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


Channel log from 19 April 2008   (all times are UTC)

00:02opapo_ has joined #ltsp
00:02ace_suares has joined #ltsp
00:11opapo_ has quit IRC
00:11ace_suares has quit IRC
00:13opapo_ has joined #ltsp
00:13opapo has joined #ltsp
00:14opapo_ has quit IRC
00:16
<vagrantc>
ogra: not having as good luck with the localdev/ldm_directx patches
00:17
ogra: caused a regression leaving the /tmp/ mount
00:17
ogra: and unmounting didn't seem to work
00:18
ogra: when not using ldm_directx
00:23
<johnny>
dberkholz, i don't know what i did wrong (or if i did) but genkernel generated initramfs are giving me trouble now
00:23opapo has quit IRC
00:23
<dberkholz>
hm
00:24
what kind of trouble?
00:24
<johnny>
lemme try with the ~arch version
00:24
<dberkholz>
that would be a good idea
00:24
<johnny>
i think i was forcing the arch version before.. but then i took it out, cuz i thought it was unnecessary
00:24
<dberkholz>
that's what releng people are using to build the release
00:24
by arch, do you mean ~arch?
00:25
<johnny>
yes.. oops
00:25
<dberkholz>
those are quite different terms, the former would mean stable if it was commonly used
00:25
<johnny>
i know this.. that was accidental :)
00:25
<dberkholz>
ok
00:26
i just spent about 20 minutes trying to explain this to some other people
00:26
<johnny>
i've been here long enough to know that
00:26
you can't survive very long if you don't.
00:27
it still doesn't quite do enough for us, even if we just want to use tmpfs bindmounting and nfs
00:27vagrantc has quit IRC
00:27
<johnny>
the modules_load file is woefully inadequate.. lots of scsi drivers.. only 2 network drivers..
00:28joebaker has joined #ltsp
00:28opapo has joined #ltsp
00:30opapo has quit IRC
00:31opapo has joined #ltsp
00:31
<dberkholz>
johnny: does the --all-initrd-modules argument help at all, or does it just copy without loading?
00:31
<johnny>
it would be the next problem after this to solve..
00:32
<dberkholz>
johnny: other ideas are to try including a custom modules_load with --initramfs-overlay, or to create an additional architecture type for ltsp archs
00:32
<johnny>
yes.. that is one option
00:32asac_ has joined #ltsp
00:33
<johnny>
just have to make it..
00:33
<dberkholz>
i suppose what we'd like is a way to overlay our own files on top of their arch, rather than really create a new arch
00:34
wonder out of busy-config config.sh kernel-config-2.6 modules_load nb-busybox.cf, how many would change
00:39
<johnny>
dberkholz, i think i just had to step away from doing ltsp stuff for a lil while
00:39
the problems seem much simpler now
00:40
<dberkholz>
good to hear
00:42
<johnny>
hopefully after 2008.0 comes out, we can get busybox updatedd in genkernel.. that'd solve a bit of the configuration issues
00:43
<dberkholz>
johnny: feel free to make local mods to genkernel, stick 'em somewhere, and make an ebuild point at it
00:43
<johnny>
then we can stop putting the ip the the kernel commandline in the pxelinux file
00:44
atm you have to do nfsroot=yourappserver:/path/to/chroot
00:44asac has quit IRC
00:44asac_ is now known as asac
00:45
<dberkholz>
append initrd=pxe/x86/initramfs ramdisk=8192 ip=dhcp root=nfs:192.168.1.106:/opt/ltsp/i386;tmpfs
00:45
i suspect that's my old genkernel-3 line
00:45
or maybe it's genkernel-4 before i fixed it to not suck
00:46
the semicolon makes me think gk4
00:46
<johnny>
yes, that is gk4
00:46
i'm saying that we dont need to use the ip there
00:46
<dberkholz>
oh absolutely, if udhcpc works
00:46
<johnny>
no.. we still do :)
00:47
until we use busybox 1.9
00:47
<dberkholz>
this was my new and improved line (still before root-path worked): append initrd=pxe/x86/initramfs ramdisk=8192 ip=dhcp root=nfs;tmpfs
00:47
<johnny>
see.. that's what we want :)
00:47
genkernel4 had busybox 1.9 :)
00:49
<dberkholz>
i think i was still updating a ton of crap in my live gk4 ebuild
00:51
<johnny>
hmm.. syntax error in /etc/lts.conf line2
01:39praveer_cool has joined #ltsp
01:53Egyptian[Home] has joined #ltsp
01:58Egyptian[Home2 has joined #ltsp
02:00mccann has quit IRC
02:06Egyptian[Home1 has quit IRC
02:16Egyptian[Home] has quit IRC
02:40Egyptian[Home2 is now known as Egyptian[Home]
02:41
<johnny>
dberkholz, this is what you told me to try right?
