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


Channel log from 22 January 2009   (all times are UTC)

00:01cyberorg has quit IRC
00:03
<warren>
stgraber: hmm, I didn't have the file command installed
00:10cyberorg has joined #ltsp
00:10Egyptian[Home]1 has quit IRC
00:12Egyptian[Home] has joined #ltsp
00:36vlt has joined #ltsp
00:37alkisg1 has joined #ltsp
00:37alkisg has quit IRC
00:38alkisg1 is now known as alkisg
00:47vlt has quit IRC
00:47vlt has joined #ltsp
00:48
<Ryan52>
alkisg, http://ryan52.info/~ryan52/ldm_2.0.30-0ubuntu1~ppa1.1_i386.deb
00:49
<alkisg>
Ryan52: wow, a whole .deb?!!! :)
00:49
Should I install all of it or just extract ldm?
00:49
<Ryan52>
install all of it.
00:50
it's directly based on stgrabers version plus my patch, so no need to go through the trouble to just get ldm.
00:50
<alkisg>
Ryan52: thanks a lot!!! I'll try it in a few hours and tell you the returned error code
00:51
<Ryan52>
alkisg, it'll say (in the log) "waiting for: PID" before the error, which will now say "waitid returned an error: BLAH" (we're using waitid, not sigwaitinfo...:P). go look through the log earlier to find out what PID really is (ssh, greeter, etc). what command PID really is (supposed to be, it might be dead) and the error message are the important part.
00:52
<alkisg>
OK. It's during logon (but it doesn't actually logout) so I guess the command will still be alive
00:52
s/logon/logout
00:53
<Ryan52>
ok.
00:53* Ryan52 will be sleeping in a few hours, but will try to help in the morning if somebody else hasn't
00:55
<alkisg>
Ryan52: ah... ignore me if I ask too much, but is it easy to also give me a version that has a "break" after "if (res == -1) { " ? So that I can actually verify that everything works OK with it?
00:55
(no need for a whole .deb, plain ldm will do)
00:56
<Ryan52>
like this? http://slexy.org/raw/s271cNktyK
00:56
<alkisg>
Yup, that look fine! :)
00:56* Ryan52 doesn't understand.
00:56
<Ryan52>
ok.
00:57
<alkisg>
Ah, maybe I didn't look at the code well enough?
00:57* alkisg is looking at ldm.c again....
00:57
<alkisg>
(The point is to make ldm exit in case of an error)
00:57
<Ryan52>
ya.
00:57
tho that'll just make ldm_wait exit.
00:58
the rest of ldm will keep doin whatever it does.
00:58
<alkisg>
hm... let me see the source...
01:01
<Ryan52>
warren, dumb question...how does file give a uniq string? if I do "dd if=/dev/urandom of=/dev/stdout bs=64 count=1 2>/dev/null | file -" I get "/dev/stdin: data" every time..is it something specific to nbd?
01:01* Ryan52 will bbiab (10 or 20 minutes)
01:04
<alkisg>
Ryan52: I think if ldm_wait() returns, the rest of ldm will close properly... So if/when you have the time, I'd like to also try the "break" patch. :)
01:08
<warren>
Ryan52: try it on a nbd device pointing at a squashfs image
01:08
Ryan52: [root@newcaprica i386]# dd if=/dev/nbd9 of=/dev/stdout bs=64 count=1 2> /dev/zero | file -
01:08
/dev/stdin: Squashfs filesystem, little endian, version 3.1, 154 bytes, 19803 inodes, blocksize: 131072 bytes, created: Thu Jan 22 01:18:39 2009
01:11
<Ryan52>
warren, okie doke
01:12* Ryan52 has no clue how to get an nbd device pointing to a squashfs image, nor does he want to :)
01:12
<warren>
that isn't a supported network rootfs method on debian?
01:12
<Ryan52>
alkisg, depends on which ldm_wait it is. but ya.
01:12
<warren>
it it way faster than nfs
01:13
<Ryan52>
*shrug*
01:15
alkisg, http://ryan52.info/~ryan52/ldm
01:16
<alkisg>
Ryan52: mucos gracias!
01:17
<Ryan52>
de nada
01:18
<warren>
I'm curious how debian/ubuntu handles a particular problem with NBD root...
01:18
our system fails to shutdown because it kills all processes during shutdown before it unmounts filesystems
01:18
nbd-client is killed...
01:18
so the system gets stuck before it unmounts
01:21* Ryan52 goes off to sleep
01:22
<Ryan52>
alkisg, highlight me with the results, or I'll forget to check the backlog :) (I'm not a morning person)
01:22
<alkisg>
Ryan52: np, I'll also talk to you when I see you online. Ty!!!
01:44tjikkun_work has joined #ltsp
01:44
<warren>
fun... X fails to start with less than 256MB RAM now...
01:50
<cyberorg>
warren, same problem here, can't shutdown/reboot after changes were made to do normal shutdown instead of force shutdown
01:50
i thought it was something in my experimental image
02:04warren has quit IRC
02:22
<Appiah>
Does no one here have experince of homedirs on the network?
02:22
I'm gonna start crying soon
02:23
<johnny>
huh?
02:23
what does that mean?
02:23
<Appiah>
me crying?
02:28kaos01 has joined #ltsp
02:29
<kaos01>
hi, trying to get ltsp-server-5 working on fedora, 4.2 worked fine but now im having trouble getting the xdm greeter up, it complains about "Unable to locate/open config file
02:29
the X server that is
02:30
<Appiah>
the xdm cofig?
02:30
<kaos01>
mmm, what xdm config ?
02:30
<laga>
ltsp 5 does not use xdm
02:31
<kaos01>
sorry im using xdmcp for now with it
02:31
i know i can use ldm
02:31
<johnny>
you should use ldm
02:31
<kaos01>
but first i just want it to work teh way it did before
02:31
<johnny>
it'll work better with ldm
02:32alkisg has quit IRC
02:32
<kaos01>
the thing is other machines work just this one doesnt and it worked with previous ltsp
02:35alkisg has joined #ltsp
02:37
<kaos01>
ok, ldm gives me : xauth: creating new authority file /var/run/ldm-xauth-Mr....
02:37
and that line just flashes on teh screen
02:39
the thing is i cant even get a shell to try to debug it
02:41
<alkisg>
kaos01: try with XSERVER=vesa and SCREEN_02=shell
02:41
So that you get a shell and run X -configure to see what's wrong..
02:42
<kaos01>
im gettign the shell but the screen blinks and one cant type into it
02:42
<alkisg>
that's the x server restarting, if you set it to vesa (or if you don't use X at all) you'll be able to type
02:42
So modify lts.conf accordingly...
02:48polytan has joined #ltsp
02:49
<nubae>
morning europeans...
02:49
<alkisg>
heh.... morning Austria! :)
02:49
<kaos01>
morning mrshians
02:50
<nubae>
yeah I suppose we kinda are... to Americans :p
02:50
<cyberorg>
johnny, Xephyr is in xorg-x11-server-extra on opensuse
02:50
<johnny>
what does that mean..?
02:51
<cyberorg>
Jan 20 03:40:20 <johnny> cyberorg, does suse have a package for xephyr ?
02:53
<kaos01>
ok, i got a shell, but cant run startx
02:54
<alkisg>
kaos01: try X -configure
02:55
<kaos01>
Failed to load module "intelmaster" ...
02:55
Failed to load module "intelbatchbuffer" ...
02:56
<johnny>
cyberorg, what i mean.. is will there be a problem if sabayon is to require it?
02:56
<kaos01>
Unable to locate/open config file: "//xorg.conf.new"
02:59
<cyberorg>
johnny, sabayon package was fixed recently changelog entry "- Require xorg-x11-Xnest and python-xdg; Sabayon won't work otherwise."
03:00
<johnny>
we're going to change it to xephyr
03:01
or at least we plan to
03:02kaos01 has quit IRC
03:09polytan has quit IRC
03:10F-GT has quit IRC
03:10dberkholz has quit IRC
03:10shamino has quit IRC
03:10Ryan52 has quit IRC
03:11F-GT has joined #ltsp
03:11Ryan52 has joined #ltsp
03:11dberkholz has joined #ltsp
03:11shamino has joined #ltsp
03:27polytan has joined #ltsp
03:38alkisg has quit IRC
03:40polytan has quit IRC
03:59alkisg has joined #ltsp
04:39
<cyberorg>
guys congratulate me, got local usb working with rdesktop+linux xrdp server :) (thanks to gadi)
04:39
<Appiah>
nice
04:39
Please put it somewhere so we know how!
04:39
<cyberorg>
ogra, removed the password export hack, now doing directly from ldm.c
04:40
Appiah, hopefully it will be in ldm-trunk soon, we still have sound to fix :)
04:40
<Appiah>
aah I see
04:40
<ogra>
cyberorg, perfect :)
04:42
<ltsppbot>
"cyberorg" pasted "nomad rdesktop from ldm with usb support" (93 lines) at http://pastebot.ltsp.org/204
04:43
<cyberorg>
ogra, i don't speak c, but what is in that patch works :)
04:44
i dont really understand sshutils.c changes, but that seem to keep the socket alive
04:45
playing movie over rdp is much faster :)
04:56
<alkisg>
Ryan52: ------ Results ------
04:56
1st try, I only extracted the ldm from the .deb. I got "Unexpected terminated process: pid = 5538". A `ps aux` while ldm was still running revealed that this was the `ssh -Y -t -M ...` session. And of course ldm hanged in an endless loop full of "waitid returned an error: No child processes".
04:56
2nd try, I installed the whole .deb - same results.
04:56
3rd try, I used the ldm binary that contained the "break" command => I still got the same errors, but without the endless loop, so from a user perspective everything was fine.
04:56
Of course "break" cures the _symptom_ and not the problem (ssh terminating unexpectedly) but I think it _should_ be there even if the main problem gets solved.
04:56
This problem constantly happened in a very slow TC of mine (300 MHz), and I've only seen it happening twice in faster ones (633 MHz).
04:56
<ltsppbot>
"alkisg" pasted "no-break-ldm.log" (73 lines) at http://pastebot.ltsp.org/205
04:57
"alkisg" pasted "no-break-psaux" (56 lines) at http://pastebot.ltsp.org/206
04:57
"alkisg" pasted "with-break-ldm.log" (68 lines) at http://pastebot.ltsp.org/207
04:58
"alkisg" pasted "with-break-psaux" (56 lines) at http://pastebot.ltsp.org/208
05:11alkisg has quit IRC
05:17isojussi has joined #ltsp
05:23
<isojussi>
i have 2 clients which worked with 7.04 I update to 7.10 and after that they stuck at starting nbd client process connecting.. activating
05:37ghaleb has joined #ltsp
05:44
<isojussi>
I need to disable quiet splash to see bootuptext
05:44alkisg has joined #ltsp
05:46
<isojussi>
it was something to do witch echo
05:46
forgot it
05:51
<alkisg>
isojussi: gksu gedit /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
05:53Shingoshi has quit IRC
05:53isojussi2 has joined #ltsp
05:55Shingoshi has joined #ltsp
05:57
<isojussi>
both machines got stuct that starting nbd client process connecting.. activating
05:57
stuck
05:57
all other clients work
06:11isojussi2 has quit IRC
06:34dirigeant has joined #ltsp
06:38
<isojussi>
it seems that it cannot start x somehow
06:39
screen blinks every few seconds
06:40F-GT has quit IRC
06:40
<isojussi>
it worked with 7.04 out of box
06:40
with 7.10 no
06:47ghaleb has quit IRC
06:49gentgeen__ has joined #ltsp
06:52F-GT has joined #ltsp
06:53chrisinanothere has quit IRC
06:55faustino3331 has joined #ltsp
06:59BrunoXLambert has joined #ltsp
07:00alekibango has quit IRC
07:01
<alkisg>
isojussi: why don't you continue with the updates to 8.04 LTS and then try to solve the X problems?
