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


Channel log from 22 February 2008   (all times are UTC)

00:49F-GT has quit IRC
01:02F-GT has joined #ltsp
01:07indradg_ has joined #ltsp
01:07indradg has quit IRC
01:48ufo_ has quit IRC
01:48ufo_ has joined #ltsp
01:50Q-FUNK has joined #ltsp
01:58mikkel has joined #ltsp
02:29Q-FUNK has quit IRC
02:33Q-FUNK has joined #ltsp
02:35
<vlt>
Hyvää Huomenta and hello. Maybe there's an easier way (even on Ubuntu 6.06's LTSP) but I don't know it, so I tried the following: I installed a SIP softphone (twinkle) into the chroot and tried to access it via `ssh -X cient.ip twinkle`. But I can't get any ssh -X app working (not even xeyes). When I connect, $DISPLAY is empty. Any idea where to look first?
02:46
<Q-FUNK>
no, terve
02:46
which version of LTSP and which distro?
02:51
<vlt>
Q-FUNK: Ubuntu Dapper 6.06's LTSP(5) packages. Maybe o_gra is the right man here ...
02:52
<Q-FUNK>
dapper is getting rather old
02:52
did you update the boot image after you installed?
03:00
<vlt>
Q-FUNK: Yes, I know. I've heard that more than once in the last days and already have decided to upgrade soon. But sometimes you need something working right now. -- No I didn't update the chroot. I installed some apps like ssh server, cups and sane, and twinkle right now. But the repo in sources.list is still only "dapper" (not "dapper-updates" for example).
03:01
<Q-FUNK>
just instaling isn't enough
03:01
sudo ltsp-update-image
03:01
right after
03:11bobby_C has joined #ltsp
04:00Egyptian[Home] has quit IRC
04:05
<vlt>
Q-FUNK: I got no image here. It's good ol' NFS mount style.
04:06
Q-FUNK: I just installed aespipe (for something completely different ;) and it instantly worked.
04:07
<Q-FUNK>
ok :)
04:07Egyptian[Home] has joined #ltsp
04:11ogra_cmpc has quit IRC
04:32elisboa has joined #ltsp
04:34elisboa has joined #ltsp
04:38bobby_C has quit IRC
04:39bobby_C has joined #ltsp
04:53ogra_cmpc has joined #ltsp
05:12ogra has quit IRC
05:18ogra has joined #ltsp
05:30
<Q-FUNK>
ogra: btw, please let me know when you're ready to flash your 61
05:35K_O-Gnom has joined #ltsp
06:05Guaraldo has joined #ltsp
06:06Guaraldo has left #ltsp
06:39sonjag has joined #ltsp
06:41
<sonjag>
Hello everyone! Can anyone tell me or direct me to instructions on how to restrict the use of the terminal from users other than administrators? Thanks!
06:46
<vlt>
sonjag: The terminal itself has a passwd protected root account. The users log in on the server, so there you have to care for permissions ...
06:50
<sonjag>
vlt, not so on my install. If a user clicks alt-F2, they get the run application menu, and then they can choose terminal from there. No password required.
06:55plamengr has joined #ltsp
06:57mccann has joined #ltsp
07:00
<vlt>
sonjag: If the user can press F2 he is already logged in.
07:00
sonjag: So, obviously no password needed then.
07:00jammcq has quit IRC
07:00
<vlt>
Alt+F2, even
07:02cliebow_ has joined #ltsp
07:05mhterres has joined #ltsp
07:13
<sonjag>
vlt, I'm not sure what you mean. Any user on my LTSP setup is currently able to use terminal without any root or other credentials. I want to restrict that to only allowing admin users to use terminal. I believe there is an executable that I can set permissions on... I'm looking for something like that.
07:14
<vlt>
sonjag: Aaah, I completey misunderstood. terminal = `terminal`. Now I got it ...
07:16
<sonjag>
vlt, ahhh, I see your confusion.... I'm trying to thwart my little hackers that found out about using the command line...
07:23
<cliebow_>
heh
07:29jammcq has joined #ltsp
07:30
<sonjag>
okay, I found the executable (gnome-terminal) and I can see the permissions. -rwxr-xr-x . The user and group root. If I change the group to admin, then change the "other" bits to off (-rwxr-x---) will that fix it so that only admin users can run terminal? Anything dangerous about doing this? Does this make sense?
07:41
<sep>
sonjag, you should encourage command line useage. not try to hide it from them. using the command line is both educational and inspiering.