02:41
/usr/lib/x11/xdm/Xsession
02:41
ooh..
02:42
<dberkholz>
with a capital X in X11
02:43
<johnny>
oh..
02:43
that does mean we have to merge xdm tho
02:46
ok.. here goes
02:46
ok.. now i just get a twm session
02:46
so.. need something else
02:46
<dberkholz>
for whatever reason, upstream has xdm supply the Xsession script instead of xinit
02:47
<johnny>
any ideas? since that doesn't work?
02:47
<dberkholz>
johnny: it really should work.
02:48
assuming you're passing an argument as a session that actually exists in /etc/X11/Sessions/ or as an executable file
02:48
Xsession sets the variable XSESSION, which is used by /etc/X11/chooser.sh
02:49
<johnny>
got it..
02:49
i had just migrated to baselayout
02:49
err openrc*
02:50
forgot to set it on my server
02:50
<dberkholz>
forgot to set what?
02:50
<johnny>
XSESSION="Gnome"
02:50
<dberkholz>
this should require that variable to be preset to work
02:50
<johnny>
in /etc/env.d/99xsession on my server
02:50
<dberkholz>
it should just require that the session name get passed to the Xsession script
02:51
i'm sorry, "this _shouldn't_ require that variable to be preset to work"
02:51
<johnny>
it's gotta default to something
02:52
gnome-session is in the list tho
02:52
on ldm
02:53
but this is good enough for now.. i need to sleep .. almost 4am
02:53
<dberkholz>
bah. defaults.
02:53
<johnny>
i set this on my server, gdm just happened to choose Gnome, ldm doesn't pass the type of session by default
02:54
my server being my normal desktop, i just didn't fix up the env.d file after upgrading to openrc there
02:55
so tomorrow i'll have one of these fine gentlemen commit that, and we can distribute and empty lts.conf file
02:55
an*
02:55
so.. just need to work out the initramfs issue, and what happens with the pxe directory, then we might have something to open to wider testing
02:56
<dberkholz>
good deal
02:56
<johnny>
i'd apprecaite if we could get some tarballs hosted somewhere official
02:56
or we could just use fedora's
02:56
<dberkholz>
tarballs for what, again?
02:56
ltsp packages?
02:56
<johnny>
yes
02:56
<dberkholz>
yeah, we'll get that going.
02:57
we've talked about it a few times now
02:57
<johnny>
relying on ubuntu's like before is impossible
02:57
<dberkholz>
we all agree that releases are a good thing
02:57ogra has quit IRC
02:58
<dberkholz>
some of us think they should be called something besides releases (Despite being releases) for various reasons, but that's not something that matters at all
02:58
<johnny>
they just have to be tarballs that match the release.conf
02:58ogra has joined #ltsp
02:58indradg has quit IRC
03:00
<johnny>
i also need to get my changes added to ltsp-trunk, i just wanna do a few more smoketests first
03:00
i'll do another when i go to work at our store..
03:01
make sure the chroot succeeds from start to finish with trivial or no modifications
03:02
well, it's bed time for now, cya
03:02
<dberkholz>
night
03:28ikonia has quit IRC
03:30ikonia has joined #ltsp
03:35praveer_cool has quit IRC
03:36sep has quit IRC
03:41indradg has joined #ltsp
03:53viking-ice has quit IRC
03:53viking-ice has joined #ltsp
04:29Pascal_1 has joined #ltsp
05:13DawnLight has joined #ltsp
05:19
<DawnLight>
how can i specify kernel parameters for clients?
