkholz>
were they broken?
00:34
<johnny>
no
00:34
upstream src is gone
00:34
<dberkholz>
they should've been on gentoo mirrors
00:34
if they're in the tree
00:35
<johnny>
hm.. yeah somethign is going on with the mirrors
00:35
oh..
00:35
maybe i broke something
00:35
but either way.. the upstream author wasn't providing that version anymore
00:35
hmm.. what did i do..
00:35
<dberkholz>
did you set GENTOO_MIRRORS="" or something?
00:36
<johnny>
i knew it ! :(
00:36
i was just thinking that
00:36
i've never used GENTOO_MIRRORS
00:36
the defaults have always been fine
00:36
so.. didn't know setting it empty would do that :(
00:37
ok.. looks like i need to adjust the profile one more time
00:37* johnny feels really dumb
00:37
<johnny>
ok.. well at least i know how to fix it
00:38
<dberkholz>
that's the hard part, the rest is just waiting
00:38
<johnny>
what's the recommended way to get the contents of the owner portage tree
00:38
err host*
00:38
UGH
00:39
no.. make.conf variables
00:39
<dberkholz>
i'll wait for that to make sense =)
00:39
<johnny>
so i can ask make.conf what it has in the host
00:39
GENTOO_MIRRORS for example
00:39
<dberkholz>
johnny: i don't understand what you're trying to do
00:39
<johnny>
i want to get GENTOO_MIRRORS from make.conf
00:40
from /etc/make.conf to /opt/ltsp/'s one
00:40
<dberkholz>
ok, i see
00:41
there are many possible ways of doing this
00:41
<johnny>
i was going over the sed oneliners page, but my sed knowledge was never very expansive..
00:41
<dberkholz>
one way: in a subshell, source /etc/make.conf and echo whatever variables you care about
00:42
another way: create a link of make.conf in the client root, and add a line at the top of the client make.conf that does 'source /etc/make.conf.server' or so
00:42
<johnny>
hmm.. i did that somewhere..
00:44
<dberkholz>
another way: just make people reconfigure for the client root, there probably aren't that many identical settings
00:44
another way: when you start the build, copy in the server's make.conf and then start tweaking it
00:44
<johnny>
no.. i just want GENTOO_MIRRORS
00:44
<dberkholz>
you probably want more than that
00:44
<johnny>
really?
00:44
like?
00:44
this is for install
00:44
they can tweak things after
00:45
<dberkholz>
why not set up as much as is reasonable?
00:45
<johnny>
i haven't?
00:45
<dberkholz>
if we've gotta do one thing, might as well do 'em all
00:45
<johnny>
i think i have been
00:45
<dberkholz>
PORT_LOGDIR, FEATURES, SYNC, CFLAGS/CHOST/ACCEPT_KEYWORDS (on same-arch builds, anyway), LDFLAGS, MAKEOPTS
00:46
CCACHE_DIR, CCACHE_SIZE, CLEAN_DELAY, maybe USE
00:46
<johnny>
some of those can be done during the initial install atm, others not
00:47
<dberkholz>
pretty much everything listed in make.conf.example that could be set in their make.conf
00:47
<johnny>
well.. some of those.. we should just link to the appropriate profiles
00:48
<dberkholz>
every variable i listed is one i have in my make.conf
00:48
not from profile
00:48
<johnny>
oops.. i forgot that make.conf comes from the stage3
00:49
not from the profile
00:49
CHOST and CFLAGS are in that
00:50
but i gotta leave something for you to do :)
00:50
i'm just trying to get it working
00:50
with near feature parity with ubuntu/debian
00:52
once that happens, we can add plenty more things i'm sure
00:53
we really need to fix genkernel so it will actually build an x86 kernel and initramfs from outside the chroot :(
00:54
it'll build the x86 kernel fine now ..
00:54
it doesn't do the right thing with UTILS_CROSS_COMPILE at all
00:54
so.. now busybox
00:54
err no busybox*
00:55
hey warren , are you about?
01:10cyberorg has joined #ltsp
01:13deavid has joined #ltsp
01:16makghosh has quit IRC
01:20makghosh has joined #ltsp
01:24indradg has quit IRC
01:49
<dberkholz>
johnny: i seem to be doing a better job of serving as a sort of expert resource than actually getting stuff done
01:50
<johnny>
is that how you want it?
01:51
sorry for charging ahead so much, i was just really excited
01:52
it's been fun to learn how it works
01:57u22 has quit IRC
02:11soneyka has joined #ltsp
02:21gvy_ is now known as gvy
02:21* gvy is back (gone 36:24:12)
02:22
<gvy>
halo
02:22
dberkholz, me too :)
02:27
<johnny>
and the perils of using upstream directly..
02:28
openrc git just got broke
02:30remitaylor has joined #ltsp
02:33ogra_cmpc has quit IRC
02:33ogra_cmpc has joined #ltsp
02:34
<remitaylor>
my thin client dhcp's properly from my ltsp-server but then it seems to re-dhcp from my gateway while booting and it loses its rootserver/rootpath info and drops to busybox ... anyone experienced this / know if i can prevent this? do i have to put something in my /opt/ltsp/i386/etc/dhcp3/dhclient.conf or ... something? using ubuntu 7.10 ... here's a screen of what it's doing: http://remi.org/ltsp-redhcp.png (gw=10.5.5.1,ltsp=.99,gw-dhc
02:37
<johnny>
yes.. heard of that. i'm sure somebody here will know the answer if you stick around a bit
02:38
<remitaylor>
whooooooo another human being! thanks :)
02:39
<johnny>
you're on at the wrong time sadly
02:39
most of the folks won't be about for a bit later
02:40
<remitaylor>
i hear that. ~ 1am my time, actually, but ... that's when all good IT work gets done around here :P
02:44
<johnny>
here too
02:44
gotta do the work when the store is closed
02:47
<remitaylor>
exactly. good for sysadmin work because the systems aren't in use. good for programming ... cause it's quiet
02:47
<johnny>
i wish i knew some other late night IT people locally..