07:42
if your system is insecure, hiding one of the command line terminal emulator programs is not going to make it secure
08:10otavio has quit IRC
08:12otavio has joined #ltsp
08:18sgonzalez has joined #ltsp
08:20sonjag has quit IRC
08:23artista_frustrad has joined #ltsp
09:03otavio has quit IRC
09:04otavio has joined #ltsp
09:04sonjagonzalez has joined #ltsp
09:06sgonzalez has quit IRC
09:22sonjag has joined #ltsp
09:22sonjag has joined #ltsp
09:34highvoltage has quit IRC
09:40highvoltage has joined #ltsp
09:42highvoltage has quit IRC
09:43bobby_C has quit IRC
09:50highvoltage has joined #ltsp
09:54vagrantc has joined #ltsp
09:55highvoltage has quit IRC
10:02highvoltage has joined #ltsp
10:14sgonzalez has joined #ltsp
10:15sonjagonz has joined #ltsp
10:16sonjagonzalez has quit IRC
10:19sonjag has quit IRC
10:20Q-FUNK has quit IRC
10:31staffencasa has joined #ltsp
10:51alumno10 has joined #ltsp
10:51
<alumno10>
hi all!
10:51
how can i change the default boot kernel on clients?
10:52
<sutula>
alumno10: Depends on how they boot. PXE?
10:52
<alumno10>
yes
10:52Blinny has joined #ltsp
10:52
<alumno10>
i have a old client that only kernel 2.6.8 support it , and i must to boot from there
10:53
<ogra_cmpc>
alumno10, install a different linux-image package in the chroot and run sudo ltsp-update-kernels
10:53
<alumno10>
but now i have 2.6.14 kernel
10:53
<ogra_cmpc>
thats unlikely to work though
10:53
udev will likely break witch such old kernels
10:53
<alumno10>
after that.. can i select between all kernels?
10:54
<ogra_cmpc>
no
10:54
<sutula>
alumno10: your dhcp server decides which kernel is pointed to for each client, right?
10:54
<alumno10>
yes
10:54staffencasa_ has joined #ltsp
10:54
<ogra_cmpc>
it will use the last installed one from the chroot
10:55
<alumno10>
humm i see... and if this 2.6.8 kernel doesn't work as i expected, there's a way to configure which kernel to use in clients?
10:55
<ogra_cmpc>
sutula, the problem is that with 2.6 kernels there is a lot userspace stuff that depends on a certain kernel version and features that provides
10:55* sutula defers to ogra_cmpc
10:55
<ogra_cmpc>
there are various ones, yes
10:56
<alumno10>
ok, thanks, then i'll install 2.6.8 kernel in chroot
10:56
<ogra_cmpc>
but dont expect it to boot with a .8 version
10:56
<alumno10>
?
10:56
<sutula>
alumno10: Sounds like you could have separate userspaces as well, but the current tools wouldn't support that automatically
10:56
<ogra_cmpc>
the easiest way for testing is probably to just edit pxelinux.cgf/default
10:57
sutula, well, the prob will be to get a chroot created that is able use such old kernelspace stuff
10:57
<alumno10>
this is the story: my ebox clients have sis7019 sound hardware, which doesn't work automatically with edubuntu 7.10 (ltsp5) just installed
10:58
<ogra_cmpc>
ugh
10:58
you run ebox with ltsp5 ?
10:58
err
10:58
s/ltsp5/ltsp/
10:58
<alumno10>
yes...why?
10:58
<ogra_cmpc>
ebox makes a nice kiosk device to run a single app but itrs way to slow as a full client
10:59
<alumno10>
my server is an athlon and it supports like 10 clients
10:59
<ogra_cmpc>
on either ltsp version (imho)
10:59
<alumno10>
i use icewm instead gnome
10:59
and it run so so
10:59
<ogra_cmpc>
well, the desktop has not much to do with the client
11:00
apart from the amount of colors probably
11:00
<alumno10>
humm, 16bit color, and 800x600 res no ssh encryption
11:00
<ogra_cmpc>
ah, k
11:00
800x600 is probably even ok with an ebox
11:00
(i never tried it with less than 1024x786,)
11:01
<alumno10>
which client machine do u recommend me?
11:01
and server... to watch youtube fluidly in 30 clients
11:01
<ogra_cmpc>
at the same time ?
11:01
<alumno10>
yep
11:02
<ogra_cmpc>
then your clients should probably have gigabit NICs
11:02
and you need a powerplant to run your network HW
11:02
<alumno10>
my clients?
11:02
my server have a gigabit
11:02
i think doesnt matter clients have a 10/100
11:03
<ogra_cmpc>
well
11:03
<alumno10>
ok, i'll check kernel thing now
11:03
<ogra_cmpc>
imagine 24frames of 800x600x24bit size being shoveled over the net ...
11:03
every second
11:04
<alumno10>
but isnt fullscreen
11:04
<ogra_cmpc>
you will ptrobably get a handfiull of clients displaying that in a fluid way
11:04
but 30 ...