07:06dirigeant has quit IRC
07:07Eghie has joined #ltsp
07:09polytan has joined #ltsp
07:09
<polytan>
hi
07:10
johnny, hello
07:10faustino333 has quit IRC
07:11alkisg has quit IRC
07:12pmatulis has joined #ltsp
07:12alkisg has joined #ltsp
07:13dirigeant has joined #ltsp
07:14evilx has joined #LTSP
07:14alkisg has quit IRC
07:16alkisg has joined #ltsp
07:19ajohnson has quit IRC
07:21
<polytan>
is there anyone ?
07:21
on ltsp/ubuntu
07:22
<Appiah>
alot of people use ubuntu
07:23* pmatulis raises his hand
07:23
<polytan>
ok
07:24
currently, I'm not able to open a usb stick on the client
07:24
and how to use the tool "student-control-panel" ?
07:24
it starts but no students are displayed :/
07:25
Here, I have 1 server and 7 clients started (and working ;) )
07:25
<pmatulis>
polytan: never used student-control but there is ltsp documentation for debugging access to a block device
07:26
<polytan>
ok
07:27dirigeant has quit IRC
07:27
<polytan>
thanks
07:27
It works indeed very well on ubuntu
07:30
<alkisg>
polytan: try to put a stick on the client BEFORE the client boots and keep it there until the user logs in. If it works this way, I have a workaround.
07:30
polytan: and instead of student-control-panel, try https://help.ubuntu.com/community/UbuntuLTSP/iTalc
07:33
<pmatulis>
polytan: https://help.ubuntu.com/community/EdubuntuFAQ#I can't get USB storage keys/devices to work on thin clients
07:33
<alkisg>
!localdevs
07:33
<ltspbot>
alkisg: Error: "localdevs" is not a valid command.
07:33
<alkisg>
!localdev
07:33
<ltspbot>
alkisg: "localdev" is (#1) Information regarding using pluggable devices and cdroms with LTSP can be found at http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev, or (#2) for ubuntu: https://wiki.ubuntu.com/EnableLTSP5LocalDevices, or (#3) for debian-edu: http://wiki.debian.org/DebianEdu/HowTo/LocalDeviceLtspfs, or (#4) http://wiki.ubuntu.com/DebugLocalDev
07:33
<polytan>
I tryied to install gnome wtach dog but there isn't amd64 packages
07:34
<pmatulis>
polytan: what is gnome watch dog?
07:34dmaran has joined #ltsp
07:35
<polytan>
something to kill user process
07:37
<pmatulis>
do you need it?
07:40ogra has quit IRC
07:40ogra has joined #ltsp
07:41
<polytan>
pmatulis, I don't know
07:42
<pmatulis>
i would forget about it for now and research it later
07:42vvinet has joined #ltsp
07:45tfl_total_nub has left #ltsp
07:45
<polytan>
yes
07:45
first I need the usb sticks
07:47
<alkisg>
polytan: which ubuntu version?
07:47
<polytan>
the last one
07:47
8.10
07:47
<alkisg>
ok, try what I told you above
07:49
<polytan>
I'm doning it now
07:58dirigeant has joined #ltsp
08:02six2one has joined #ltsp
08:02polytan has quit IRC
08:07cliebow has joined #ltsp
08:10polytan has joined #ltsp
08:11brendan0powers has joined #ltsp
08:11
<polytan>
alkisg, on my ubuntu, tftp tries to download from /ltsp/i386 but the generated image is in /ltsp/amd64
08:11brendan0powers has left #ltsp
08:11
<polytan>
where do I have to go to chenge the default location ?
08:12
<alkisg>
polytan: /etc/ltsp/dhcpd.conf
08:12
polytan: but do you really have amd64 clients?
08:12
<polytan>
yes
08:12
<alkisg>
ok
08:12
<polytan>
why is there a file here !
08:13
<alkisg>
?
08:13
<polytan>
the dhcp file is not in /etc/dhcp3 ?
08:13
<alkisg>
No. This is specific to ltsp.
08:13
/etc/ltsp/dhcpd.conf is the file you need to modify.
08:13
<polytan>
that's a no-sens
08:15
I understand that ltsp purpose a confiog file
08:15
but it stay in the right folder
08:16
nevertheless, it works, I can boot amd64 machines
08:19dirigeant has quit IRC
08:21
<_UsUrPeR_>
morning all
08:23
<polytan>
hi
08:23
alkisg, usb works well now
08:23
thanks :)
08:23
I will try to install italc now ;)
08:24
is there any interest to use amd64 ?
08:24
for thin client I mean
08:31polytan has quit IRC
08:32polytan has joined #ltsp
08:34vvinet has quit IRC
08:35
<alkisg>
polytan: usb works with what? with the amd64 clients?
08:36Gadi has joined #ltsp
08:36
<polytan>
yes, on the amd64 machines
08:36
<alkisg>
So there's no need for workarounds? good!
08:37
<polytan>
I'm doing my tests
08:37
<sbalneav>
Morning all
08:38
<alkisg>
!s
08:38
<ltspbot>
alkisg: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
08:38
<alkisg>
:P
08:38
<polytan>
alkisg, is there any interest to have a 64bit image on thin client ?
08:38
<alkisg>
polytan: what do you mean "interest"?
08:38
<cyberorg>
!g
08:39
<ltspbot>
cyberorg: "g" is Gadi!!!!!!!!!!!!!!!!!!!!!!!!
08:39
<cyberorg>
Gadi, got sound and USB both now, hopefully in an acceptable manner :)
08:39
<polytan>
alkisg, I mean power increase, for example
08:40
<alkisg>
Ah, I don't know, I've never had 64bit clients
08:40
<polytan>
and it works well (with usb) on both base system (amd64 and i386)
08:40
<Gadi>
cyberorg: sounds good
08:41hanthana has quit IRC
08:42hanthana has joined #ltsp
08:42
<alkisg>
sbalneav: if/when you have the time, could you have a look at today's logs (http://www.nubae.com/logs/ltsp20090122_pg1.html - [03:56]) ? I think a small patch (http://slexy.org/raw/s271cNktyK) is needed in ldm.c, because on some TCs logout doesn't work...
08:43
ssh dies unexpectedly, so ldm goes into an endless loop... but with a "break" there, everything works fine.
08:44
<ltsppbot>
"cyberorg" pasted "nomad rdesktop from ldm with local cd/dvd/usb and sound" (110 lines) at http://pastebot.ltsp.org/209
08:44bobby_C has joined #ltsp
08:45
<cyberorg>
Gadi, and export `cat $HOME/.ltspvars` && rm $HOME/.ltspvars || true in /etc/bash.bashrc.local for sound
08:45ajohnson has joined #ltsp
08:48
<Gadi>
cyberorg: certainly looks cleaner. does xrdp not support sound redirection?
08:48
<cyberorg>
Gadi, nope, doesnt look like it will any soon
08:49dirigeant has joined #ltsp
08:49
<Gadi>
sounds like they are determined to produce a feature-free implementation of rdp
08:49
:)
08:50
<stgraber>
Gadi: bugger
08:51
Gadi: LTSP_LOCALAPPS_MENU_ITEMS is called after a su -
08:51
Gadi: *used
08:51
<Gadi>
so?
08:51
<stgraber>
Gadi: so ltsp-genmenu is always run with an empty environment :)
08:51
making LTSP_LOCALAPPS_MENU_ITEMS always empty :)
08:51
<Gadi>
oh
08:51
<cyberorg>
will put comments and see if there is anything else that can be improved, then will send in the patch
08:51
<Gadi>
hehe
08:51
lets fix that
08:52
<stgraber>
we can just send that var when calling the su
08:52
LTSP_LOCALAPPS_MENU_ITEMS=${LTSP_LOCALAPPS_MENU_ITEMS}
08:53mikkel has joined #ltsp
08:53
<polytan>
alkisg, I've got an error when I chroot the client and install italc-client
08:53
a problem with mDNS
08:53
and I can't start italc on the server (don't know why)
08:54
started
08:54
<pmatulis>
liw: confirmed, the nfs "mounter" will wait for the nfs server, i'll make them both 'autostart' and see what happens
08:56
<stgraber>
[ -x "/usr/bin/ltsp-genmenu" ] && su - ${LDM_USERNAME} -c "LOCAL_APPS_MENU_ITEMS=${LOCAL_APPS_MENU_ITEMS} /usr/bin/ltsp-genmenu install"
08:57CAN-o-SPAM has joined #ltsp
08:58
<sbalneav>
alkisg: Yup, that's definitely a bug.
08:58
<alkisg>
sbalneav: ty :)
08:58
<stgraber>
Gadi: pushing that upstream
08:58
<alkisg>
I could only regularly reproduce it on the very slow client, though
08:59
<Gadi>
stgraber: just as I was thinking
08:59
<stgraber>
pushed
09:01
<sbalneav>
alkisg: I'll patch your patch a bit so it can be i15d
09:01
<alkisg>
i15d? :) I know i18n and i10n only!!!
09:02
But I'm sure you'll do the right thing! :)
09:02vvinet has joined #ltsp
09:02
<alkisg>
(Gadi did the patch, I only suggested the "break" there)
09:03
Sorry, I meant Ryan52
09:03
<sbalneav>
internationalized
09:04
pushed 1058
09:05warren has joined #ltsp
09:07
<alkisg>
l10n = localization, i18n = internationalization, i15d = internationalized - and the numbers are just the # of characters between the starting and the ending letter... Got it at last! :D
09:09
<warren>
ogra: https://www.redhat.com/archives/fedora-devel-list/2009-January/msg01389.html
09:10
ogra: I hear Ubuntu had this problem before? How did you solve it?
09:11
<ogra>
warren, ubuntu doesnt use such evlish shutdown methods :)
09:12* warren read "elvish" at first and thought ESR.
09:12
<warren>
ogra: how do you ensure that all processes are killed?
09:12
<ogra>
i dont think something like killall is used in any official script in debian or ubuntu
09:13faustino333 has joined #ltsp
09:13MeW_ has joined #ltsp
09:15
<ogra>
warren, oh, i take that back
09:15
we do the same evil stuff
09:15
<warren>
haha
09:15
<ogra>
sending -15 and then -9
09:16* ogra cant belive killall is used in that script
09:16jammcq has joined #ltsp
09:16
<jammcq>
hello #ltsp
09:16
<ogra>
i was whacked so often in debian and ubuntu using killall
09:16
<warren>
ogra: and you don't have trouble turning stuff off and unmounting *after* the nbd block devices are gone?
09:16dirigeant has quit IRC
09:16
<ogra>
warren, why do you care ? its a redonly mount anyway
09:17
<warren>
ogra: I care because my system gets stuck
09:17
<ogra>
as long as poweroff/reboot works
09:17
<warren>
during shutdown
09:20faustino3331 has quit IRC
09:20
<alkisg>
sbalneav: I don't think both log errors are needed, probably the line 'logerr(_("sigwaitinfo returned an error"));' should be completely removed now. Ryan52 changed the error message because waitid is used and not sigwaitinfo...
09:20
<warren>
ogra: somebody says Ubuntu had this problem with nbd before
09:21
<ogra>
warren, i dont know, we used to run halt -fp and reboot -fp the last time i looked
09:22
that skips the whole shutdown process
09:22
if you get any hangs with that, it will be ACPI probs and not init related at all
09:22
<warren>
ogra: wasn't it YOUR decision to change ldm to shutdown "cleanly"?
09:22wild_oscar has joined #ltsp
09:22
<ogra>
my "suggestion"
09:22
we didnt implement it yet
09:23
or did anyone ?
09:23
<warren>
ogra: upstream did a long long time ago
09:23
<ogra>
how ?
09:24
by just blindly calling telinit ?
09:24* ogra didnt know someone just changed without properly preparing the chroot to cope
09:25
<ogra>
ltsp-build-client needs a new RC0_WHITELIST and RC6_WHITELIST ... and the sendsigs initscript needs to be skipped
09:25
well, probably better a blacklist
09:26
such stuff needs to be properly speced before blindly switching over
09:26
you also want to skip the initscritpt that shuts down networking functions
09:26
<warren>
does ubuntu use killall5 during shutdown?