02:48indradg has joined #ltsp
02:51
<remitaylor>
they're probably lurking in the shadows. who knows? the only thing you can know for sure is ... obviously _their_ ltsp setups are working :P
02:51
<johnny>
mine works.. mostly
02:51
i need to implement something so it doesnt' make the music skip on the server
02:54
<remitaylor>
i can't wait to have such problems ... it'll mean i've got it up and running okay
03:04mikkel has joined #ltsp
03:47makghosh has quit IRC
04:01
<ogra_cmpc>
johnny, there is only one answer to that question (at least for everything before ubuntu 8.04) ....
04:02
dont use more than one dhcp server in one network
04:03
hardy just recently got code to run dhcp on a different port for the adventourous minds :)
04:03
remitaylor, ^^^
04:07remitaylor has quit IRC
04:35
<laga>
ogra_cmpc: did ou upload a fixed ltsp?
04:36
<ogra_cmpc>
laga, didnt you get mail from the publisher ?
04:36
i even upoladed under your name
04:36
<laga>
ogra_cmpc: didn't get anything.
04:37
oh, thanks. even you made some more changes. :)
04:37
i left my name in there because i initially assumed no more changes on your part were necessary for that upload.
04:39
<ogra_cmpc>
yeah, the patch didnt apply properly, i had to fiddle a bit
04:39
(and it was late at noght after a 14h day, so please test the change)
04:41
<laga>
i had a 14h days as well, so i'll make sure to test it :)
04:41
i dont have my laptop handy right now so it'll have to wait
05:12
<warren>
johnny: ?
05:15
<ogra_cmpc>
morning warren
05:17
<warren>
hi
05:17
I woke up because of the pain
05:18
<ogra_cmpc>
mah, thats bad
05:20
<warren>
this is normal for me =(
05:21
<ogra_cmpc>
yeah, you said so before
05:21
it helps a lot to have a good bed though
05:23
<warren>
I have a good bed
05:23
the pain seems to be caused by sitting too much
05:23
for work
05:24
<ogra_cmpc>
work standing upright then ...
05:25
<laga>
or take a laptop to bed
05:25
<ogra_cmpc>
i know a company where they have only tables you can stand on and a treadmill at every workplace
05:25
so you stay fit all the time ... the CTO says it increased the productivity in a noticeable way
05:33kwak has joined #ltsp
05:34
<kwak>
hi, does anyone here know of a presentation about opensource not just ltsp but also using Linux. I will present open source and the use of Linux OS in our school in two weeks
05:35
I want to show them the PROs and the CONs and other information.
05:35
#k12ltsp
05:48muh2000 has joined #ltsp
05:51
<muh2000>
how to change volume ?
05:55
with "VOLUME = 100" in the lts.conf ?
05:59kwak has quit IRC
06:06jammcq has quit IRC
06:08viking-ice has joined #ltsp
06:13TelnetManta has quit IRC
06:21daduke has quit IRC
06:30ffhghgf has joined #ltsp
06:31soneyka has quit IRC
06:47jummel has joined #ltsp
06:47
<jummel>
hello?
06:49
does anyone know how i can set the default session in LDM so it logs into KDE rather than the strange iconless, barless DE\
06:49
?
06:52
is anyone here?
06:52
<ogra_cmpc>
jummel, distro ltsp version ?
06:52viking-ice has quit IRC
06:53
<jummel>
I'm running sidux
06:53
I think version is 5
06:53
<ogra_cmpc>
sidux is a debian clone, right ?
06:54
<jummel>
no, I think it's better described as managed debian
06:54
it draws packages directly from debian
06:54
<ogra_cmpc>
well, its a debian based one, thats what i was after
06:54
<jummel>
unlike ubuntu which has its own packages
06:54
<ogra_cmpc>
update-alternatives --config x-session-manage
06:54
*x-session-manager
06:54
select kde
06:55
<jummel>
do i chroot into the i386 directory first?
06:55
<ogra_cmpc>
no
06:55
on the server
06:55
ldm executes /etx/X11/Xsession ....
06:56
Xsession looks for ~/.xsession and if thats not here iot falls back to the x-session-manager alternative
06:56
(which defines teh systems default desktop)
06:57
(at least thats how it works in ubuntu and debian)
06:57
(not sure the sidux guys do something differently)
06:57
<jummel>
ok thanks
06:57
I'll give that a go
06:58
<ogra_cmpc>
do you know which repo sidux bases off ? sid or testing ?
06:58
(might be that the sid ltsp version is broken atm)
06:59jammcq has joined #ltsp
06:59
<jammcq>
g'morning
06:59
<ogra_cmpc>
hey jammcq
07:00dcordes has joined #ltsp
07:00cyberorg has quit IRC
07:00imperfect- has joined #ltsp
07:01
<jummel>
IT WORKED!
07:01
YAY!
07:01
I think Sidux always points to the sid debian repositories
07:01
otherwise the name would be pretty stupid
07:02
why sid is the name of a release level or something is even more confusing i think
07:02
anyway, thanks, CYA!
07:02jummel has quit IRC
07:02deavid has quit IRC
07:03
<ogra_cmpc>
i guess he didnt watch toy story :)
07:05TelnetManta has joined #ltsp
07:10cyberorg has joined #ltsp
07:17
<imperfect->
So I'm told you guys might be PXE ninja's?
07:21
<ogra_cmpc>
not really ... but we know about all the surrounding bits :)
07:22
whats your prob ?
07:23vagrantc has joined #ltsp
07:23
<imperfect->
Well
07:23
I'm doing somehting not related to LTSP
07:23
But I was told you guys were good with PXE ;)
07:23
Just trying to figure out how to correctly identify machine types via isc dhcp
07:24
Alas, there seems to be a two stage deal where it sends PXEClient:Arch:00000:UNDI:003016, before it sends the string I'm looking for in my match statement
07:24indradg_ has joined #ltsp
07:24
<ogra_cmpc>
well, thats rather dhcp then :)
07:24
<imperfect->
Aye
07:24
I know
07:24
Twas suggested I come her e;)
07:24
<ogra_cmpc>
see, its always the surrounding bits :)
07:25
<imperfect->
nod nod
07:25
Any ideas?