05:32K_O-Gnom has joined #ltsp
05:43Q-FUNK has joined #ltsp
05:48deavid has joined #ltsp
06:03spectra has joined #ltsp
06:13Q-FUNK has quit IRC
06:57J45p3r has joined #ltsp
07:12barbosa has joined #ltsp
08:03Pascal_1 has quit IRC
08:08indradg has quit IRC
08:45indradg has joined #ltsp
08:54K_O-Gnom has quit IRC
08:54tux_440volt has joined #ltsp
09:10tux_440volt is now known as Subhodip
09:12Nubae has joined #ltsp
09:13
<Nubae>
ogra: if you are there, what was the url for the RC page
09:14
ah found it
09:15
ok... first... when I hit f4 I get oem install as the last option not LTSP install
09:15
in fact I dont get LTSP install at all
09:19
I've double checked and it is the CD built on 17/4/08 so its the right onw... the ltsp option is not in there
09:24Q-FUNK has joined #ltsp
09:25
<laga>
Nubae: you using the normal ubuntu install?
09:26
Nubae: or the alternate disk?
09:35barbosa has quit IRC
09:37
<Nubae>
ogra: are u there?
09:42
<ogra>
Nubae, well, its definately on the alternate disk, i promise :)
09:42
<Nubae>
shit... did I download the desktop cd
09:43
or normal server cd
09:43
<ogra>
do you have two or four entries in the modes menu ?
09:44
desktop only has two afaik
09:44
<Nubae>
maybe there should be a direct link to the alternate cd here - https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
09:44
4
09:44
so must have been server, and I thought that was the right onw
09:44
<ogra>
nope
09:44
thats why it says "alternate CD" everywhere :)
09:44
<Nubae>
I no it says it clearly in the instructions
09:45
but I was in a rush and downloaded normal server cd without looking
09:45
anyway, without redownloading, should I just install normal ubuntu
09:46
<ogra>
setting a fixed link means that it will need to be updated for 8.04.1 .2 .3 etc
09:46
<Nubae>
that would have suited me fine ;-)
09:46
<ogra>
i'll make "alternate CD" bold in the page
09:46
<Nubae>
but I know what you mean yea
09:47
so, shoudl I redownload?
09:47
i was so concentrated on the 64 bit part, the alternate bit escaped me
09:48
<ogra>
ah, well, for 64bit you should raher use the manual method anyway
09:48
the package selection on the 64bit cd still doesnt contain ay 32bit packages .... so it will build an amd64 chroot out of the box
09:49
<Nubae>
yeah its been like that from day 1
09:49
<ogra>
intrepid will get an option if network is detected to buiod a 32bit client from the network archive during install
09:50
<Nubae>
downloading a lot faster today than yesterday in any case
09:50
<ogra>
indeed
09:50
<Nubae>
so install from the regular ubuntu server cd should not cause issues
09:51
<ogra>
RC and release day are always hard for the servers
09:51
<Nubae>
wont be able to beta test u're instructions though
09:51
<ogra>
err
09:51
server doesnt gain you anything
09:51
you need a desktop
09:51
<Nubae>
ah sorry, I have desktop
09:51
<ogra>
-alternate or -desktop
09:52
<Nubae>
ubuntu-8.04-RC-desktop
09:52DonSilver has joined #ltsp
09:52
<ogra>
thats fine
09:52
<Nubae>
amd64
09:52
ok
09:53
gonna try writing again with brasero, the regular cd hangs
09:53
at startup
09:54Subhodip has quit IRC
09:55
<Nubae>
hope u didn't get too much mail yesterday/today :-)
09:56vagrantc has joined #ltsp
09:57
<Nubae>
yesterday I asked, but people were quite busy... are there any recommendations for actual thin terminal clients ( a website url preferably)
10:03
<Q-FUNK>
ogra: is there any way to prevent invoke-rc.d from running inside the chroot, when upgrading packages?