09:26
<ogra>
in case fedora has something like that
09:26chrisinajar has joined #ltsp
09:26
<ogra>
yes
09:27alkisg has quit IRC
09:27
<ogra>
/etc/rc0.d/S20sendsigs and /etc/rc0.d/S35networking need to be skipped in ubuntu
09:29
<warren>
My diskless client doesn't use the network service, so it wont bring it down
09:29
<ogra>
right
09:30
well, but we need something similar to the RC2_WHITELIST and RCS_WHITELIST vars
09:30
you will have similar probs with nfsroot i guess
09:31
if nfs is brought down to early you wont be able to halt
09:32* ogra wonders if anyone here has a clue about qemu in -nographics mode
09:40faustino3331 has joined #ltsp
09:41isojussi_ has joined #ltsp
09:42
<isojussi_>
what was name of that package that enabled sounds with firefox+flash like in youtube?
09:42
<kc8pxy>
johnny: still here?
09:44MeW_ has quit IRC
09:44
<warren>
isojussi: no package is needed anymore for flash 10
09:44
isojussi: what distro are you using?
09:45
<isojussi_>
ubuntu 7.10
09:45
<ogra>
libflashsupport then
09:46wild_oscar has left #ltsp
09:46
<warren>
flash 9 needs libflashsupport, but it causes crashes
09:46
<ogra>
not in that iteration
09:46
in 7.10 it was stable
09:47
and i also thing 7.10 had flash 7 or 8
09:47
*think
09:47
its so long ago :)
09:47
<warren>
ogra: all versions of flash that needed libflashsupport have a race condition in sound output that causes crashes.
09:48
<ogra>
well, i used it for 6 months without a single crash
09:48
the crashing started in 8.04
09:48
with flash9 and the newer pulse
09:48
<isojussi_>
how i can determine whic version flash i have installed?
09:48
<kc8pxy>
johnny: i got the gui to work, using my account on the server, once i enabled X11forwarding on the sshd. please add openrc to the ldm deps, as well as keywords for udev-133 and sysvinit. i also (had to?) keyword and unmask portage2.2_rc20. those should be in the chroot-building scripting.
09:48
<ogra>
about:plugins in firefox
09:49* ogra would consider upgrading to 8.10 anyway ... 7.10 will be EOL in 3 months
09:51
<laga>
or maybe use an LTS release :)
09:51
<isojussi_>
i have 5 servers in different locations. a bit work to update them, but those are under "construction"
09:56isojussi_ has quit IRC
09:57faustino333 has quit IRC
09:59cliebow has quit IRC
10:03
<warren>
Ryan52: https://bugzilla.redhat.com/show_bug.cgi?id=478976 dude, awesome
10:06
<Ryan52>
:)
10:06
<warren>
Ryan52: was this submitted upstream?
10:07japerry has quit IRC
10:07
<Ryan52>
dunno
10:17
warren, so y'all don't just throw patches into packages and call it done? :)
10:17
<warren>
Ryan52: because we're not fucking Ubuntu
10:18
<sbalneav>
Classy
10:18
<Ryan52>
:)
10:18
<warren>
sbalneav: I can say it again while holding a champagne glass.
10:18* nubae kicks moodle
10:18
<SDuensin>
Morning.
10:20
<Ryan52>
heh
10:20cliebow has joined #ltsp
10:21
<cliebow>
davidj:your leg all healed?
10:22Eghie has quit IRC
10:30spectra has joined #ltsp
10:31
<warren>
Ryan52: they're requesting that you please submit it upstream
10:36faustino3331 has quit IRC
10:42staffencasa has joined #ltsp
10:47japerry has joined #ltsp
10:53tjikkun_work has quit IRC
10:56cliebow has quit IRC
10:57cliebow has joined #ltsp
11:08otavio has quit IRC
11:08otavio has joined #ltsp
11:08sepski has joined #ltsp
11:14lilly_bn has joined #ltsp
11:14
<lilly_bn>
hi
11:14
can some one help me with nfs
11:15
i have this errors
11:15
unable to mount root fs on unknown block
11:19japerry_cat has joined #ltsp
11:20
<sbalneav>
lilly_bn: Perhaps you could give us some more information?
11:21
Like, if this is in relation to LTSP, what version it is, what host os you're running it under, etc.
11:32bobby_C has quit IRC
11:33japerry has quit IRC
11:35warren has quit IRC
11:40pmatulis has quit IRC
11:41pmatulis has joined #ltsp
11:42otavio has quit IRC
11:42otavio has joined #ltsp
11:43
<lilly_bn>
i running gentoo
11:43
but i found now that nfs running on 127.0.0.1
11:43
how can i change to run on eth0 internet
11:44
interface*
11:44
or on all interfaces
11:47
i need to configure nfs
11:47
:D
11:47vvinet has quit IRC
11:47
<lilly_bn>
only that
11:49
<nobios>
you configure nfs in /etc/exports
11:50japerry_cat has quit IRC
11:50
<lilly_bn>
./opt/ltsp-4.2/i386 0.0.0.0/0.0.0.0(rw,fsid=0,insecure,no_subtree_check,sync)
11:51
/var/opt/ltsp/swapfiles 192.168.234.2/255.255.255.0(rw,fsid=0,insecure,no_subtree_check,sync)
11:51
without "."
11:52
i can mount like this
11:52
mount -t nfs 127.0.0.1:/opt/ltsp-4.2/i386/ /root/
11:52
<nobios>
I think you can swap 0.0.0.0/0.0.0.0 to *
11:52
<lilly_bn>
but i cant like this mount -t nfs 192.168.234.2:/opt/ltsp-4.2/i386 /root
11:52
swap nevermind
11:52pmatulis has quit IRC
11:52
<lilly_bn>
i need ro mount over network
11:53
now i only can mount over lo
11:53
<nobios>
maybe it's because of the dot ?
11:53
<lilly_bn>
no dot is mistake in print
11:53
:D
11:53
<nobios>
ok :)
11:53
<lilly_bn>
nfs deamon runing on 0.0.0.0
11:54
<nobios>
so, rpc is running?
11:54pmatulis has joined #ltsp
11:54
<lilly_bn>
tcp 0 0 *:nfs *:* LISTEN -
11:54
tcp 0 0 *:33961 *:* LISTEN -
11:54
tcp 0 0 localhost:sunrpc *:* LISTEN 13819/portmap
11:54
tcp 0 0 *:39153 *:* LISTEN 18590/rpc.statd
11:54
tcp 0 0 *:37460 *:* LISTEN 18651/rpc.mountd
11:54
tcp 0 0 *:ssh *:* LISTEN 8523/sshd
11:54
sorry
11:55
is running
11:55
rpcbind: server 192.168.234.2 not responding, timed out
11:56* nobios put * in /etc/exports and try to restart portmap and nfs
11:56lilly_bn is now known as Guza
11:58
<polytan>
how to build a ubuntu kernel with ubuntu tools ?
11:58Blinny has joined #ltsp
11:58
<polytan>
(I mean not using make, etc"
11:58
)
11:59RobertLaptop has quit IRC
11:59
<Guza>
done but with no succes
12:00alekibango has joined #ltsp
12:03
<nobios>
Guza: look at your logs and see if you find something that can help
12:04
<Guza>
[ 9191.103421] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
12:04
[ 9191.104712] NFSD: starting 90-second grace period
12:04
only this
12:04
in dmesg
12:08
my version of nfs 1.1.3 and nfsv4
12:08Lns has joined #ltsp
12:09
<Blinny>
Lns: You had any luck w/ testing those firefox changes? I've been following on ltsp-discuss
12:09
<Lns>
Blinny: I have, and they work great. Hold on thou - on phone
12:09faustino333 has joined #ltsp
12:09
<Lns>
(nspr fixes from PPA is what fixed it)
12:11
I'm wondering if there are further optimizations as far as pixmap caching (in Firefox about:config) that I haven't touched though, FF3 is working but they say it crashes "about as much" as FF2 - which is not much, but still
12:11
<Blinny>
still
12:11
That's the local xserver RAM running out right?
12:13
I forget - are you rolling your own binary or just tweaking :config values?
12:13
<Lns>
Blinny: correct
12:13
Blinny: I installed the updated nspr packages from the PPA described in the bug report
12:14
Tweaking about:config also helps a lot
12:14
<Blinny>
Dig.
12:15
I haven't been hit with that yet, but as I bring on more terminals with less over-engineered servers, I may see some issues.
12:16six2one_ has joined #ltsp
12:16dirigeant has joined #ltsp
12:17
<Lns>
Blinny: It mostly happens when users launch FF3 at the same time. One at a time doesn't affect it *as much* - but my sites are mainly student computer labs where they come in all at once and are instructed to do things like launch FF all at the same time.
12:18
<Blinny>
Right.
12:18six2one_ has quit IRC
12:18six2one_ has joined #ltsp
12:18
<Blinny>
Here we have weekly staff meetings from which everyone exits stage left, and walks straight to their desks/terminals.
12:19six2one_ has quit IRC
12:19six2one_ has joined #ltsp
12:19six2one has quit IRC
12:20
<polytan>
I'm building an other kernel for the server and the thin client
12:20
how to put the kernel in place ?
12:21
I mean, I can copy the kernel in /tftpboot/ltsp/i386
12:21
but for the initrd ?
12:21
how to modify it ?
12:21
<ogra>
https://help.ubuntu.com/community/Kernel/Compile
12:23
<polytan>
thanks ;)
12:27
Alternate Build Method: The Old-Fashioned Debian Way => yeah
12:28otavio has quit IRC
12:29six2one_ has quit IRC
12:35japerry has joined #ltsp
12:37otavio has joined #ltsp
12:39
<Guza>
i have this error on client terminal: Kernel Panic - not syncing: VFS Unable to mount root fs on unknown-block(8,3)
12:40
<stgraber>
Gadi: around ?
12:40warren has joined #ltsp
12:40
<Guza>
where to configure root?
12:40
<stgraber>
Gadi: I'm testing ltspfs with a standard ldm session and it doesn't seem to work
12:40
Gadi: do you have a few minutes to help me debug ?
12:41
Gadi: ltspfs_fstab is correct but the device isn't mounted
12:42
Gadi: it's the same usbkey that works correctly when using rdesktop instead of ldm
12:42
<warren>
stgraber: saw my post?
12:44
<Gadi>
stgraber: I'm here
12:44
stgraber: do you see /media/root on the client?
12:44
<stgraber>
Gadi: no
12:44
Gadi: ltspfs_fstab is fine /var/run/drives/usbdisk-sda is here but nothing is mounted
12:45
<Gadi>
sda?
12:45
not sda1?
12:45
<stgraber>
Gadi: looks like the application server didn't get the mount xprop (or similar, I don't really know how it works)
12:45
Gadi: that's not the issue, /dev/sda is fine, it doesn't have a partition table
12:45
<Gadi>
ok
12:45
<stgraber>
Gadi: as I said, the same works fine using rdesktop
12:45
<Gadi>
ok, ok
12:45
<stgraber>
something broke with ldm
12:46vvinet has joined #ltsp
12:46
<Gadi>
can you:
12:46
unset SSH_CONNECTION
12:47
<stgraber>
sure
12:47
<Gadi>
eetc
12:47
etc
12:51
<stgraber>
do I have to do something else than the ltspfsmounter blabla add ? something ldm-specific ?
12:51
I know that local mounting in /media/root/... works
12:52
ok, I set DISPLAY and XAUTHORITY
12:52
then did the ltspfsmounter and the result is ... nothing :)
12:52
I got: Error: /media/root/ is not mounted
12:53
but it shouldn't even use that directory
12:53
Gadi: ^
12:57
<Blinny>
Guza: What distro / LTSP version?
12:57
Guza: You can google "nfs ltsp VFS Unable to mount root fs" and get some information
13:01polytan has quit IRC
13:05alkisg has joined #ltsp
13:05
<SDuensin>
Is it possible to remove the "Places" menu in GNOME?
13:06
Locking down this server is like herding cats.
13:07
Grr. You can still edit the locked-down menus, too, by right-clicking them.
13:08
<johnny>
locked down how?