07:25
;)
07:25
<ogra_cmpc>
you can match against vendor-class-identifie
07:25
r
07:25
<imperfect->
Right
07:26
<ogra_cmpc>
http://people.ubuntu.com/~ogra/dhcpd.conf
07:26
<imperfect->
And what I'm saying is vci comes back PXEClient:Arch:00000:UNDI:003016
07:26
for like maybe 4 or 5 DHCPOFFERS
07:26
<ogra_cmpc>
that at least gives you distinction between ppc and x86 arches
07:26
<imperfect->
then it switches to the one I want
07:26
<ogra_cmpc>
wait for Gadi, he's deep into such stuff
07:26
(and should show up soon (new york time))
07:27
<imperfect->
AAPLBSDPC/i386 <-- looking for that
07:27
It switches option 60 in the midst
07:27
and then starts the pxe bios id again
07:27
rinse, repeat.
07:28
The real issue I have here, as I see it, is that I've got a shitton(tm) of LWAPP's
07:28
PCs that need PXE service and these apples.
07:28
The LWAPPS won't be a problem. They play nices and always send option 60 like I'd expect.
07:29
So I think my plan will be to class for Apples, Cisco LWAPPS and have it fall through to straight PXE ;)
07:29
Sound reasonable?
07:29makghosh has joined #ltsp
07:30elisboa has joined #ltsp
07:34viking-ice has joined #ltsp
07:42indradg has quit IRC
07:43slidesinger has joined #ltsp
07:45makghosh is now known as makghosh|afk
07:46Guaraldo has joined #ltsp
07:50mhterres has joined #ltsp
08:01deavid has joined #ltsp
08:03imperfect- has left #ltsp
08:11praveer_cool has joined #ltsp
08:22cliebow has joined #ltsp
08:40Gadi has joined #ltsp
08:50makghosh|afk is now known as makghosh
08:52
<gvy>
halo
08:52
vagrantc, ayt?
08:54
<vagrantc>
gvy: hmmm?
08:54* ogra_cmpc dances around his new lappie
08:54
<gvy>
vagrantc, re merge. :) if you might find some more time, would be glad to understand when to show up
08:54
<ogra_cmpc>
just in time for installation testing :)
08:55
<gvy>
i've just sorta got from under the pile of rub^H^H^Hjob and led too, so just discussed next merge and what should be better to do
08:55
ogra_cmpc, 'grats :)
08:56
vagrantc, so far i think that doing functions hook (for distro-specific functions) and moving part of stuff deeper for now should be an easier start
08:56
so that when folks have time, distro-specific scripts can be reviewed and if reasonable/more generic, generalized
08:57
so that it's no specific anymore
08:57
<vagrantc>
sure
08:58
i also think that some functions need to remain distro-specific, but we should use common function names that behave the same way
08:58
<gvy>
yeah
08:58
<vagrantc>
i.e. get_arch () { dpkg --print-architecture }
08:58
for debian and ubuntu
08:58muh2000 has quit IRC
08:59
<vagrantc>
get_arch () { case $(uname -a) ; i*86) echo FOO ;; esac }
08:59
<gvy>
or rather for "standard dpkg-based distros" -- e.g. mepis might jump in some day too
08:59
<vagrantc>
gvy: well, ubuntu may even want a different function ...
08:59
<gvy>
but the hierarchy of differences is like locks: they can be more fine-grained over time, not immediately :)
08:59
<vagrantc>
i.e. i think ubuntu defaults to installing i386 on amd64, for example.
09:00
gvy: but at any rate ... you have any new branches to propose for merging? or you want to discuss first before working on them?
09:01
gvy: there were some remaining changes as yet unmerged, because they came mixed in with some changes that i definitely don't want to merge without more discussion and probably re-writing them a bit.
09:02
it's been a while since i looked at them, but that's what i remember.
09:04
<gvy>
vagrantc, well one was about boolean_is_{true,false} and somewhat philosophic discussion on trating defaults
09:04
when you told that each and every default value is to be defined in a common config (which is reasonable and was omitted by me while thinking things over beforehand)
09:05
another one was an obvious bugfix for selinux check, i've posted test cases for that sort of shell scripting mistake i used to repeatedly make too :)
09:06
<vagrantc>
gvy: i think the selinux check was fixed differently.
09:06
<gvy>
that is, [ ... ] && ... as a last command might result in whole script's exit code being non-zero
09:06
vagrantc, well, i hope :)
09:07
vagrantc, so i think that for the functions part, we'd do a hook and put it under ALTLinux/ atm
09:07
is that ok with you? (with hopefully a merge-up some day)
09:08
<vagrantc>
gvy: depends on how it's implemented i guess.
09:08
<gvy>
indeed
09:08
i mean the direction
09:08
<vagrantc>
gvy: have you looked at the current bzr of ltsp-trunk ?
09:08
<gvy>
vagrantc, i have beein looking into image build system and corresponding hooks/bugs/packages for these two weeks...
09:09
*been
09:09
will have a look as led syncs up, probably
09:09
<warren>
did you even LOOK at it recently?
09:09
the selinux line wont exit the script non-zero
09:10
<vagrantc>
gvy: just to keep in mind, please keep proposed upstream changes in different branches than altlinux-specific stuff
09:10
<gvy>
warren, no as i've said :) but hope it's all ok there by now
09:10
<vagrantc>
gvy: and ideally, each proposed upstream change in it's own branch
09:11
<gvy>
vagrantc, ugh. i understand why but with that heavy branches it's going to be so slow :( okay, we'll try
09:12
<warren>
gvy: if you don't split out the alt specific stuff from upstream rewrites then the alt specific stuff will get in much slower.
09:12
if you intend on forking permanently then the latter is fine.