11:05
<alumno10>
humm , i called to dell and they recommend me a xeon
11:05
<ogra_cmpc>
flash and ltsp are not going well together as long as firefox runs on the server ...
11:05
to much network traffic that way
11:05
<alumno10>
humm.. do you think flash could run locally in this tinny machines?
11:05
<ogra_cmpc>
there is no compression or anything for your video frames they go over the net in raw format
11:06
24 of them per second per client
11:06
<alumno10>
hummm
11:06
<ogra_cmpc>
there is no localapp implementaition that works yet ... so currently you would have to implement a locally running firefox yourself
11:07
<alumno10>
ok...
11:07
hey, i tried to install kernel, and i got this :/
11:07
run-parts: executing /etc/kernel/postinst.d/ltsp-update-kernels
11:07
Could not find where the current root file system is mounted!
11:07
Please make some arrangements for your system to boot the new
11:07
kernel (like running LILO, loadlin, SILO, QUIK, VMELILO, or
11:07
getting a boot floppy).
11:07
Please hit return to continue.
11:07
(im chrooted)
11:07
<ogra_cmpc>
ugh
11:07
dont chroot
11:08milesd has joined #ltsp
11:08
<alumno10>
but, if i install kernel without chroot, then server will boot with old 2.6.8 kernel?
11:08
<ogra_cmpc>
you need to install the kernel chrooted
11:08
but not run ltsp-update-kernels chrooted
11:09
<alumno10>
i didn't do it, the system did by itself
11:09
:/tmp# dpkg -i kernel-image-2.6.8-20050531_BlueBox.1.0_i386.deb
11:09
<ogra_cmpc>
using an ubuntu linux-image packaged kernel
11:09
<alumno10>
Configurando kernel-image-2.6.8-20050531 (BlueBox.1.0) ...
11:09
run-parts: executing /etc/kernel/postinst.d/ltsp-update-kernels
11:09
Could not find where the current root file system is mounted!
11:09
<ogra_cmpc>
that wont work
11:09
<alumno10>
..
11:09
humm
11:09
<ogra_cmpc>
you need to package the kernel with that packaging system for the current release
11:10
<alumno10>
could you please tell mehow?
11:10
<ogra_cmpc>
thats definately a debian and not an ubuntu kernel
11:10
that cant work
11:10
<vagrantc>
/etc/kernel/postinst.d/ltsp-update-kernels should be a symlink to /usr/lib/ltsp/update-kernels
11:10
<ogra_cmpc>
ubuntu usews a totally differnt packaging setup for its kernels
11:10staffencasa has quit IRC
11:11
<vagrantc>
ubuntu doesn't use /etc/kernel/postinst.d ?
11:11
<ogra_cmpc>
it does
11:11
<alumno10>
how can i get installed kernel 2.6.8 on clients then?
11:12
<vagrantc>
alumno10: it is probably a very complicated task involving a lot of hacking.
11:12
<ogra_cmpc>
build one yourself from vanilla source might work
11:13
<alumno10>
or, what about if i insmod sis7019 chrooted?
11:13
<ogra_cmpc>
that would try to insmod it to the server
11:13
<vagrantc>
for every ebox attempted to be used with LTSP, $deity kills a kitten
11:14
<ogra_cmpc>
++
11:14spectra has joined #ltsp
11:14
<ogra_cmpc>
vagrantc, well ...
11:14
<alumno10>
where goes the modules in client filesistem?
11:14
<ogra_cmpc>
vagrantc, not actually true ...
11:14
<alumno10>
lol!
11:14
<vagrantc>
ogra_cmpc: really? :)
11:14
<ogra_cmpc>
my firewall/router is my ebox sample :) .... booting a special squashfs
11:14* vagrantc was pretty sure vagrantc heard a kitten screaming in agony when alumno10 booted the ebox
11:15
<alumno10>
when i was a child i burnt a kitten so it doesnt matter
11:15
lol
11:16
it was an accident...a sad accident
11:16
modprobe sis7019
11:16
FATAL: Module sis7019 not found.
11:16* ogra_cmpc is trying to patition an nbd device ....
11:16
<alumno10>
where i put the sound module?
11:17
now im using 2.6.14-386 kernel
11:18
<vagrantc>
*finally* ltsp in sid is in a good state again.
11:18
<alumno10>
and why when i do uname -a in a chrooted terminal, i get server kernel, but when i do the same in tty1 console i get the real client kernel?
11:18
<ogra_cmpc>
vagrantc, seen the recent post to ltsp-discuss ? where the guy needed to install xdialog ?
11:18
<vagrantc>
alumno10: because chroot'ing doesn't change what kernel you're running.
11:19
<ogra_cmpc>
i wonder what went wrong there
11:19* vagrantc admits to not reading ltsp-discuss
11:19
<ogra_cmpc>
he's on debian and had a message that he should install xdialog on first terminal boot
11:20
<vagrantc>
ogra_cmpc: it's probably because the ltspfs in lenny depends on a version of ldm that's not yet in lenny, so instead of installing ldm it installed sdm-terminal.