13:08
<SDuensin>
I set them how I want them for restricted users and disabled the menu editor and all the "Run..." things I can find.
13:08
But they seem to still be able to edit them by right-clicking a menu and selecting "Edit".
13:09
<johnny>
how did you do that?
13:09
<SDuensin>
Which part?
13:09
<johnny>
any of it
13:09
<SDuensin>
hehe
13:10
Mix of GConf and the menu editor in GNOME's prefs.
13:10
<nubae>
johnny: how close is gentoo now with ltsp, I'd like to try it out
13:10
<johnny>
SDuensin, sabayon will be nicely patched up soon
13:10
give us a few days
13:10* SDuensin cheers wildly!
13:11
<johnny>
you'll have to install it from a ppa probably tho
13:11
<SDuensin>
I can do that. NOOoooo problem!
13:11
<johnny>
nubae, i'm about to do another test run, make sure the chroot works perfectly from boot (minus the initramfs issues, and setting up dhcp/tftp manually)
13:11
nubae, having a bit of a problem
13:11
with my computer
13:12
very confused.. but i'll get through it
13:13
<nubae>
I just think gentoo is a good choice because of the level of customisation possible
13:13
<johnny>
nubae, talk to me in a few days.. hopefully my issues will be sorted by then
13:13Blinny has quit IRC
13:13
<johnny>
i can't build xulrunner, nor wine atm
13:13
<nubae>
cool
13:14
<johnny>
i accidentally had 32bit stuff that shouldn't have been 32bit
13:14
pretty sure it was my fault
13:14
i just don't recall how :)
13:16vagrantc has joined #ltsp
13:16hanthana is now known as hanthana|sleepin
13:18hanthana|sleepin has quit IRC
13:18
<sbalneav>
ogra: You still managing the ltsp-drivers group?
13:19
nubae: you've got a launchpad account, right?
13:19
<ogra>
do i ?
13:19
<nubae>
yyeah
13:19
<sbalneav>
Well, who adds people to ltsp-drivers?
13:19
<nubae>
I wouldnt have been able to push docs otherwise
13:19
<sbalneav>
right
13:19
<ogra>
sbalneav, "Team owner: Jim McQuillan"
13:19
for https://launchpad.net/~ltsp-drivers
13:20
<nubae>
I'll ask jammq when he shows up
13:20
<ogra>
wht for do you use ltsp-drivers ?
13:20
we never used it for anything afaik
13:21
<sbalneav>
nubae's got his fat client script, I suggested he just commit it and work on it within -trunk. I realize it's not part of the main install, but if he's producing stuff that's being used by people, it should just end up being part of ltsp upstream.
13:21
<cliebow>
sbalneav: doyou replicate your ldap across the internet?
13:21
<nubae>
sounds translated from German.... für was benutzt man ltsp-drivers :-)
13:21
<sbalneav>
Don't we use it for write access to the bzr -trunk?
13:21
cliebow: No.
13:22
<cliebow>
any recommends about doing it ?
13:22
<nubae>
ldap across the net, really??
13:22
<sbalneav>
I was just going to suggest that nubae get write access to bzr so he can start off by putting it in the contrib or extras section.
13:22
<ogra>
sbalneav, well, if he should get access to ltsp-upstream it makes more sense to add him to ltsp-upstream ... though note that this gives him god proveleges on all branches
13:22
<cliebow>
well..replcaiting
13:22
<sbalneav>
ah, that's what I meant
13:23
<ogra>
and for that each distro has its admin
13:23
doesnt block on me
13:23
<sbalneav>
Yeah, nubae's done some doc work, and I'll vouch for him as a regular upstanding LTSP citizen.
13:23Lns has quit IRC
13:23
<ogra>
each distro representative is a master
13:23
<nubae>
heh, hopefully I'm not considered malicious
13:23
<jammcq>
and I'll vouch for sbalneav, so it's doubly good :)
13:24
<sbalneav>
Anyone with a problem with nubae becoming upstream?
13:24
hey jammcq!
13:24mpih has joined #ltsp
13:24
<ogra>
sbalneav, and, you know there is that Scott Balneaves guy who is an admin for the team too ;)
13:24
<sbalneav>
Ah, right, I can do it.
13:24
<nubae>
heh
13:25
<sbalneav>
I forget what power I (potentially) wield
13:25
<johnny>
that guy? wtf?
13:25
hehe
13:25
<sbalneav>
heh
13:25
So, going once????
13:25
Going twice???
13:25
All in favour say aye
13:25* vagrantc would like to see more developent in branches, and merge complete features
13:26* sbalneav looks around
13:26* ogra agrees with vagrantc but wont stand in the way
13:27
<sbalneav>
Well, that's fine
13:27
<stgraber>
Gadi: still around ?
13:27
<sbalneav>
That's just workflow, right?
13:27
<ogra>
especially sine i wont be the guy who merges them
13:27
<nubae>
ok, so create my branch and push that way
13:28
<vagrantc>
sbalneav: perhaps, but when bringing new people on, it's possibly a good time to ensure consistant development policies...
13:28
or at least evaluate them.
13:28
<stgraber>
Gadi: oh, sorry, I have an idea of what's wrong :)
13:28
<ogra>
LP has this neat new "propose branch for merging" button btw :)
13:28
<sbalneav>
Do we have a development policy document?
13:28
<stgraber>
Gadi: /dev/fuse not writeable ... sounds like an openvz bug
13:28
<vagrantc>
suppose i should really work on crafting a development policy which documents this sort of thing, so we can have people read it...
13:28
<nubae>
fatclient now is just ubuntu, but I'd like to make it fedora and gentoo too
13:29
<ogra>
vagrantc, ++
13:29
<sbalneav>
I'd love to see one, seeing as how it usually ends up being me messing things up :)
13:29
<vagrantc>
nothing too absurd, but giving a brief codingstyle explanation, documenting cleaner workflows and such...
13:30
<ogra>
nubae, well, then working in a branch is just perfect
13:30
<nubae>
that totally sounds awsome
13:30
sounds like a howto for ltsp development
13:30
<sbalneav>
nubae: ok, so branch, add in, and when you're ready to know, let us know.
13:30
<nubae>
ok
13:30
<ogra>
nubae, and with the metioned propose feature, if you select the ltsp-upstream branch, all members of ltsp-upstream will get an autonotification so it wont get lost
13:31
"Propose for merging into another branch" on the page of your branch ...
13:31
<vagrantc>
maybe a good guideline woudl be to have people do a few merge requests before getting direct commit access? that gives us some room to evaluate beforehand, comment on the code, and still gives a good idea of how they work
13:32
<ogra>
yeah
13:32Lns has joined #ltsp
13:33bobby_C has joined #ltsp
13:33
<ogra>
"If you want to join the team please find the right team member for the distro you work with and commit your patches through that person for a while before applying."
13:33
vagrantc, ^^^
13:34Guza has quit IRC
13:34
<ogra>
https://launchpad.net/~ltsp-upstream already says that
13:34
<vagrantc>
ogra: sounds pretty good to me.
13:34
ogra: even better :)
13:34
<ogra>
should just be a bit extended
13:34
<vagrantc>
yeah.
13:34
<ogra>
like "work in your own branches for a while"
13:34
<nubae>
yes, coding conventions, what to work on, etc
13:34
would be nice to have some sort of wishlist that devs could get involved with
13:34
<ogra>
"get frequent reviews from team members"
13:36RobertLaptop has joined #ltsp
13:36
<ogra>
vagrantc, btw, since i got you here, do you have any idea how to use a second serial port in qemu ? (i'm trying to establish a fifo for IPC inside the vm through that)
13:36
<vagrantc>
ogra: no. i've only used a single serial port as a console ... the syntax for serial stuff in qemu is a little tricky, for sure.
13:36RobertLaptop has quit IRC
13:37
<johnny>
so.. what about the dbus efforts..
13:37
<ogra>
well, i have a single serial port already attached to a fifo which i use for the output
13:38
i'd like a second one for the input and error catching so i can feed return values into the wrapper script
13:38
<nubae>
ogra: I saw u started a fatclient branch a while ago (7.10 I believe), what happened to that?
13:38
<ogra>
nothing
13:39
i was hoping to get it approved by the ubuntu techboard ...
13:39pasmen has joined #LTSP
13:39
<ogra>
but i didnt, so i didnt work on it
13:39
<nubae>
so there's nothing worth ripping from there?
13:39
<ogra>
nah
13:40
i think i developed the first webkiosk version for ubuntu in it
13:40
yours is already far beyond
13:40
<nubae>
u know, I'm surprised at the amount of people that started using the fatclient script... must be in the hundreds now...
13:42
<Lns>
nubae: it's a great alternative to remote X sessions, that's for sure
13:42
<nubae>
wish it had a better management integration
13:42
<_UsUrPeR_>
where does gnome store it's monitor resolution settings in user /home files?
13:43
ubuntu 8.10 ^^
13:43
<sbalneav>
_UsUrPeR_: it's in gconf
13:43
<_UsUrPeR_>
sbalneav: thx
13:43
<nubae>
if I could visually install whatever on the fatclient on the server, that would be great
13:43
<sbalneav>
use gconftool-2 to set/unset
13:43
<nubae>
I suppose I could use zenity
13:44
<sbalneav>
_UsUrPeR_: /desktop/gnome/screen/default/0
13:44
usually
13:44
<_UsUrPeR_>
sbalneav: ok, thanks
13:44
<Lns>
Any particular reason why cron isn't installed in ubuntu ltsp chroot by default? Is it simply not needed?
13:44
<warren>
Ryan52: ping
13:45
<ogra>
oh, new gdm gets considered for jaunty+1
13:45
<nubae>
hmm I thought it was... I don't install it extra
13:45
<sbalneav>
Lns: not needed, usually
13:45
and on older slower boxes, one less process to run.
13:45
<Lns>
sbalneav: right
13:45
ok.. well i'm going to install it and write a wiki howto for doing things like automating client shutdown via cron. :)
13:46
<nubae>
hey, last we talked about memory usage... we concluded that ldm was using an absurd amount of ram, has that been fixed?
13:46
<vagrantc>
seemed like the switch from python ldm to C ldm mostly reduced the memory useage ...
13:46
<sbalneav>
Lns: Better still why not write it up, email it to me or nubae and we'll put it in the docs.
13:46
<nubae>
reduced to what?
13:47
<Lns>
sbalneav: sure, np.
13:47
<vagrantc>
nubae: well, how bad was it "last we talked" :P
13:47
<ogra>
nubae, 8M resident or some such
13:47
<nubae>
think 25 mb
13:47
ok 8mb is really good then
13:47
<ogra>
its like that since years ?
13:48
<nubae>
no, this was about 6 months ago....
13:48
<vagrantc>
i haven't looked at the memory footprint in a while, but i do most of my testing in 64MB ram thin clients
13:48
<nubae>
we did a test to see what ram was used on startup
13:48
I brought it up because there are still people wanting to run on 32mb
13:48
<sbalneav>
ps -aux on my ws right now shows 4 megs for ldm
13:48
<nubae>
niiice
13:49
<sbalneav>
the ssh uses way more
13:49
<ogra>
nubae, i doubt you looked at resident (actual used) ram
13:49
it uses a lot virtual and shared
13:50
which is irrelevant
13:50* ogra likes to point people to htop to get more decent numbers than with top
13:51
<Lns>
I just discovered htop the other day, too...awesome tool!!!
13:51
<nubae>
yeah thats what we used,u recommended it long time ago ;-)
13:51
there are many cool top apps
13:52
<alkisg>
htop isn't in the default chroot - and it's only 200kb ... :)
13:52
<vagrantc>
neither is atop!
13:52
<nubae>
true should be a seed
13:54
<ogra>
for ubuntu that would force it to main ...
13:54
which requires papaerwork
13:54
which i doubt anyone will do
13:55
<nubae>
not for a 200kb program, lol
13:55
<Lns>
what are we, the highway patrol?? =p
13:55
j/k
13:56
<nubae>
hmmm... so doing an MIR is that hard?