09:13
<gvy>
warren, i wouldn't spend time and bytes then :]
09:14
<vagrantc>
gvy: the easier it is for us to review a merge, the quicker it will go in. it's very simple.
09:14
<gvy>
vagrantc, warren, yeah, i do understand
09:14
and it's all reasonable :)
09:15
<vagrantc>
i feel like i've already spent too many hours trying to sort out what's what.
09:15
<gvy>
maybe :(
09:16
<vagrantc>
i'm debating weather to make my ltspfs changes in trunk or not ... it's a pretty significant change from previous behavior.
09:16slidesinger has quit IRC
09:16
<vagrantc>
but i think it's the right direction
09:16
calling both cdpinger and ltspfsd from udev, rather than from the ltsp-client* init scripts (or wherever you call them from)
09:25deavid has quit IRC
09:29ffhghgf is now known as sunny
09:29sunny is now known as sunnyrise
09:36
<gvy>
vagrantc, so we'll do a functions hook as discussed back then (in a separate branch)
09:37
like
09:37
<Led> [ -f /usr/share/ltsp/finctions-$VENDOR ] && . /usr/share/ltsp/finctions-$VENDOR
09:38
<vagrantc> basically, yes. although we might want to hard-code $VENDOR when building the package.
09:38
<gvy> vagrantc, i just told "build time" here 8)
09:38
<vagrantc>
lets skip $V?ENDOR
09:38Blinny has joined #ltsp
09:38
<gvy>
vagrantc, well, maybe you do the hook then as you see best fit? i think we're ok with any working solution :)
09:39
since we might just lose more time waving branches if pulling from ftp.linux.kiev.ua is that slow
09:40
was sore about that the last time :(
09:42
<vagrantc>
gvy: well, i'd rather not have to detect $VENDOR at run-time ... seems easier to handle by installing at build-time
09:43
so ...
09:43
<gvy>
vagrantc, we seem to have already agreed hardwiring is the way to do it :)
09:43
<vagrantc>
if [ -f /usr/share/ltsp/functions-distro-specific ] ; then . /usr/share/ltsp/functions-distro-specific ; fi
09:43
<gvy>
ok
09:43
<vagrantc>
although i'm open to a better name ...
09:44
but that can go at the bottom of /usr/share/ltsp/ltsp-common-functions
09:44
<gvy>
seems good enough
09:44
yup
09:45
<vagrantc>
thus allowing you to override common functions without getting overly complicated.
09:45mccann has joined #ltsp
09:45
<vagrantc>
currently, i on;y see boolean_is_true in common functions
09:46
<gvy>
vagrantc, led suggests "ltsp-vendor-functions" as more in line with "ltsp-common-functions"
09:46
<vagrantc>
sounds good
09:46
<warren>
so I replace the contents of /usr/share/ltsp/functions-distro-specific?
09:47
<gvy>
yup
09:47
<vagrantc>
warren: each distro can optionally include or not include the file ... thoughts ?
09:47
<warren>
it is OK because it doesn't have distro names in the file/directory names
09:47
<vagrantc>
the other idea i've had is to merely append to ltsp-common-functions at build time
09:47
<gvy>
vagrantc, led started to write if [ -d ltsp-vendor-functions.d/ ]; then ... :) -- "just to be done with it"
09:48
<vagrantc>
gvy: well, i'd like to avoid doing a ltsp-vendor-functions.d ... but that can easily be added in the ltsp-vendor-functions on a per-distro basis
09:48
avoid mandating it
09:49
<warren>
vagrantc: I hope this function override thing doesn't go in until May at the earliest?
09:49
<vagrantc>
warren: why?
09:49
<warren>
vagrantc: I'm still pushing trunk directly into Fedora
09:49
<vagrantc>
warren: you're not frozen yet?
09:49
<warren>
well, frozen means different things here...
09:49
<vagrantc>
ah.
09:49
more like slushy?
09:49
<warren>
vagrantc: ok, not THAT long, but I need at least a week
09:49bobby_C has joined #ltsp
09:50
<vagrantc>
warren: it's a trivial change...
09:50
<warren>
if you're certain it wont break anything, I guess.
09:50
<gvy>
warren, it doesn't break anything for you :)
09:50
<vagrantc>
warren: it will result in one additional filesystem call if you don't actually use it.
09:51
<gvy>
vagrantc, re .d -- reasonable, and if we find some day it's too widely implemented, then it's be time to pull upwards
09:51
(if ever)
09:52
<vagrantc>
gvy: exactly.
09:52
gvy: last i looked, you had a custom run_parts_list function ... ?
09:53
<gvy>
vagrantc, led says that was custom long ago, but he moved to common code by now
09:55
<vagrantc>
it looked like my ugly hack for the fedora branch before warren started actively working on it
09:55
<gvy>
ah
09:56
well, i don't remember who started the fedora adaptation effort but if that was you, my hat actually off
09:56
<vagrantc>
committed and pushed
09:56
gvy: well, i did a quick and dirty "this builds a fedora chroot and in theory enables the initramfs to do network boot" proof of concept
09:57
<gvy>
/nick led
09:57
<vagrantc>
because people kept asking about it, and figured if someone who knew debian could get that much, then people who knew fedora could actually make it happen.
09:57
<warren>
hmm, something changed in rawhide in the last week and broke LTSP
09:57
<gvy>
"first it didn't fit us, but then some day i saw it's ok"
09:57
/nick gvy
09:58
warren, if you build images to test then build logs might help figure out the package difference
09:58
if not... then maybe upgrade logs but that's harder to me
09:58
<warren>
gvy: hundreds of packages changed in the last week
09:58
<gvy>
ah, slushy.
09:59vlt has joined #ltsp
09:59
<gvy>
we use to do slushy freezes (like "is it freeze-freeze or freeze-freeze-freeze?" (c) qa@)
09:59
<warren>
no, we just hit freeze
09:59
<gvy>
as well :-/
09:59
<vagrantc>
warren: what sort of breakage?