10:04
<Nubae>
q funk, enable proc and sys
10:04
then it wont matter
10:04
damn... desktop cd hangs
10:07
cd hangs on all types of startups, I'll wait for that actual alternate cd to download
10:07
maybe the iso was downloaded badly
10:08
<laga>
verify the md5sum
10:14tux_440volt has joined #ltsp
10:17
<Nubae>
where are the 8.04 hashes?
10:18
<laga>
on the mirrors
10:20
<Nubae>
ok mdsum is correct
10:21
turn off apic and lapic?
10:23
even clicking on check cd for defects freezes the system
10:27
well 40 minutes to wait to download alernate, but will the same thing happen
10:29
<vagrantc>
Q-FUNK: LTSP_HANDLE_DAEMONS=false chroot /path/to/your/ltsp/chroot
10:29
Q-FUNK: or just export LTSP_HANDLE_DAEMONS=false
10:30
Q-FUNK: that'll trigger the policy wrapper so invoke-rc.d won't do anything
10:30
<Q-FUNK>
ah ok
10:32DawnLight has left #ltsp
10:35
<Nubae>
weird, written 2 cds, both cause the computer to freeze then eject the disc saying error reading disk
10:36
<StevenR_>
tested the drive with a known-good cd?
10:45
<Nubae>
yes
10:45
20 minutes till alternate downloads, just gonna wait
10:45tux_440volt has quit IRC
10:46
<Nubae>
weird, one cd hangs on a screen with a white blinking -
10:46
the other extracted the cd after some time saying there was an error with it
10:46
burnt one with brasero, one with what comes standard
10:53
<vagrantc>
ogra: did you say you could confirm the regression regarding the ltspfs /tmp/.USER-ltspfs/DEVICE issue?
10:56
<Nubae>
5 dias eh... cool
10:56
I wonder, has ubuntustudio changed much at all
11:02
weird, been testing mandriva, and it comes with the mic settings to full as standard, so you can hear yourself talk...
11:02
lol
11:07tux_440volt has joined #ltsp
11:13
<Nubae>
ok, I have the alternate Cd downloaded now, can I follow the same instructions even if its 64bit?
11:13Q-FUNK has quit IRC
11:13
<laga>
Nubae: if your clients are 64bit
11:13
<Nubae>
they're note
11:13
not
11:14
<laga>
16:45 < ogra> ah, well, for 64bit you should raher use the manual method anyway
11:14
16:46 < ogra> the package selection on the 64bit cd still doesnt contain ay 32bit packages .... so it will build an amd64 chroot out of the box
11:14
<Nubae>
who would have 64 bit thin clients
11:15
right, but I can build another chroot after without problems right?
11:15
using the --arch flag
11:15lns has quit IRC
11:15
<laga>
afterwards, yes
11:15
<Nubae>
k, but forced to install 64 bit chroot, get it
11:17Q-FUNK has joined #ltsp
11:17Pascal_1 has joined #ltsp
11:19
<Nubae>
ASUS, a leading company in the new digital era, today announced the release of a comprehensive software developer's kit (SDK) for the ASUS Eee PC. The new SDK enables third-party developers to easily enhance and extend the popular, mobile, open source platform for educational, telecommunications, and many other uses. The Eee PC SDK provides a complete development platform, including an Eee PC-compatible Open Circulation Edition of the X
11:19
a Qt4 toolkit, a developer's guide, sample applications, and a multilingual VMware testing and debugging environment." - they really are beating classmate and xo to the punch
11:20davidj has quit IRC
11:30
<Nubae>
http://www.everex.com/products/cloudbook/cloudbook.htm - that looks interestig
11:30tux_440volt has quit IRC
11:32lns has joined #ltsp
11:33
<Nubae>
jeez, the alternate cd hangs too..
11:36Nubae has left #ltsp
11:37Nubae has joined #ltsp
11:38
<vagrantc>
warren: what's the deal with: [ -f /selinux/enforce ] && /sbin/restorecon -R $TFTPDIR &> /dev/null
11:38
warren: can we just use > /dev/null ?