13:56
<_UsUrPeR_>
in ubuntu 8.10, where is ltsp-localapps supposed to be? I just installed a fresh install from the 8.10 Alternative cd and updated to full, and can't find ltsp-localapps anywhere. localappsd is in the chroot... just not the executor.
13:56
<johnny>
it's in the example directorry or something
13:56
it's not working out of the box til jaunty
13:57
it was hidden on purpose
13:57
<nubae>
theres something in the wiki which ogra wrote I believe
13:57
<ogra>
https://wiki.ubuntu.com/LTSPLocalAppSetup
13:57otavio has quit IRC
13:58
<warren>
hmm... the nomad patches look ok, except most of it is done in the wrong place
13:58
<nubae>
hehe, the Suse syndrome...
13:58
<ogra>
heh
13:59
nubae, nah, the suse syndrome would mean it randomly rewrites your files in /etc while running as well :P
13:59Lns has quit IRC
14:01otavio has joined #ltsp
14:03
<stgraber>
Gadi, vagrantc: As far as I'm concerned ltspfs works just fine so I'd be for tagging it. Do you have something else to test before tagging ?
14:04spectra has quit IRC
14:05
<Gadi>
stgraber: all issues are resolved for you?
14:05Lns has joined #ltsp
14:06
<Gadi>
stgraber: I believe vagrantc was having some issues on his NFS+bind-mount system
14:06
<stgraber>
Gadi: tup
14:06
Gadi: yup
14:06
<Gadi>
if it works for you, I say tag it
14:06
<stgraber>
Gadi: well, I count on vagrantc to upload it so ... :)
14:06
<Gadi>
and then we can see if vagrantc can retest with that same version
14:06
on his system
14:06
and increment a fix if it breaks on NFS+bindmount
14:06
<Lns>
allllllllllrighty then... hahaha.. I Installed cron in the chroot, added a line IN THE CHROOT to 'halt' at noon (6 min ago), ltsp-update-image, and at 12, the SERVER shuts off. Any takers? ;)
14:07
<Gadi>
also, warren can test on his setups
14:08
<warren>
Gadi: test what?
14:08
<Lns>
'sudo chroot /opt/ltsp/i386 apt-get install cron' , at the end, said "Starting Cron [ ok ]". Maybe this is a clue.
14:09
<Gadi>
warren: the ltspfs that stgraber will tagsoon
14:09
<warren>
Lns: our packaging guidelines forbid starting services at that time...
14:09
Gadi: test what specifically?
14:09
<Gadi>
and, with that package, testing installing ltspfs in the chroot
14:09
along with ltspfsd
14:10
to see if a /media/root/usbdisk-sda1 is created in the chroot upon inserting a USB stick
14:10
<Lns>
warren: ?
14:10
<warren>
root?
14:10
Lns: that looks like a debian/ubuntu bug
14:10
<Gadi>
(this way, one can: rdesktop -r disk:drives=/media/root ....)
14:10
and get localdev in rdesktop
14:11
<warren>
root user?
14:11
<Lns>
hmm...not sure how to tackle this one
14:11
<Gadi>
warren: unfortunately, yes- udev runs ltspfsmounter as root, and lbmount has hardcoded paths based upon username
14:11
hardcoded because it is setuid on the server
14:11
<nubae>
Gadi... so with this nomad rdesktop implemenation, one can disconnect a laptop and continue using it, then connect the next day, and synchronisation automatically happens?
14:11
<Gadi>
so for security
14:12
nubae: in theory
14:12
<nubae>
theory?
14:12
<Gadi>
no synchronization
14:12
<_UsUrPeR_>
sbalneav: there's no screen directory in ~/.gconf/desktop/gnome . Where else would this information be stored?
14:12
<Gadi>
nothing syncs
14:12
<warren>
nubae: nomad requires an entirely separate X server and patched rdesktop, it is a bit of a mess.
14:12
<Gadi>
your session continues to run on the server
14:12
when you disconnect
14:12
<nubae>
ok.. so if ofs was used at the end, it would do that though, right
14:12
?
14:12
<Gadi>
hehe
14:12
sure
14:13
warren: really?
14:13
I dont think so
14:13
<warren>
Gadi: yes it does.
14:13
<Gadi>
maybe to get some extended features?
14:13
<warren>
no
14:13
Gadi: kind of like NX and Xvnc are their own X server
14:13
<alkisg>
_UsUrPeR_: have a look at ./config/monitors.xml
14:14
<nubae>
hmmm, cyberorg made it sound like its more efficient than X over ssh
14:14
<Gadi>
warren: ah
14:14
<warren>
nubae: it is
14:14
<_UsUrPeR_>
alkisg: aha! :D
14:14
<Gadi>
bbiab
14:14
<nubae>
so whats the problem then?
14:15
<warren>
nubae: if you are using Ubuntu or Fedora, you need some pretty massive changes that your distro lacks
14:15
non-upstream changes
14:16
<nubae>
what kind of changes?
14:16
<johnny>
the patched rdesktop for one.. and the different xserver.. :)
14:17
<warren>
patched rdesktop is small
14:17dirigeant has quit IRC
14:17
<warren>
different X server is huge
14:18
<nubae>
its just, now with netbooks making the rounds in schools, it makes sense to be able to disconnect and reconnect to an ltsp server with synching
14:19
<Lns>
Ok this is really strange now. I had already unlocked/created a pw for my chroot's 'root' acct a while back. I logged into the chroot (tty1) as root. I went to tty7, opened up gnome-terminal, su'd to root, and it tells me "Added user root." wtf is that?
14:20
<nubae>
lag :-)
14:20gate_keeper_ has joined #ltsp
14:20
<Lns>
har, nubae, har. :p
14:20
<nubae>
:D
14:21otavio_ has joined #ltsp
14:22japerry_cat has joined #ltsp
14:23
<vagrantc>
stgraber: i won't have a chance to test ltspfs till tomorrow
14:23
so, what's the simplest way to add all users to the fuse group automatically?
14:23
pam-group ?
14:24
or even better, add categories of users ...
14:24
but let's start with the simplest case first.
14:26
<warren>
stgraber: saw my post on ltsp-developer?
14:26
stgraber: I want to change file to sha1sum, file isn't in my chroot, and sha1sum will do the job equally as well
14:27
<ogra>
how can you live without file ? :)
14:27otavio_ has quit IRC
14:27
<vagrantc>
file isn't installed in the debian chroot by default either, i think.
14:28* ogra would have thought that comes with debootstrap
14:29
<ogra>
ah, i see, in ubuntu it comes with ubuntu-standard
14:30
though defoma deps on it ... so it should be in the debian chroot as well
14:31* vagrantc checks
14:31
<ogra>
i cant imagine you dont have defoma in the chroot
14:32
<davidj>
cliebow: Hello!
14:33
<vagrantc>
file is installed by default, i think....
14:34
<davidj>
cliebow: Yes, my leg is about well, thanks for asking.
14:36litlebuda has joined #ltsp
14:36otavio has quit IRC
14:37* alkisg thinks defoma-user has the best man page he has ever seen
14:37
<warren>
Hmm, did I piss off stgraber? I've been asking since yesterday with no response.
14:38
should I just go ahead and do it?
14:39
<vagrantc>
warren: i wouldn't assume something wrong with just lack of response...
14:40otavio has joined #ltsp
14:46
<alkisg>
How can I install ltsp-server using the ubuntu desktop cd and with minimal downloading from the internet? (1) uncomment the CD in sources.list, (2) ltsp-build-client --copy-sourceslist --mount-cdrom ?
14:46
<warren>
brb
14:46
<laga>
you can't do that using the desktop CD.
14:46
<nubae>
alkisg: aptcacher
14:47
<alkisg>
Agh... laga, why? Aren't most of the packages already on the CD?
14:47japerry has quit IRC
14:47
<laga>
alkisg: the CD contains a pre-installed image
14:47
<alkisg>
nubae, aptcacher will only work if the packages were downloaded once from the internet
14:47
<nubae>
true, I kinda figured that was the case
14:48
<alkisg>
laga, really? I thought that debian-installer was still running behind the scene... pitty :(
14:48
ty!
14:48
<laga>
alkisg: you'd need the alternate disk for that
14:49
<alkisg>
That's what I was using so far, but I thought the desktop CD would be more useful (for recovery, gparted etc)
14:50dirigeant has joined #ltsp
14:54
<alkisg>
...and a similar one: If I use the amd64 alternate CD, and my clients are i386, what can I do?
14:54dmaran has quit IRC
14:54
<ogra>
alkisg, right, debian-instaler is the backend of ubiquity (the graphical installer) but it operates on top of an installed system
14:54
<alkisg>
ogra, thanks... :(
14:55
<ogra>
i.e. it runs in a livesession ... on the target device the d-i partitioner is run, the content of the live system gets copied over to the target, the system configuration is done by d-i again and then unwated packages (i.e. langpacks not used for the target language) are removed
14:55otavio_ has joined #ltsp
14:56
<alkisg>
Ah, makes sense. I thought it was installing packages, but this would need twice the space for no benefit.
14:58
What's the usual process to install ltsp on an amd64 server with i386 clients? Download the alternate cd, install, remove the chroot (it's an amd64 chroot, isn't it?) and then run ltsp-build-client --arch=i386?
14:58
<SDuensin>
alkisg - That's what the docs say. I've not tried it.
14:59
It's on my list of things to do. :-)
14:59
<alkisg>
So there's really no good reason to choose the alternate CD in this case
14:59otavio has quit IRC
14:59
<SDuensin>
Dunno. I start with that and then am scripting all my customizations. Makes a nice base.
15:00
<alkisg>
Hm... yeah, I think there's an option to load an installation script from a disk or from the network somewhere in the installer... that could be handy.
15:01
<SDuensin>
Yea. The OEM install. One step at a time. :-) I'm still fighting with locking down the terminals.
15:01
<alkisg>
Did you look at pessulus?
15:02
<SDuensin>
That's the "Lockdown Tool" in Ubuntu, yea? It's part of what I'm using.
15:02
<alkisg>
Yup... I hope sbalneav gets sabayon working again
15:03
<SDuensin>
You and me both. I've run into a few things I can't figure out and hope that will help.
15:03
<alkisg>
...like?
15:03cliebow has quit IRC
15:03
<SDuensin>
The locked down user can still right-click the GNOME menu bar and select "Edit".
15:03
That's a big one.
15:04
<stgraber>
warren: hi :) I thought I answered it (by mail to ltsp-devel ...)
15:04
<alkisg>
But if you lock the menus from pessulus (I think) then the users can't modify anything in the panels nor in the menus
15:04
<SDuensin>
Maybe I'm not using pessulus.
15:05
<stgraber>
warren: argh, sent it from the wrong mail address ...
15:05
warren: sending again
15:05
<warren>
oh
15:05nobios has quit IRC
15:06
<SDuensin>
Oh yea, I used that before. I set the same stuff in GConf.
15:06
<warren>
http://en.wikipedia.org/wiki/How_does_one_patch_KDE2_under_FreeBSD%3F
15:06
whaaa?
15:07
<SDuensin>
Even with the panel and everything else I can find locked, I can still right-click and load the menu editor.
15:07
<moquist>
I've got an Ubuntu 8.10 LTSP server (OOTB, + updates) with a TC that fails to boot with an NBD_DISCONNECT error right after "Setting up LTSP client". Is that familiar to anyone?
15:07
<alkisg>
SDuensin: you're right, I just saw that happening with pessulus also
15:09pasmen has quit IRC
15:09
<SDuensin>
alkisg - That's f'ed up. :-)
15:09pscheie_ has quit IRC
15:09
<SDuensin>
moquist - Not me. Sorry.
15:12* SDuensin should write a book about setting this thing up.
15:12vagrantc_ has joined #ltsp
15:12vagrantc has quit IRC
15:12
<stgraber>
moquist: the reconnect is the expected behavior as I have to drop the initial nbd link and reconnect it with -persist
15:12
moquist: though the boot usually continues afterwards :)
15:12
<moquist>
got it. ran dist-upgrade in the chroot, ltsp-update-image, and now it booted.