09:59
<gvy>
ah, then wish you nice anti-cyclone :)
10:00
<warren>
vagrantc: most seriously, it gets stuck during bootup before it gets to initscripts
10:00
<vagrantc>
i guess i'll postpone my push to get ltsp_chroot functionality in ltsp-trunk before ubuntu and fedora totally freeze up
10:00supreme has joined #ltsp
10:00
<supreme>
hi all
10:00
<warren>
vagrantc: or rather until we finalize what we're pushing
10:01
<gvy>
hi supreme
10:01Q-FUNK has joined #ltsp
10:01
<vagrantc>
i need to get new upstream versions of ltsp and ltspfs into debian before we start hitting some freeze cycles.
10:01
<supreme>
somebody knows about 'Corrupted MAC on input' error?
10:01
<vagrantc>
ldm, too, though i'm not much worried about that.
10:02
<warren>
supreme: dude, I already told you what the problem might be last night.
10:02
<gvy>
supreme, quick google seems to know something, i haven't hit that
10:02
<warren>
supreme: you need to talk to your distro's kernel people
10:02
<vlt>
Hello. Any idea where to find the time when a user has logged out? auth.log just shows me logins.
10:03
<supreme>
warren, i asked again cause now there is more people than last night
10:03
<warren>
supreme: your problem is not specific to LTSP
10:03
it is a kernel problem
10:04
<gvy>
and quite old one it seems (2005 known)
10:04
<supreme>
warren, i know, but somebody could help me here,... ok, whatever
10:04
<gvy>
http://ubuntuforums.org/showthread.php?t=202076 ,
10:04
?
10:06
<vagrantc>
vlt: sounds like you're basically hit by: http://bugs.debian.org/471793 ... ?
10:10
<vlt>
vagrantc: Hmmm, yes, seems to be exactly that :( Is there any skript that is executed on every (KDE) logout I could use?
10:10deavid has joined #ltsp
10:11
<vagrantc>
vlt: i don't know, sorry.
10:15
<supreme>
vagrantc, have you ever used edubuntu as thin client server right?
10:16
<vagrantc>
supreme: not once, no.
10:16* vagrantc primarily uses and develops debian
10:16
<supreme>
ok
10:17
<vagrantc>
there's a lot of similarities, so i can sometimes be helpful in that regard. but at the end of the day i try to pass the buck to ogra_cmpc if it's too hard for me :)
10:20Q-FUNK has quit IRC
10:21
<warren>
supreme: stop fishing, I told you what the problem likely is, and none of the people are responsible for the broken component
10:23
<supreme>
warren, im asking for edubuntu users cause I have to buy 2 servers and I must to ensure that it will work ok, im not asking for the MAC problem
10:25* warren used /ignore for the first time ever.
10:28
<vagrantc>
supreme: there is an edubuntu channel, you might ask there.
10:29
<supreme>
its curious since i got there: supreme: try #ltsp, they might be more help
10:29
lol
10:29
anyway,, i have to go
10:29
thanks, see you
10:29supreme has quit IRC
10:31staffencasa has joined #ltsp
10:38joebaker has quit IRC
10:42
<gvy>
vagrantc, thanks for r689 :)
10:52ogra has joined #ltsp
10:56daduke has joined #ltsp
10:58sepski has joined #ltsp
11:00praveer_cool has quit IRC
11:06ogra has quit IRC
11:11ogra has joined #ltsp
11:17Blinny has quit IRC
11:22mhterres has quit IRC
11:24mhterres has joined #ltsp
11:26gentgeen__ has quit IRC
11:33gentgeen__ has joined #ltsp
11:34muh2000 has joined #ltsp
11:35vagrantc has quit IRC
11:36Guarald1 has joined #ltsp
11:37Guaraldo has quit IRC
11:46ogra has quit IRC
11:47ogra has joined #ltsp
11:48tux_440volt has joined #ltsp
11:51cpunches has joined #ltsp
11:53Guarald1 has quit IRC
11:56Guaraldo has joined #ltsp
12:11
<johnny>
i broke my chroot .. oops
12:12
warren, sup with that LDM_SESSION idea ?
12:13indradg_ is now known as indradg
12:22ogra has quit IRC
12:23indradg_ has joined #ltsp
12:41indradg has quit IRC
12:47likuidkewl has joined #ltsp
12:47Blinny has joined #ltsp
12:48
<johnny>
dberkholz, i removed almost all the overrides i did in the client chroot's /etc/lts.conf
12:48
i make CONFIGURE_X to false, and added the symlink for the default theme to ltsp (until we have a good gentoo theme)
12:48
so now we're only left with LDM_SESSION
12:50
if we want to remove that, we have to either add another elseif to ldm, or implement warren's idea
12:51
<Blinny>
I don't want to file a bug if this is transient or already known - ltsp-build-client is broken in current Hardy packages - I have output if anyone needs it.
12:52
<laga>
Blinny: how current are your packages?
12:52
Blinny: please post the output on a pastebin
12:52
!paste
12:52
<ltspbot>
laga: Error: "paste" is not a valid command.
12:52
<laga>
!pastebin
12:52
<ltspbot>
laga: Error: "pastebin" is not a valid command.
12:52
<laga>
stupid bot.
12:52
<Blinny>
heh
12:52
<laga>
use www.pastebin.ca or your favourite pastebin :)
12:53
<Blinny>
As current as 'aptitude update && aptitude safe-upgrade' can make you.
12:53spectra has joined #ltsp
12:53
<laga>
Blinny: apt-cache policy ltsp-server?
12:53
<Blinny>
http://pastebin.org/28792
12:54
apt-cache policy ltsp-server: http://pastebin.org/28793
12:56
<laga>
Blinny: ok, you're up to date and that issue looks transient.. i guess you should wait a few hours and then try again
12:56
Blinny: and/or ping ogra
12:56
<Blinny>
Right-O. Cheers.