11:39
<warren>
vagrantc: let me test that
11:39
vagrantc: yeah, i guess so, go ahead and change it.
11:39
<vagrantc>
warren: cool.
11:46
<johnny>
ok.. guys
11:46
here's the magic line
11:46
/usr/lib/X11/xdm/Xsession
11:46
commit that to ldm for me?
11:47
<vagrantc>
johnny: does it accept arguments?
11:47
<johnny>
sorry for taking so long.. i had some serious issues with my routing
11:47
you mean the ones in the dropdown ?
11:47
gnome-session and ll?
11:47
<vagrantc>
johnny: i.e. /usr/lib/X11/xdm/Xsession FOO ... starts the FOO window manager
11:48
<johnny>
what sets the default ?
11:48
<vagrantc>
i don't know how gentoo works.
11:48
<johnny>
on a normal desktop
11:48
<vagrantc>
define normal
11:48
<Nubae>
argh... is there some easy way for me to make an iso onto a usb stick instead?
11:48
<johnny>
the default is an env var called XSESSION
11:48
<vagrantc>
johnny: i can explain the behavior on debian and ubuntu systems, but every distro appears to implement this stuff differently.
11:48
<johnny>
vagrantc, the one you're on right now :)
11:49
prolly
11:49
<vagrantc>
hopefully we can fix all the distros. :)
11:49
exactly what goes on behind the scenes i don't care about.
11:49
<Nubae>
all the distros... lol
11:50
<vagrantc>
but i would like to see all supported Xsession's handling two behaviors: no argument means set some reasonable default for your distro, and the first argument is which session to run.
11:50
that's how debian works
11:50
and it seems reasonable.
11:51
at the moment, though, LDM actually doesn't use Xsession when you've manually selected a session
11:51
<johnny>
ok.. so i'll test to make sure the dropdown works
11:52
<vagrantc>
if you use the dropdown, it will skip Xsession entirely.
11:52
<johnny>
so how can i test whether it takes an argument that actually works? :)
11:52
<vagrantc>
read the code?
11:53
i would think it should be near the top of the code...
11:53
johnny: i guess it's not really important at the moment, since not all distros support it. but i would like to be able to support it in the future.
11:54
<johnny>
ok.. i selected failsafe xterm, while the default is Gnome
11:54
and it worked
11:54
<vagrantc>
maybe we can make it configurable ... but then it will break cross-distro logins.
11:54
<johnny>
so yes.. it worked
11:55
<vagrantc>
it worked, as in your system allows you to run arbitrary commands... that does nothing to demonstrate weather your Xsession supports passing an argument.
11:58
johnny: like i said, don't worry about it now. we'll fix it when we get to it.
11:58
<johnny>
i was told that it did
11:58
<vagrantc>
it did what?
11:58
<johnny>
supports passing an argument
11:59
<vagrantc>
ok, great :)
12:14Q-FUNK has quit IRC
12:15
<Nubae>
damn cdrom drive, changed it now it works great
12:15
I thought HP was supposed to be good stuff
12:15Pascal_1 has quit IRC
12:17
<Nubae>
hp proliant dl... both the network ports have broken and now the cdrom drive
12:17
I hightly dis-recommend that piece of machinery
12:24k has joined #ltsp
12:24 Morning everyone
12:26 I just bought an hp 2207h widescreen lcd and am trying to get a client to display proberly in k12ltsp5
12:26
<johnny>
think you can commit that for me guys?
12:27 The desktop fills the display but images are stretched a bit too wide... ;-(
12:28slipttees has joined #ltsp
12:28Q-FUNK has joined #ltsp
12:29MacIver has quit IRC
12:31jammcq has joined #ltsp
12:31MacIver has joined #ltsp
12:31
<johnny>
vagrantc, lemme know when you do
12:32
i'm pretty close to an OOTB install
12:32
just need to setup the tftp server, and copy the initramfs and kernel to the chroot now
12:32
err out of the chroot now
12:32 the lcd says it is at 1680x1050 but images are too wide
12:32
<johnny>
that's temporary for the copying tho
12:39
<Q-FUNK>
http://launchpadlibrarian.net/13579502/buildlog_ubuntu-hardy-i386.xorg-server_2%3A1.4.1~git20080131-1ubuntu10_CHROOTWAIT.txt.gz
12:39
*sigh*
12:46k has quit IRC
12:51
<vagrantc>
johnny: do what?