15:13
different versions of NBD on TC and server?
15:13
<SDuensin>
I love how all the docs on locking the system down tell you how to do it...but not how to apply that to new or existing users.
15:13
<johnny>
SDuensin, dont bother until we fix sabayon..
15:14
<SDuensin>
johnny - I hope it's soon. Kinda need to build this thing to keep eating.
15:14
<stgraber>
moquist: maybe a kernel issue ? there was quite a lot of kernel update since intrepid was released
15:15
<johnny>
SDuensin, i kinda need a way to keep eating to build the thing
15:16
<SDuensin>
hehehe
15:16
<johnny>
i haven't been able to afford to give it the time it deserves
15:16
<SDuensin>
I looked at it, but it just blows up. :-(
15:16Gadi has left #ltsp
15:17
<SDuensin>
My GConf stuff can go in /etc/skel, but the crazy user name is in the menu configs. :-(
15:18
<moquist>
stgraber: right.
15:18* SDuensin predicts a bit of sed in his logon script soon.
15:20
<stgraber>
vagrantc_: pam_group works fine
15:20
vagrantc_: *;*;*;Al0000-2400;fuse,audio,video
15:20
vagrantc_: and you get everyone in fuse,audio and video
15:21
<vagrantc_>
stgraber: thanks!
15:22
<johnny>
SDuensin, you ready to fork over some loot? :)
15:22
<warren>
vagrantc_: would you prefer that we not require file?
15:22
<vagrantc_>
warren: i think it's by default on debian, so i'm ambivalent.
15:22
<SDuensin>
johnny - if I had it!
15:23
<warren>
oh
15:23
<vagrantc_>
warren: though i like a checksum... that really makes sure it's the same.
15:23
<warren>
vagrantc_: given that the output of file has a timestamp of creation on it, that really isn't a problem
15:27
<vagrantc_>
warren: well, just trying to give other reasons to not use file, since you don't seem to want it :)
15:27
overall, the NBD checks look kind of wonky to me...
15:28
<warren>
wonky indeed
15:30mpih has quit IRC
15:32CAN-o-SPAM has quit IRC
15:35pmatulis_ has joined #ltsp
15:35japerry has joined #ltsp
15:35mopey has quit IRC
15:36faustino333 has quit IRC
15:38markit has joined #ltsp
15:38faustino333 has joined #ltsp
15:39mopey has joined #ltsp
15:40warren has quit IRC
15:40RobertLaptop has joined #ltsp
15:42
<markit>
hi, once installed ltsp, to have X-Window (KDE) access to the server, do I have to enable it as usual, or is something ltsp take automatically care of, or is not needed at all?
15:42mikkel has quit IRC
15:43
<vagrantc_>
!release
15:43
<ltspbot>
vagrantc_: "release" is please mention the linux distro and release you're using :)
15:43
<vagrantc_>
markit: ^^
15:43evilx has quit IRC
15:45alkisg has quit IRC
15:45faustino3331 has joined #ltsp
15:45johnny has left #ltsp
15:45F-GT has quit IRC
15:46Patina has quit IRC
15:47Patina has joined #ltsp
15:48faustino333 has quit IRC
15:48F-GT has joined #ltsp
15:48
<markit>
debian sid 64, ltsp ltsp-server 5.1.47-1
15:48
vagrantc_: sorry :)
15:48BrunoXLambert has quit IRC
15:49japerry_cat has quit IRC
15:51vagrantc_ has quit IRC
15:52johnny has joined #ltsp
15:54gate_keeper_ has quit IRC
16:03
<markit>
mmm /etc/exports does not exist at all... what package should I add? I've nfs-common installed, seems there is not a nfs-server, neither in the repositories
16:06pmatulis_ has quit IRC
16:07sepski has quit IRC
16:20
<Lns>
Can anyone tell me why my chroot time (not date, just time) is way off and random on different TCs?
16:20
Gnome/server time is correct, BIOS time on TCs are correct.
16:22
It's preventing me from being able to execute cron scripts in the chroot at the right time :(
16:23
<johnny>
is it just way off.. or random?
16:23
i doubt it can be both
16:24
Lns, maybe add ntp-client to boot
16:24
or whatever runs ntpdate
16:24
<Lns>
johnny: but do you know why it would be off in the first place?
16:24
johnny: it's pretty off..it's ~2:30pm here now, and my chroot says it's 10:22pm
16:25
the other TC says it's 6:00am
16:25
<johnny>
utc?
16:25
oh
16:25
weird
16:25
<Lns>
yeah
16:25
<johnny>
i'd suggest ntpdate
16:25
<Lns>
:( I'd rather not have to roll new images if i can fix it otherwise :(
16:26
<johnny>
it should possibly be there by default.
16:26
on anything that runs ANYTHING locally
16:26
<Lns>
nptdate is there actually
16:27
<johnny>
does it run?
16:27
<Lns>
i would have to assume no... ;)
16:27
lemme verify
16:27
not in init.d
16:27
<markit>
anyone can help about how to configure a dnsmasq that is on a different host than ltsp server?
16:30
<johnny>
Lns, it's probably a package
16:30
markit, there's a file in the examples/ directory of the documentation of ltsp-server
16:31
<Lns>
johnny: hrm, /etc/network/if-up.d/ntpdate exists in chroot..
16:31jammcq has quit IRC
16:32
<johnny>
Lns, but i don't think that runs
16:32
i could be wrong tho
16:33
<markit>
johnny: I've seen it, and I'm coping it's config, but I've the feeling that is ok only in case you have dhcp server AND ltsp server on the same host,or am I wrong?
16:33
<johnny>
you can try by logging itno the chroot and running it manually
16:33
markit, i don't recall.. i'd have to see it again :) lemme check
16:33
<Lns>
johnny: yeah - i'm digging deeper, /etc/default/ntpdate exists, and uses NTPDATE_USE_NTP_CONF=yes - but there's no /etc/ntp.conf (which that var requires)
16:33
that might be it
16:35
<johnny>
markit,
16:36
dhcp-boot=net:pxe,ltsp/x86/pxelinux.0,beep,192.168.2.4
16:36
dhcp-option=17,"192.168.2.4:/opt/ltsp/x86"
16:36
that's what i have
16:36
where beep is my ltsp server, and so are the ips
16:36vvinet has quit IRC
16:37
<markit>
johnny: oh "beep" is the hostname? ok
16:37
with 192.168.2.4 as ip?
16:37
<johnny>
whatever your server ip is
16:38
<Lns>
ugh..this is weird. /etc/default/ntpdate says in comments it's for ntpdate-debian, not upstream nptdate package...but no ntpdate-debian package exists in chroot
16:38
<markit>
sure, but you have a ltsp server called "beep" with 192.168.2.4 ip, right?
16:38
<johnny>
yes
16:38
does it exist at all ?
16:39
<Lns>
oh, it looks like ntpdate package contains ntpdate-debian as well
16:39
<markit>
johnny: and I skip enable-tftp etc, since I will have to install a tftp server on ltsp server, right? any suggestion to what use?
16:39
<johnny>
tftp-hpa is probably fine
16:40Egyptian[Home] has quit IRC
16:40
<markit>
johnny: thanks a lot :)
16:41Egyptian[Home] has joined #ltsp
16:45japerry_cat has joined #ltsp
16:48
<markit>
johnny: in the sample config, dhcp-boot argument starts with "/" (/ltsp...) while your does not
16:48
<johnny>
you'll find it
16:48
out*
16:48
try both
16:48
i just got something working and didn't touch it anymore.. but it was in the middle of 3 or 4 broken things at the time
16:48
<markit>
oh :)
16:48
<johnny>
so.. you'll find out what works for you
16:49
<markit>
ok, thanks, I've version 5, let's cross fingers, lol
16:51faustino333 has joined #ltsp
16:51F-GT has quit IRC
16:52faustino3331 has quit IRC
16:54F-GT has joined #ltsp
16:55* Lns thinks he's going mad
16:55
<Lns>
my chroot sources.list file has reverted to the default 2-line repository....
17:02pem725 has joined #ltsp
17:03japerry has quit IRC
17:03
<pem725>
anyone have luck with nvidia twinview on thin clients?
17:04mathesis has joined #ltsp
17:05
<johnny>
pem725, and you have the nvidia-drivers installed in the client.. and server?
17:07
<pem725>
well I have the nvidia-drivers installed in the chroot
17:07
not on the server per se though.
17:07
<markit>
mm I issued ltsp-build-client --arch i386 --dist sid then stopped some seconds after start. Now I removed /opt/ltsp/i386 and issued the same command again, but does NOTHING, any clue?
17:08
(I removed also the log file from /var/log)
17:08
<pem725>
johnny - any hints?
17:09
<johnny>
it needs to be on the server
17:09
tha'ts the only hint i have
17:09
i've never done it before tho
17:09
<markit>
ok, works now (don't know why)
17:09
<pem725>
ok, thanks.
17:10
<markit>
oh, stupid me, I pasted the "#" in front of the command :)
17:11vagrantc has joined #ltsp
17:11
<markit>
I've the bad feeling that I will not be able to make it work in a reasonable time... I would like to demo it to a friend of mine tomorrow morning, and is midnight right now, and 1 hour since I started
17:11
:)
17:15
<johnny>
didn't you already build a chroot?
17:17otavio_ has quit IRC
17:17otavio has joined #ltsp
17:19
<markit>
johnny: is downloading packages right now
17:19
I finished building, then realized that I did a amd64 install, so dropped
17:20
then figure out what packages are needed, set dnsmasq, installed packages
17:20
and restarted client-cheroor-building
17:20
ehm, tons of mistypes :)
17:27
cryptsetup: WARNING: could not determine root device from /etc/fstab
17:27
hope is not a problem
17:28Lns has quit IRC
17:33
<vagrantc>
markit: shouldn't be.
17:33
<markit>
ok, I've tftp running also, let's do the first test with a client
17:33
oh :(
17:34
<vagrantc>
markit: the cryptsetup warning shouldn't be a prolbme, that is.
17:34
<markit>
oh, I misunderstood your answer.. maybe I'm too tired
17:35
do I have to set X to let access from thin clients, or ltsp works differently?
17:35
<johnny>
no
17:35
it's all ssh tunneled by default
17:36
altho you might have to set to allow x11forwarding in sshd config.. but i'm not sure of debian's default config
17:37
<vagrantc>
debian allows X11Forwading by default
17:37
since etch, at least...
17:38
<markit>
sshd_config:X11Forwarding yes
17:38
correct :) thanks
17:41
it's booting the kernel! :)))
17:41
nfsmount need a path :(
17:41
rootpath 192.168.1.30/don'tremember
17:41
maybe the lines of dnsmasq we were talking before
17:41
and the "/" question
17:42
<vagrantc>
stgraber: pam_group seems to work fine for me on debian etch, but it's not working for me with hardy ...
17:42pem725 has left #ltsp
17:43
<vagrantc>
stgraber: ah, i don't seem to have fuse stuff installed...
17:43bobby_C has quit IRC
17:43
<stgraber>
vagrantc: IIRC pam_group isn't enabled by default in Hardy, you have to add it in /etc/pam.d/common-* by hand
17:45
<markit>
mmm does not change, maybe is another parameter that matters
17:46
<johnny>
maybe you need the /
17:46
oh.. you tried
17:46
<markit>
dhcp-option=17 I missed a ":", sigh
17:47RobertLaptop has quit IRC
17:48
<markit>
ltsp login!!!!
17:48
johnny: thanks for your help :)
17:49
and thanks to the community for ltsp and debian :)
17:49
<vagrantc>
stgraber: yeah, i didn't have ltspfs installed on the hardy server :)
17:49* vagrantc hopes it's compatible with the version in debian experimental
17:49
<markit>
mmmm I fear for the health of my lcd monitor... image seems really broken
17:51
many vertical trumbling bars at a certain distance each other
17:52
wondering how can I improve recognition ability without linking the thin client with mac address and a config file
17:54
<vagrantc>
stgraber: doesn't appear to work as well with more current ltspfs ...
17:54
this will be strange. me, of all people, maintaining a backport of ltspfs for ubuntu hardy ...