13:07deavid has quit IRC
13:09K_O-Gnom has joined #ltsp
13:15
<johnny>
ok.. so.. my chroot probably installs from scratch now via ltsp-build-client :)
13:15herson has joined #ltsp
13:15
<johnny>
i need to get my branch merged at some point, or at least put on launchpad somewhere
13:16
uggh.. NO.. i accidentally committed something i shouldn't have..
13:18
<StevenR>
johnny: and that's why one uses version control :)
13:19
<johnny>
obviously..
13:19
i rarely have to revert tho, so i dislike it
13:20
<herson>
jammcq: You are here?
13:23Guaraldo has left #ltsp
13:26tux_440volt has quit IRC
13:27
<johnny>
dberkholz, are you about?
13:30muh2000 has quit IRC
13:33indradg_ has quit IRC
13:33Guaraldo has joined #ltsp
13:33ogra has joined #ltsp
13:36vagrantc has joined #ltsp
13:36likuidkewl has left #ltsp
13:37Guaraldo has quit IRC
13:37indradg has joined #ltsp
13:38Guaraldo has joined #ltsp
13:38sepski has quit IRC
13:39Guaraldo has quit IRC
13:42Guaraldo has joined #ltsp
13:56viking-ice has quit IRC
13:57
<Blinny>
ogra: I assume you already know this, but ltsp-build-client is broken in current i386 Hardy packages -- output is here: http://pastebin.org/28792
13:58
<ogra>
thats debbootstrap, not ltsp :)
13:58
i pinged the udev maintainer
13:58
when did that happen ? i talked to him ths afternoon
14:00
<Blinny>
I just did that this afternoon.
14:00
Thanks.
14:01
<ogra>
should be fixed (soon)
14:01
<Blinny>
Not bothering me at all. Thank you.
14:02
<ogra>
tanks for reporting :)
14:06makghosh|afk has joined #ltsp
14:07makghosh has quit IRC
14:12
<dberkholz>
johnny: yes
14:13
<johnny>
had a question about how we handle XSESSION on gentoo
14:13
since i'm not wanting to modify LDM too much for gentoo than adding a elseif
14:14
<dberkholz>
johnny: ok..
14:15
<johnny>
if i do just LDM_SESSION="/etc/X11/Sessions/Xsession, it's just gonna give me twm
14:15Q-FUNK has joined #ltsp
14:16
<johnny>
atm i have Gnome in there hardcoded, since that's all i have here
14:18
<dberkholz>
johnny: try /usr/lib64/X11/xdm/Xsession
14:19herson has quit IRC
14:19makghosh|afk is now known as makghosh
14:19
<dberkholz>
that one looks like it'll behave more like people expect
14:20
<johnny>
i hope /usr/lib will work..
14:21
:)
14:23deavid has joined #ltsp
14:31viking-ice has joined #ltsp
14:32Blinny has quit IRC
14:39muh2000 has joined #ltsp
14:42dcordes has quit IRC
14:50Gadi has left #ltsp
14:53ogra has quit IRC
14:57makghosh has quit IRC
14:59
<Q-FUNK>
ogra_cmpc: is anyone looking into closig the libsals2 upgrade blocker?
15:03mikkel has quit IRC
15:07deavid has quit IRC
15:10makghosh has joined #ltsp
15:17cliebow has quit IRC
15:29ogra has joined #ltsp
15:30
<warren>
asac: ogra: who is your pulseaudio maintainer?
15:31
<asac>
ogra: luke?
15:31
<ogra>
it ws crimsun for a while
15:32
but he suspended all work, pitti did a lot on pulse this cycle
15:32TelnetManta has quit IRC
15:33
<warren>
so there's no consistent participator upstream?
15:33
<Q-FUNK>
ograseems that hardy will release without -geode after all
15:34
ogra: unless you manage to convince vorlon to make an exception
15:34
<warren>
asac: does the next ubuntu release use pulseaudio by default?
15:34
asac: (alsa emulated device as default too?)
15:36
<asac>
warren: i try to stay as much out of sound as possible :) ... what i can say is that pulseaudio will be on by default
15:36
what is "alsa emulated device" ?
15:37
<ogra>
Q-FUNK, did you fix the deps ?
15:37
<Q-FUNK>
ogra: yup. waiting to be sponsored into debian and sync'ed
15:38
<ogra>
<bryce> slangasek: btw, there is a sync request that's been in for -amd, that didn't make it in before the cut off (we've been testing the fix, but unfortunately whomever does syncs didn't get to it in time).
15:38
<slangasek> bryce: I'll have a look at all those goodies today/tomorrow
15:38
Q-FUNK, dont worry :)
15:38
<Q-FUNK>
:)
15:38
there is a santa in scotland, after all
15:38
<vagrantc>
and he drinks scotch
15:38
<ogra>
seems maddog comlained to sabdfl directy on monday (they met somewhere)
15:39
<Q-FUNK>
however, we'd first need to get it sponsored into debian NOW :)
15:39
<ogra>
pfft
15:39
<Q-FUNK>
yeah, maddog has been in touch with me too, over the last few days
15:39
<ogra>
right
15:40
<Q-FUNK>
I told him that it might be a good idea for Koolu to directly get involved with the launchapad bugs, rather thna blindly file support ticket at Canonical's commercial support team
15:41
I directly contacted him and other Kollu guys to point them to the bug where it's made obvious that the BIOS vendor they use on their hardware screwed up
15:41bricode has joined #ltsp
15:41
<Q-FUNK>
and speaking of the devil ;)
15:42
bricode: nice of you to join us. we were actually discussing the geode driver issues
15:42
<bricode>
me? :)
15:42
Well, I'm around now.
15:43
The award bios is *quite* broken.
15:43
<Q-FUNK>
bricode: apparently, maddog's sermon with sabdfl had an effect. they will sync the -geode driver on time for the release (probably makign an exception to the freeze) and look into possibly missing patches for X core too.
15:44
bricode: it indeed is. the lintop guys noticed it too.
15:44
bricode: Insyde still is the most stable one. general software also got their act together, eventually.
15:44
<bricode>
I'm looking at INT 15 stuff and it's either broken, or non existent.