12:53
<johnny>
commit that line
12:53
for the xsession
12:53
<vagrantc>
ah.
12:57
<Nubae>
is it normal that building thin client sits at 50% for a very long time?
12:57
<laga>
Nubae: yes
12:57
Nubae: alt+f4 to see progress
12:58
<Nubae>
just started compressing image so ok now
12:59
<laga>
it's amazingly fast to build the image from disk
13:00
<Nubae>
well now comes all the hard part, migrating from gutsy
13:03
<vagrantc>
johnny: pushing now...
13:04
<johnny>
i'm going to do a smoketest tonight when i head to work
13:04
then maybe i might have something to recommend for merging
13:06shogunx has quit IRC
13:11vagrantc has quit IRC
13:11
<johnny>
testing now
13:12
so, is anybody hosting tarballs that match the release.conf versions?
13:12shogunx has joined #ltsp
13:13vagrantc has joined #ltsp
13:13
<johnny>
vagrantc, line 464 is that correct?
13:14
i think you need to remove it
13:14
<vagrantc>
johnny: of?
13:14
<johnny>
ldm.c
13:14
here's the error
13:14
ldm.c:461: error: expected expression before ‘else’
13:14
<vagrantc>
well, i think i need to add another
13:15
<johnny>
oh yeah
13:15
need to add another } :)
13:15
<vagrantc>
can you add another at after the debian segment ... test it, and i'll commit
13:16
johnny: by test, i mean just see if you get the compile error.
13:18
<johnny>
yes
13:18
it built
13:19
<vagrantc>
pushing...
13:23
<johnny>
build complete, now testing :)
13:23
if this works, i dont' have to distribute an /etc/lts.conf with any overrides that might get accidentally taken out by the user
13:24
yay
13:40
is it me or is getltscfg too strict?
13:40
syntax error in /etc/lts.conf, line 3
13:40
there's nothing on that line :)
13:40
<laga>
line 2 might be faulty?
13:40
<johnny>
line two is [default]
13:41
<laga>
maybe it's expecting something? do the following lines have content?
13:41
<johnny>
there are no following lines?
13:41
<laga>
i wonder if just having
13:41
[default]
13:41
<johnny>
ubuntu distributes that way iirc
13:41
<laga>
with not content for that section is valid
13:41
oh?
13:41
<johnny>
pretty sure it has a bunch of commented lines
13:41
and then [default]
13:42
oh.. it's just comments
13:43
i coulda swore i tried that, which is why i added [default] in the first place :(
13:43
ok.. here goes..
13:43
i think it should allow [default] tho
13:43
that way users dont' have to remember to add it
13:44
and wonder why they get syntax errors
13:44
<Nubae>
I've got my new install up now, but it jumps to busybox with mount: Mounting /rofs failed:infalid argument
13:44
I've been through this issue before
13:45
<laga>
is inetd.conf ok?
13:45
<vagrantc>
johnny: if you have an lts.conf, you need something in there.
13:46
johnny: or no default entry
13:46
<Nubae>
looks correct, this is off a totally fresh install
13:46
with no messing around at all
13:46
it was i386 I downloaded in the end, so built the ltsp chroot and everything
13:46
but then getting this error
13:47
<laga>
Nubae: dhpcd.conf correct?
13:48
<Nubae>
considering its pointing to the same place as always, I guess so
13:49
<laga>
Nubae: /opt/ltsp/images/i386.img existing?
13:49
<Nubae>
anywayes
13:50
sorrry was supposed to be just yes
13:50
<johnny>
yay.. i rock :)
13:54
<Nubae>
I've tried rebuilding the image, maybe that will work?
13:55
nope didn't help
13:55
this is totally strange, a fresh i386 install, and I get that
13:56
Its looking for rootserver 192.168.1.1, what is that?