17:54
<stgraber>
vagrantc: does fuse appear when you do "groups" ?
17:54
<vagrantc>
stgraber: yeah, fuse group is kicking in...
17:55
stgraber: there may be enough difference in ltspfs from hardy and current ltspfs that it might just not work...
17:56
<stgraber>
ok, what happens if you try to execute ltspfsmounter all cleanup
17:58
<vagrantc>
stgraber: aha!
17:58
stgraber: /dev/fuse is not writeable
17:59
stgraber: i reported this bug on debian ...
17:59
<stgraber>
ok, had a similar issue with Intrepid in OpenVZ this afternoon :) that's how I found it
17:59
<vagrantc>
stgraber: /dev/fuse isn't writeable on initial package install
17:59
<stgraber>
it should be 660 root.fuse
17:59
<vagrantc>
because udev is started before the fuse group exists
17:59
<stgraber>
stgraber@sahal:~/code/ldm-trunk$ ls -lh /dev/fuse
18:00
crw-rw----+ 1 root fuse 10, 229 2009-01-14 16:23 /dev/fuse
18:00
that's in default Intrepid (on my laptop) so the bug has been fixed in Intrepid but the fix hasn't been backported ot hardy it seems
18:01
<vagrantc>
restarting the server fixed it on debian
18:01
hopefully rebooting it will work, too
18:01* vagrantc wonders if the ltspfs_token stuff is in hardy
18:02
<vagrantc>
seems to be working :)
18:02
<stgraber>
ltspfs in hardy is 0.5.0
18:03
<vagrantc>
~bzr2008...
18:03
<stgraber>
yeah, that's basically a one year old snapshot :)
18:06
<vagrantc>
i think i'm going to disable floppy disks...
18:06
<stgraber>
bzr st -r date:2008-01-09
18:06
in ltspfs-trunk :)
18:06
ltspfs was basically rewritten in a year :)
18:06
<vagrantc>
not totally re-written... still seems at least somewhat compatible.
18:07litlebuda has quit IRC
18:09
<markit>
after changing lts.conf, do I have to reboot something apart the client ?
18:10Lns has joined #ltsp
18:10
<Lns>
johnny: what were you saying before re: UTC time in the chroot? This *might* be my issue
18:10
it's weird, because I'm GMT+7 - but one client shows about 7 hours ahead, the other 7 hours behind.
18:11
markit: no
18:11
<johnny>
well you should have the time of the bios in the client to be in utc
18:11
not local time
18:11
<Lns>
I see a ton of things regarding setting the system clock in the chroot, so i don't want to band-aid it with some cronjob and ntpdate if the code is there already
18:12
<johnny>
but.. if you use ntp, it matters less as long as the clock time in the chroot is set to localtime
18:13* Ryan52 wonders what warren wanted..
18:14
<Lns>
johnny: well i think i want to steer clear of ntp because the init scripts for setting the time are already there. maybe setting the bios time is the best bet
18:14
<johnny>
i disagree
18:15
because... you never have to do it again
18:15
on any machine
18:15
<Lns>
true
18:16
It almost seems that Ubuntu 8.04 chroots set up ntpdate halfway
18:16
<johnny>
you might also want to setup an ntpd on the server, so you can just query locally
18:17
<Lns>
yeah - i would almost think this should be a default, to keep the times in sync..though im not sure what purpose that would serve apart from showing the correct time in ldm
18:17
and my special case of course
18:17
<johnny>
which is less and less of a special case with local apps
18:17
<Lns>
very true
18:17
<johnny>
and/or fatclients in general
18:18
<Lns>
So maybe "we" (us all) should look into default ntp synchronization in ltsp
18:18
unless it's already happened in intrepid/jaunty/other dists
18:20
<vagrantc>
there's already code to use ntpdate in ltsp
18:20
<Lns>
vagrantc: then why aren't my TCs using it? :(
18:21
<vagrantc>
Lns: it's probably not enabled by default
18:21
<Lns>
aaaah
18:21
<vagrantc>
set TIMESERVER and make sure ntpdate is installed in the chroot
18:21
and set up ntpd on the server
18:22
although it was added fairly recently... not sure if the version you're running has it or not
18:23
<Lns>
vagrantc: where's TIMESERVER set? docs don't have it ..
18:23
just an environment var?
18:24
ntpdate is indeed in my chroot by default
18:24
<vagrantc>
lts.conf, like everything else.
18:25
<Lns>
hrm, the upstream docs don't have that var
18:25
<vagrantc>
the first version in debian that had it was 5.1.31-1
18:25
er...
18:25
5.1.34 ...
18:26
definitely in 5.1.34, not in 5.1.31 or earlier ... not sure about 5.1.32 or 5.1.33
18:26johnny has left #ltsp
18:26
<Lns>
8.04 is at 5.0.40 :(
18:26johnny has joined #ltsp
18:27
<markit>
what is mac address syntax in lts.conf? ":" as separator? all uppercase or lowercase ?
18:27johnny is now known as Guest21536
18:27
<vagrantc>
Lns: it wouldn't be hard to add in... it's pretty much a one-line patch.
18:28
<Lns>
vagrantc: can you tell me where and what to change possibly?
18:29
<vagrantc>
Lns: add an RCFILE_NN with: ntpdate $ip_of_server
18:29
<Lns>
vagrantc: oh well that's what i was gonna do anyway =p
18:29
<vagrantc>
if you want to make it forwards-compatible with upstream:
18:29
[ -n "$TIMESERVER" ] && ntpdate $TIMESERVER &
18:29
and set TIMESERVER in lts.conf
18:30
<Lns>
vagrantc: awesome, thank you =)
18:30* Lns is learning!
18:36RobertLaptop has joined #ltsp
18:38
<Lns>
vagrantc: generic bash q: the -n is "if the variable exists" right? Is the && a cleaner way than using if/else since it only executes if true?
18:40
<vagrantc>
if [ -n "$TIMESERVER" ]; then
18:40
ntpdate $TIMESERVER &
18:40
fi
18:40
Lns: is basically equivalent
18:40
<Lns>
vagrantc: cool, that's what i thought
18:41
i'm sort of putting all the pieces together from what i've learned in the past. I need to pick up a good book/url on bash scripting. I always just google when i'm trying to figure something specific out.
18:42
<markit>
to my question: uppercase is needed
18:53Guest21536 is now known as johnny
18:54
<markit>
when I move a window in the thin client, I see ALL the lights of my switch blink! maybe there is a problem with IP?
18:54
is something that sounds familiar with one of you?
18:54* markit tries wireshark but is not an expert
18:54
<Lns>
markit: are you using a hub??
18:55
<markit>
no, otherwise would be normal...is a 1Gbit 16 port switch
18:55
but you have got the strange synthom I see
18:55
is like it were an hub
18:56
<Lns>
markit: is it a managed switch? anyone messing with the config?
18:56
I see 2 lights blink when i move windows :)
18:57
<markit>
wireshark does not show broadcast
18:57
I'm confused, but sure there is something wrong
18:58
<Lns>
i dunno...try power cycling the switch maybe
18:58
<markit>
do you have encrypted traffic also, or you disabled that?
19:00
<ltsppbot>
"Lns" pasted "Horribly written directions on automated thin-client shutdown at a specified time" (41 lines) at http://pastebot.ltsp.org/210
19:00staffencasa has quit IRC
19:00
<Lns>
vagrantc: does that look sane to you ? ^^^
19:01
I need to take off but i'll write more descriptive directions for inclusion in docs if you want
19:01
<markit>
mm having 2 interfaces plugged in the same switch seems get traffic on both from the thin client
19:01
<vagrantc>
Lns: other than the restarting cron on server ...
19:02
<Lns>
vagrantc: right..well at least it's not my own insanity there ;)
19:02
<vagrantc>
Lns: wait, why are you editing /opt/ltsp/i386/etc/init.d/ntpdate ...
19:02
<markit>
even if one is on a different subnet
19:02
<Lns>
vagrantc: for inclusion in the RCFILE_01 script
19:02
err...line
19:03
<vagrantc>
Lns: /etc/init.d/ scripts have their own conventions ... put it somewhere else... /etc/ltsp
19:03
<Lns>
vagrantc: ok gotcha.
19:04
i was always used to putting RCFILE_nn in chroot /etc/init.d because that's what the docs tell me to do =)
19:06
<markit>
now works fine, but can't understand why second adapter plugged gave that problems
19:08
<Lns>
ugh. I've had enough of today. I'm going home. =)
19:08
Cheers all
19:09Lns has quit IRC
19:09otavio has quit IRC
19:15otavio has joined #ltsp
19:23twinprism_ has quit IRC
19:29nobios has joined #ltsp
19:38vagrantc has quit IRC
19:46johnny has left #ltsp
19:49MeW_ has joined #ltsp
19:51tinpardo has joined #ltsp
19:53tinpardo is now known as help
19:53jammcq has joined #ltsp
19:53help is now known as Guest2028
19:54Guest2028 is now known as tinpardo
19:56tfl_totalnub has joined #ltsp
20:02try2free has joined #ltsp
20:04dirigeant has quit IRC
20:07johnny has joined #ltsp
20:09
<try2free>
all my client default resolution is only 640x480, how to set to 800x600. because same client pc to another ltsp server get 800x600 as default. it's because setting memory vga too low in bios?
20:12
all my client default resolution is only 640x480, how to set to 1024x768. because same client pc to another ltsp server get 1024x768 as default. it's because setting memory vga too low in bios?
20:17
<tinpardo>
but if i install flash players it crash
20:17
kernel panic
20:18
system freeze
20:30markit has quit IRC
21:17ccherrett has quit IRC
21:24ccherrett has joined #ltsp
21:28warren has joined #ltsp
21:41
<cyberorg>
warren, nomad uses modified rdesktop only for composite extension, it works fine using un-modified rdp clients too, you will not get wobbly windows that is all
21:41
<Ryan52>
warren, pong (delayed)
21:42
<warren>
cyberorg: the modified rdesktop is the problem, the entirely separate X server is.
21:43
cyberorg: is it simple to build that separate X server as its own RPM and parallel install it with whatever you have?
21:43
Ryan52: the kvm folks ask that you please submit the patch upstream
21:43
Ryan52: let me know if you will not do so
21:43
<cyberorg>
warren, modified rdesktop is not mandatory, you dont get compiz taking advantage of local graphics hardware
21:43
warren, there is no modifed X server either
21:44
<warren>
cyberorg: what acts as the X proxy server?
21:44
<Ryan52>
warren, I noticed, but as was later commented, it can be done a better way.
21:44
<cyberorg>
it is xorg-x11-xrdp and xdmx
21:44
again only modifications are support for composite extensions
21:44
<warren>
cyberorg: what exactly is that? is it going to upstream xorg?
21:45
<jammcq>
xdmx was written by Kevin Martin, a redhat guy
21:45
<cyberorg>
http://xrdp.sourceforge.net/
21:45
warren, Xdmx provides funky things like DnD/copy/paste between multiple X sessions
21:46
<jammcq>
xdmx allows you to treat a bunch of separate machines as one big display
21:46
way cool stuff
21:46
<warren>
cyberorg: I mean xrdp
21:47
cyberorg: the earlier nomad pages I read made it seem that xrdp is its own X server
21:47
<Ryan52>
warren, btw, what is up with you @redhat.com people? :). I submit the patch, no comment for 9 days, then all of the sudden 4 people reply within a half hour...
21:47
<mistik1>
Xdmx is awesome
21:47
<cyberorg>
warren, it starts Xfake
21:47
<warren>
cyberorg: our own X guys said "you may not add this to fedora because we will not maintain yet another X server"
21:47
<cyberorg>
warren, no, redhat guys just wait for suse devs to perfect something then include it :)
21:47
<warren>
Ryan52: I'm not sure, I personally didn't see that bug until today, but it was irritating the hell out of me for months.
21:48
cyberorg: what are other examples of this?
21:48
<cyberorg>
Xen
21:48
<warren>
cyberorg: that is utter bullshit.
21:48
We did far more work on Xen than Novell.