15:44
<Q-FUNK>
I wouldn't be suprised.
15:45
did the eon always use that brand of bios?
15:45
<bricode>
Q-FUNK: Are the linutop guys using General in their Linutop2?
15:45
Q-FUNK: Seem to have. As long as we've been carrying them...about a year and a bit.
15:45
<Q-FUNK>
they buy the exact same stuff as you guys, in china.
15:45
<bricode>
Q-FUNK: The DTRI units used the Insyde, so I've been testing back and forth.
15:46* bricode wonders if the Linutop guys know the ION has a limited life left.
15:46
<Q-FUNK>
I doubt that they do.
15:47
<bricode>
Q-FUNK: The geode driver seems to be reasonably stable for me apart from the 800x600 default res.
15:48
<Q-FUNK>
this being said, it looks like our dbe62 will end up with a CF socket and might even get an SODIMM socket too. you're all welcome to make further inquiries ;)
15:48
<bricode>
Q-FUNK: sexy.
15:48
Q-FUNK: What about the lifespan of the whole Geode line?
15:48
<Q-FUNK>
AMD gaurantees availability until 2012
15:49
<bricode>
huh. That's a few years out.
15:49
<Q-FUNK>
subject to an extension if the OLPC is a real hit
15:49
yup. 4 years.
15:49
but that's for the traditional GX/LX line
15:50
<bricode>
Q-FUNK: Yeah, I was banking on the fact that OLPC would have been a homerun and that Geode supplies would be safe.
15:50
<Q-FUNK>
they're merging two product lines, starting next year. we might end up being offered something based on recent athlon core instead.
15:50
<bricode>
Q-FUNK: same power envelope?
15:50
<Q-FUNK>
the whole thin client industry is
15:51
no and that's what worries me. nobody except VIA seems to have addressed the concerns of the thin client market adequately. however, VIA keeps on discontinuing their chipsets too often. absolutely unreliable supply line.
15:51
<bricode>
Q-FUNK: I thought that the dbe62 platform finalized?
15:52makghosh has quit IRC
15:52
<bricode>
Q-FUNK: I'm looking forward to the menlow chipsets, although everyone seems to be cramming them into those damn UMPC's. 100K chips manufactured in a day seems to be a reasonable supply...
15:53
<Q-FUNK>
bricode: what we sent you was preproduction and, as I expected, the only good feedback we got was from the LTSP market. for the green workstation market, everyone pointed us to the ION as what they'd want.
15:53
we looked into menlow. it's a no-go. not a real x86 chipset.
15:53
<bricode>
Q-FUNK: Why's that? Too much shaved off for power reasons?
15:54
<Q-FUNK>
no longer compatible at code level
15:54
it really is a brand new architecture, half-way between ARM and x86.
15:54
<bricode>
Q-FUNK: Huh. That's no good.
15:54
<Q-FUNK>
it's not a low-power x86 at all.
15:55
it indeed is closer to x86 than to ARM, but still no longer a real pentium-compatible chipset.
16:01johnny_ has joined #ltsp
16:01
<bricode>
Q-FUNK: Interesting. Hence a bunch of funding to Canonical to develop Ubuntu Mobile...
16:02
<Q-FUNK>
yup
16:02
that's my conclusion too
16:03
or how to save a chipset doomed to fail on the market
16:04pscheie has joined #ltsp
16:04
<bricode>
Q-FUNK: So Geode and VIA. Any other options close by?
16:05
<Q-FUNK>
bricode: anyhow, for the dbe62, we ended up refocusing the design. it was a hard sell, but I finally managed our investors to fund a complete redesign of the motherboard layout so that we can accomodate CF and SODIMM, while keeping essentially the same features
16:06mcfloppy has joined #ltsp
16:06
<mcfloppy>
hello
16:06
<bricode>
Q-FUNK: So with the SODIMM, will there still be onboard RAM?
16:06
<Q-FUNK>
bricode: rumor has it that intel might have something closer down to earth by summer. I'll believe it when I see it.
16:06
bricode: no, we'd be completely dependant on the SODIMM
16:06
<mcfloppy>
how can i test my ltsp from a running windows pc? is there a way to emulate a terminal?
16:07
<bricode>
Q-FUNK: Will you have VGA support in the BIOS?
16:07ogra_ has joined #ltsp
16:08mhterres has quit IRC
16:08
<Q-FUNK>
we'll keep the "everythign soldered-on and welded shut" approach for the next 61 revision, which will be strictly for the thin client market. 62 will be a full motherboard redesing, oriented for the green desktop market - still, at the end of the day, both products with the same components - except for ram and mass storage.
16:08
bricode: 62 will use gsw bios
16:08ogra has quit IRC
16:09
<Q-FUNK>
however, the X driver is already scheduled to be upgraded with libDDC support
16:09
<bricode>
Q-FUNK: CF for mass storage? or IDE?
16:09
Q-FUNK: Yeah, I saw that.
16:10
<Q-FUNK>
CF
16:11
we still won't include a mini IDE disk, because we avoid rotating components
16:11
we're ging the SSD route, instead
16:11
<bricode>
Q-FUNK: That's better. It's where the world's heading.
16:12
Q-FUNK: CF port internal? What are the timelines?
16:14
<Q-FUNK>
yes, internal
16:14
we'd still rather assemble and bolt everything shut ourselves.
16:15ogra_ is now known as ogra
16:15
<johnny_>
mcfloppy, use virtualbox and setup netboot
16:15
hey guys!
16:19
<bricode>
Q-FUNK: Cool.