13:56
<laga>
Nubae: what's in the pxelinux config?
13:57
<Nubae>
just standard DEFAULT vmlinuz to initrd=initrd.img
13:57
<laga>
odd
13:57
<Nubae>
i took out quiet and splash to see what was happening
14:00
any more ideas?
14:03Nubae has left #ltsp
14:12Furlow has joined #ltsp
14:14
<Furlow>
hi would anyone know why my client says its trying to boot from /etc/pxelinux.cfg.<some code here> and then each time it takes when leter off the end and entually boots from default
14:15
also my client is booting but the keyboard is all messed up although I have the right key config
14:15
<Q-FUNK>
ogra: does LTSP have a list of known X drivers it uses to parse whta X configure returns? could it be that -geode isn't there but -amd is?
14:31Furlow has left #ltsp
14:32vagrantc has quit IRC
15:09DonSilver has quit IRC
15:43jammcq has quit IRC
15:51leio has quit IRC
15:55leio has joined #ltsp
16:03
<johnny>
oops.. i'm a bad person :(
16:03
i forgot to commit my updated quickstart file since ages
16:04
smoke testing will begin in t minus 1 hour :)
16:51Pascal_1 has joined #ltsp
16:56Pascal_1 has quit IRC
16:59deavid has quit IRC
17:06mccann has joined #ltsp
17:13viking-ice has quit IRC
17:45Q-FUNK has quit IRC
18:07vagrantc has joined #ltsp
18:27
<vagrantc>
hrm.
18:36
yay!
18:37
figured out the double-mount problem.
18:37
ldm's rc.d handling just looks at the beginning of the filename, and doesn't exclude things like .old and such.
18:37
so i had multiple scripts in the directory.
18:37
and so it was getting mounted multiple times ...
18:44
<warren>
a bunch of .d scripts either have to enforce
18:44
1) never write .old or .rpmnew or .rpmwhatever files on version upgrades
18:44
(always overwrite)
18:44
2) or have an exclusion list of suffixes
18:44
3) or only operate on files that end in a certain fixed suffix
18:45
we had to patch upstart to ignore certain filenames in /etc/event.d
18:45
for example
18:52slipttees has quit IRC
19:05
<vagrantc>
debian uses run-parts to handle most of that stuff
19:06
we could probably move run_parts_list into ltsp-common-functions and use that.
19:11Q-FUNK has joined #ltsp
19:21
<Q-FUNK>
hm
19:24izigo has quit IRC
19:47Q-FUNK has quit IRC
19:50
<vagrantc>
warren: i mentioned all this at one point.
19:52Q-FUNK has joined #ltsp
19:55
<Q-FUNK>
am I the only one whose chroot no longer works on Hardy since a couple of days?
20:01vagrantc_ has joined #ltsp
20:14vagrantc has quit IRC
21:08Q-FUNK has quit IRC
21:09
<warren>
vagrantc_: I asked if we could switch to debian's run-parts, I was told no.
21:19
<vagrantc_>
warren: yeah, that's why we have the run_parts_list function
21:20
warren: it basically can do the same thing.
21:30
warren, ogra, dberkholz, johnny, cyberorg, gvy: preparing to tag ltspfs soonish.
21:31
<warren>
vagrantc_: I'll take a quick look to be sure there's nothing I had queued
21:32
vagrantc_: that's without the udev based thing right?
21:35
vagrantc_: ok, looks OK
21:35
vagrantc_: I'm assuming ogra's changes are working fine for you?
21:40
<vagrantc_>
warren: yes, they're working fine. i had some troubles at first, but it was because of mistakes i had made.
21:41
but yes, i had brought up a while ago using something to ensure backup files or whatever, aren't executed.
21:44captain_magnus has quit IRC
21:48spectra has quit IRC
21:49captain_magnus has joined #ltsp
22:09J45p3r has quit IRC
22:20captain_magnus has quit IRC
22:22captain_magnus has joined #ltsp
23:26
<vagrantc_>
tagged and pushed ltspfs-trunk
23:27vagrantc_ has quit IRC