21:48
<cyberorg>
warren, just kidding you
21:49
ah, but you guys kept on saying it was not ready till it was included in SLES 10, then promptly added it to RHEL5 :D
21:49alekibango has quit IRC
21:49
<cyberorg>
some more background about nomad http://lists.freedesktop.org/archives/compiz/2009-January/003264.html
21:49
reasons why it is more efficient way of doing things
21:50
<warren>
cyberorg: RHEL5's dev schedule was just later than SLES
21:51
cyberorg: you are not answering the question, is xrdp yet another X server? according to my people it is.
21:53
<cyberorg>
warren, i dont know the details, that is why the link to the home page :P
21:53hanthana has joined #ltsp
21:54
<cyberorg>
warren, can the src package of xserver be 380K?
21:55
http://download.opensuse.org/repositories/home:/hfiguiere:/Factory/openSUSE_Factory/src/
21:55
hmm, its 17M
21:55
xorg-x11-server-rdp-7.3.99-3.24.src.rpm 20-Jan-2009 20:22 17M
21:55
xorg-x11-server-dmx-7.3.99-22.12.src.rpm too
21:59
<warren>
cyberorg: yes, this is a fork of xserver itself
21:59
and an old pre-release version
21:59
cyberorg: will be interesting to see if they modernize it to 7.6 and upstream it
22:01
<cyberorg>
warren, i sure hope so, if it is as ugly as you think it is, no one would want to maintain it
22:01CaScAdE^1arAway has joined #ltsp
22:01
<warren>
well, it depends on how much resources Novell has behind it
22:01japerry_cat has quit IRC
22:01
<warren>
cyberorg: folks like NX have been forever maintaining a very ancient fork of X
22:02
<cyberorg>
from the conference call, they gave us impression that it is the next big thing
22:02
<warren>
It may well be, but long-term maintainability is a big deal.
22:02
upstreaming it will make it easier to maintain
22:02
<cyberorg>
they have to maintain for 7 years at the least, due to SLE life time
22:03
<warren>
ok
22:06* johnny feeds warren another chill pill
22:07
<warren>
johnny: I was only stating the truth.
22:07
<johnny>
often.. you only tell the truth in anger.
22:07
<warren>
I asked my people, and they have no interest in using nomad.
22:08
<johnny>
i do hope the nx issue gets resolved some day
22:08
<Ryan52>
what's this nomad thing everybody's talking about?
22:09
<johnny>
Ryan52, it was on ltsp-developer
22:09
<Ryan52>
o, i c.
22:10
<warren>
johnny: what nx issue?
22:10
<johnny>
the ancient x fork you are speaking of
22:10
<warren>
Ryan52: looks better than nx
22:10
well, anything looks better than nx
22:11
<johnny>
warren, are your people working on something like launchpad for rpms?
22:11
<warren>
johnny: yes.
22:11
<Ryan52>
heh
22:11
<johnny>
or.. are you just gonna add rpm support to launchpad
22:12
working on launchpad would be fun..
22:12
<warren>
johnny: although we already do most of launchpad's functions just scattered in 5 places in a confusing way.
22:12
johnny: we're not working on launchpad code itself. it isn't open yet, and it is bzr-only centric.
22:12
<johnny>
i know you aren't
22:12
i was just kidding around
22:12
altho.. it could happen..
22:12
once it is open
22:13japerry has joined #ltsp
22:13* cyberorg gotta go give lecture
22:13
<johnny>
me.. i think bzr is slightly better than git in the way it can be extended.. and the way metadata is stored
22:13
<warren>
the way metadata is stored
22:14
sure, they keep redoing the metadata format
22:14
<johnny>
not saying bzr is perfect
22:14
git is fast
22:14
<warren>
hahah, icantbelieveitsnotbtr
22:14
<johnny>
monotone is still my favorite..
22:14
so i'm no bzr fanboy or anything..
22:14
<warren>
you need to put this on the kernel cmdline for our installer to allow you to format btrfs
22:14
<johnny>
hehe :)
22:15
i'll be happy enough with ext4
22:15
i'm contemplating doing it on this fedora10 box
22:15
the kernel is new enough
22:15
i liked the way mtn handles arbitrary certificates.. and the fact that it uses sqlite..
22:16
certficiates == metadata
22:16
and the signing is enforced
22:17
altho.. who knows what we'll find in this new era of schema-less databases..
22:18* johnny looks for nubae
22:20
<johnny>
warren, iirc fedora 11 will use ext4
22:20
<warren>
yes, and btrfs is available for testing if you are crazy enough
22:20
it still can't tell you if it runs out of space
22:20
it just panics
22:20
<johnny>
lol
22:20
like frozen kernel?
22:21
<warren>
I think so
22:21CaScAdE^FarAway has quit IRC
22:37tinpardo has quit IRC
22:37Egyptian[Home] has quit IRC
22:39Egyptian[Home] has joined #ltsp
23:07tfl_totalnub has left #ltsp
23:10alkisg has joined #ltsp
23:17japerry has quit IRC
23:17japerry has joined #ltsp
23:19
<alkisg>
Ryan52: hi, did you see the backlogs? All ok, break fixed it.
23:20
<Ryan52>
yep
23:20
<alkisg>
Thanks! :)
23:20
<Ryan52>
yw :)
23:21* Ryan52 also saw the commit, which makes the "error" messages even more confusing :)
23:22
<Ryan52>
sigwaitinfo errors, waitid errors, and everything is working like it should be...that's gonna confuse somebody.
23:22
<alkisg>
Yeah I told sbalneav to remove the first error message but I guess he was busy and didn't hear me... :(
23:23
<Ryan52>
even the second error message shouldn't be shown if errno is equal to whichever "error" it is that you were getting.
23:23
do you know off the top of your head which error it was? (if not, I can check backlog)
23:23
<alkisg>
Yeah, no such process or something
23:23
Because ssh died before waiting for it to die...
23:23
<Ryan52>
ESRCH No such process (POSIX.1)
23:23
makes sense.
23:24
<alkisg>
sbalneav just wanted to add _() for internationalization
23:24
<Ryan52>
no, it should be this one: ECHILD No child processes (POSIX.1)
23:24
ya.
23:25
my patch wasn't upstream ready.
23:25
(or I would have pushed it myself)
23:26
<alkisg>
Well, when I pinged sbalneav I was looking for a bug talk, but instead he just commited the patch :)
23:26
Anyway, please do fix it properly if you have the time
23:32alkisg has quit IRC
23:41
<kc8pxy>
johnny: you around?
23:41
<johnny>
hi
23:41
my box might be alive again..
23:41
almost..
23:41
<kc8pxy>
you get my message?
23:41
<johnny>
no
23:42* johnny hopes the machine comes up again
23:42
<johnny>
then i'm going to build a new image
23:42
<kc8pxy>
congrats on mostly getting your box back
23:42
<johnny>
i had so many updates to do
23:42
i still have about 12 packages that don't install
23:42
but they aren't that important
23:42
<kc8pxy>
IMHO< openrc should go in the ldm deps.
23:43
<johnny>
ldm deps??
23:43
more like ltsp-client
23:43
<kc8pxy>
ok
23:43
something in there
23:43
<johnny>
i was hoping not to have to add deps at all.. and that openrc would go stable
23:43
but that process is taking longer than i expected
23:43
i was under the assumption that it was going to happen a few months ago
23:43
<kc8pxy>
but that, and udev and sysvinit need to be keyworded.
23:44OnePutt has joined #ltsp
23:44
<johnny>
kc8pxy, i'm going to just fix it in quickstart
23:44
until it goes stable
23:44
<OnePutt>
hi all
23:44
<kc8pxy>
kk
23:44
<johnny>
instead off adding it to deps
23:44
uggh.. one more problem i see..
23:44
all i see are squares :)
23:44
no text
23:44
<kc8pxy>
on your test gentoo?
23:47try2free has left #ltsp
23:47
<OnePutt>
my client after pxe boot drops out of X to a screen talking about Disabled Privacy Extensions then goes back into X windows repeats a few times then reboots the thin client. I just started with this a couple hours ago so sorry if this is n00b mistake
23:48
<kc8pxy>
OnePutt: distro of server?
23:48
<OnePutt>
server
23:48
pxe boot
23:48
dhcp
23:49
client is a wyse s10
23:49
<kc8pxy>
OnePutt: what distro is the server running?
23:49
<OnePutt>
x64
23:49
latest
23:49
<kc8pxy>
what linux?
23:49
<OnePutt>
sorry
23:49
k12linux
23:50
<kc8pxy>
OnePutt: ugh.. i have barely heard of it, let alone touched it.
23:50
sorry.
23:50
<OnePutt>
iK12Linux F10 Stable 2 x86-64 '
23:50
oh
23:50
well i just grabbed one to start with
23:50
whats reccomended?
23:50
<kc8pxy>
i can try, but i'm not sure how much help i can be.
23:50
<OnePutt>
server currently is a amd x64
23:51
<kc8pxy>
OnePutt: are you new to linux?
23:51
<OnePutt>
well ive been away for a while
23:51
<kc8pxy>
kk
23:51
welcome back :)
23:51
<OnePutt>
ive only used linux for web hosting
23:51
and just got into the apple world 3 years ago
23:51
unix
23:52
<kc8pxy>
OnePutt: good application.(of linux)
23:52
<OnePutt>
back into linux i go
23:52
<kc8pxy>
hmmm.
23:52
why k12?
23:52
<OnePutt>
no reason
23:52
i just picked it
23:52
<kc8pxy>
just because it was simple to get?
23:52
kk
23:52
<OnePutt>
saw thay it had a x64 iso
23:53
and it said "fastest way to pxe booting"
23:53
something like that
23:53
<kc8pxy>
hehe
23:54
<OnePutt>
ive just started using debian a couple weeks ago
23:54
<kc8pxy>
OnePutt: i would say , if you are using it primarily for just ltsp, you probably want (ed)ubuntu or gentoo. me and johnny are hammering out the details nicely :) )
23:54
<OnePutt>
should i install onto a debian box?
23:55
<kc8pxy>
johnny does the fixing, and i do the head-vs-wall beating, trying to get it working :)
23:55
<johnny>
OnePutt, they are all nice
23:55
use whatever you are most familiar with
23:55
<OnePutt>
suse
23:55
heh
23:55
<johnny>
the major distros except madriva and slackware have good support
23:55
<kc8pxy>
OnePutt: what did you use before you linux drought?
23:55
<johnny>
OnePutt, suse is fine too
23:55
<OnePutt>
redhat
23:56
slakware
23:56
mitel e-smith server
23:56
<johnny>
i'm currently on a fedora laptop.. but i operate regularly with gentoo and ubuntu boxen
23:56
<OnePutt>
kde
23:56
<johnny>
they are all nice
23:56dpepin_ has joined #ltsp
23:57* kc8pxy has been converting a decent number of people to gentoo recently :)
23:57
<OnePutt>
well whats a good distro for amd x64?
23:58
<kc8pxy>
OnePutt: IMNSHO, gentoo :)
23:58
<OnePutt>
k
23:58
gentoo it is
23:58
<johnny>
OnePutt, they are all fine.. but..
23:58
<OnePutt>
my friend swears by it
23:58
<kc8pxy>
OnePutt: your friend swears by gentoo?
23:58
<johnny>
it's only the non open source apps you'll have trouble with.. on any distro
23:58
everything else is fine with 64bit
23:58
<OnePutt>
he says "you compile everything specific to your hardware!, its the shizzz"
23:58
<kc8pxy>
johnny: yeah, like flash :)
23:59
<johnny>
luckily flash finally put out a 64bit
23:59
<OnePutt>
yea
23:59
<kc8pxy>
johnny: when??
23:59
<johnny>
it's still "alpha" works fine here tho
23:59
kc8pxy, it's masked in gentoo
23:59
<dpepin_>
ubuntu intrepid works well for me
23:59
<johnny>
but give it a shot
23:59
no need for nspluginwrapper
23:59
<kc8pxy>
johnny: hardmasked?
23:59
<johnny>
now we just need to get skype