16:21slipttees has joined #ltsp
16:25topslakr has quit IRC
16:26topslakr has joined #ltsp
16:28Guaraldo has quit IRC
16:32Q-FUNK has quit IRC
16:33topslakr has joined #ltsp
16:34leio has joined #ltsp
16:36Q-FUNK has joined #ltsp
16:38Q-FUNK has joined #ltsp
16:49
<mcfloppy>
johnny thanks
16:56vagrantc has quit IRC
17:01J45p3r has joined #ltsp
17:07Egyptian[Home] has quit IRC
17:08Egyptian[Home] has joined #ltsp
17:11mcfloppy has quit IRC
17:13Q-FUNK has quit IRC
17:16bobby_C has quit IRC
17:20jammcq has quit IRC
17:29vagrantc has joined #ltsp
17:37K_O-Gnom has quit IRC
17:46johnny_ has quit IRC
17:48
<vagrantc>
otavio: any thoughts regarding how we go about limiting the architectures for ltsp, ldm and ltspfs ?
17:48
otavio: i'm thinking the process should go something like ...
17:49
otavio: upload new packages with architecture: i386, amd64, powerpc, arm, armel
17:49
otavio: request that all the old versions be removed (ftp-masters?)
17:57
otavio: this is all regarding: http://lists.alioth.debian.org/pipermail/pkg-ltsp-devel/2008-April/001341.html
17:57slipttees has quit IRC
18:33petre has joined #ltsp
18:33
<petre>
evening all
18:36IRCzito has joined #ltsp
18:38IRCzito has joined #ltsp
18:39
<IRCzito>
Guys, where jetpipe is launched on ltsp-client?
18:40
<vagrantc>
IRCzito: linux distro and release?
18:40
<IRCzito>
I want welp implement client printers on Fedora
18:41
The rigth is launch it by udev?
18:41* vagrantc isn't sure how fedora handles it
18:41
<IRCzito>
or use some scripts on rc... or sysinitv
18:42
<vagrantc>
definitely wanted to move in the direction of launching it from udev in general.
18:42
but haven't really gotten there yet, at least on debian.
18:44
<IRCzito>
vagrantc: why not use cups on the client?
18:46
vagrantc: Today jetpipe work on ubuntu?
18:47bricode has quit IRC
18:48Q-FUNK has joined #ltsp
18:48jammcq has joined #ltsp
18:49
<vagrantc>
IRCzito: i don't know where it works.
18:49
IRCzito: cups is a little heavy, but you can configure it manually, if you want.
18:52
so, to properly get ltspfs and cdpinger started from udev, i feel like we need to move the binaries into a new location, and write a wrapper script that doesn't restart itself.
18:53
er, doesn't start again if another is already running ... works for cdpinger ... not sure about proper behavior for ltspfsd
18:53
since it seems like ltspfsd starts one instance of itself for each mounted device...
18:55staffencasa has quit IRC
19:07viking-ice has quit IRC
19:10viking-ice has joined #ltsp
19:16IRCzito has quit IRC
19:18vagrantc has quit IRC
19:32viking-ice has quit IRC
19:39TelnetManta has joined #ltsp
19:46ogra has quit IRC
19:47
<warren>
asac: I'm going to be approaching Adobe at some point in the future about cooperating during testing of their next flash version in order to get rid of all the crashes that plague us
19:47
asac: could you join us from Debian/Ubuntu in some official capacity so we have more weight in our negotiations with them?
19:47
asac: (we're kind of in the same boat now)
19:52J45p3r has left #ltsp
20:03ace_suares has joined #ltsp
20:04
<ace_suares>
hi guys
20:05spectra has quit IRC
20:19ogra has joined #ltsp
20:33vagrantc has joined #ltsp
20:37joebaker has joined #ltsp
20:52
<lns>
hey ace_suares
20:52herson has joined #ltsp
20:52
<ace_suares>
hi Ins
20:53
<lns>
haha..everyone calls me Ins
20:56lns is now known as askdjfksjdkfsjkd
20:56askdjfksjdkfsjkd is now known as lns
20:57lns has joined #ltsp
21:18herso1 has joined #ltsp
21:20herso1 has left #ltsp
21:21herson has quit IRC
21:24herson has joined #ltsp
21:30
<warren>
cyberorg: who is suse's firefox maintainer?
21:39Faithful has joined #ltsp
21:43BGomes has joined #ltsp
21:46mccann has quit IRC
21:53Faithful has quit IRC
21:54petre has quit IRC
21:56herson has quit IRC
22:13Faithful has joined #ltsp
22:35
<johnny>
hey warren :)
22:35
<warren>
hi
22:35
heading to sleep soon
22:41yanu_ has joined #ltsp
22:45Faithful has quit IRC
22:46
<johnny>
warren, thanks for making my life easier
22:46
!
22:47
<warren>
johnny: huh?
22:48
in what way?
22:48
<johnny>
the stuff you did to remove the debianisms :)
22:48
<warren>
johnny: yeah, but now they're busy removing fedoraisms
22:49
and i'm not even done adding more
22:49
=)
22:49
<johnny>
i'll have to stop you on that one :)
22:49
so.. i finally have a working install now
22:50
<warren>
what distro are you?
22:50* warren can't keep everyone straight
22:50
<johnny>
gentoo
22:51
my real deployment is on ubuntu tho
22:51
since ubuntu was already in place at the store mostly
22:52
<warren>
I'm surprised anything I did was useful for non-Fedora
22:52
I don't remember what I did
22:53
<johnny>
removed usages of update-alternatives
22:53
and dpkg --architecture, etc
22:53
<warren>
I don't remember doing that
22:53
<johnny>
that's what your blog says
22:53
<warren>
and I realized today that something else in F9 changed breaking LTSP entirely
22:53
havne't been able to figure it out
22:53
=)
22:54
<johnny>
ouch
22:54vagrantc has quit IRC
22:56
<warren>
oh. I think I know what it is.
22:59yanu has quit IRC
23:00
<warren>
sigh, this appears to be a last minute kernel regression
23:01cpunches has left #ltsp
23:01Faithful has joined #ltsp
23:03xachen has quit IRC
23:05
<johnny>
:(
23:12BGomes has left #ltsp
23:18xachen has joined #ltsp
23:37
<warren>
hmm, doesn't appear to be the client's kernel
23:37
just tried a dozen kernels going backwards
23:54asac_ has joined #ltsp