11:20
<ogra_cmpc>
yeah, thats what i suspected
11:20
<alumno10>
where can i put sound modules, so the system find it and finally can do insmod module?
11:21
<vagrantc>
i've actually dropped sdm-terminal with the last ltsp upload to debian
11:21
<ogra_cmpc>
damned ....
11:21
the partitioning works fine ... but i got no devices for the partitions afterwards
11:21
<vagrantc>
ogra_cmpc: what you messing around with partitioned nbd devices for?
11:22
<ogra_cmpc>
creating them with mknod doesnt help either
11:22
vagrantc, i try to write a script to create a proper grub booting image that you just can dd to an usb key
11:22
grub, fdisk and mkfs all edxpect a blockdevice, they dont work on images
11:23
so i'm nbd loop mounting the image file
11:23
then i fdisk it which works fine
11:23
<vagrantc>
ogra_cmpc: the debian-live code does that using loop devices somehow
11:23
<ogra_cmpc>
using grub and ext3 ?
11:23
<vagrantc>
ogra_cmpc: i believe so
11:23
<ogra_cmpc>
inside the image
11:24* ogra_cmpc cant imagine debian live using either of these
11:24* vagrantc is floored at how many things debian-live supports
11:24
<vagrantc>
they're insane.
11:25
<ogra_cmpc>
i'd not expect grub
11:25
but i'll take a look, thanks for the hint
11:29
<vagrantc>
they definitely support grub
11:32
<ogra_cmpc>
hmm, is debian-live a package ?
11:33
<vagrantc>
not by that name ...
11:34
there's several related packages ...
11:34
live-helper probably has what you're looking for
11:37* vagrantc is confused by ltsp related popcon ratings ... in the last week, some system has used a binary from the package at least once: ldm 8, ltsp-server 238, ltsp-client 1, ltspfsd 5, ltspfs 75
11:38
<vagrantc>
i would expect similar results for ltsp-client, ldm and ltspfsd ...
11:38
i guess they're all pretty low numbers ...
11:38plamengr has quit IRC
11:39
<vagrantc>
but what are the 4-7 systems using ltspfsd and ldm but not using ltsp-client ?
11:39daya has joined #ltsp
11:40daya has joined #ltsp
11:40
<alumno10>
damn, i installed 2.6.8 kernel
11:40
clients doesnt boot now :S
11:40* vagrantc notes that alumno10 had been warned
11:40
<alumno10>
i try to remove or purge that kernel (always chrooted)
11:40
but i get an error
11:41
Desinstalando kernel-image-2.6.8-20050531 ...
11:41
run-parts: executing /etc/kernel/postrm.d/ltsp-update-kernels
11:41
Cannot open ``/boot/nbi.img-2.6.22-14-generic'':File exists
11:41
run-parts: /etc/kernel/postrm.d/ltsp-update-kernels exited with return code 1
11:41
Failed to process /etc/kernel/postrm.d at /var/lib/dpkg/info/kernel-image-2.6.8-20050531.postrm line 260.
11:41
dpkg: error al procesar kernel-image-2.6.8-20050531 (--remove):
11:41
el subproceso post-removal script devolvió el código de salida de error 2
11:41
Se encontraron errores al procesar:
11:41
<vagrantc>
!pastebot
11:41
<ltspbot>
vagrantc: "pastebot" is The LTSP pastebot is at http://pastebot.ltsp.org. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. A link to the content will be pasted in the channel.
11:41
<alumno10>
kernel-image-2.6.8-20050531
11:41
E: Sub-process /usr/bin/dpkg returned an error code (1)
11:41
<vagrantc>
alumno10: please use the pastebot in the future
11:42
<alumno10>
ok, sorry
11:42
<vagrantc>
alumno10: you can just delete /boot/nbi.img*
11:42
alumno10: and then remove your kernels ... they should get regenerated.
11:44
*sigh*
11:44
<alumno10>
where are the kernels in chrooted?
11:44
<vagrantc>
alumno10: the kernels in the chroot should be in /boot
11:45
alumno10: the nbi.img* are generated by ltsp hooks to the kernel postinstall/postremove stuff
11:46
<alumno10>
ok
11:47
<vagrantc>
so ... now i just need to find myself a nice m68k system to upload the next ltsp with ... so then i don't have to wait for the ridiculously slow m68k buildd, which hasn't built ltsp packages in 10 months and is preventing me from uploading ldm ...
11:47
er, not 10 months ... not since october ... more like 6 months :)
11:48
<cliebow_>
vagrantc:you are in Macintosh hell?
11:48* vagrantc notes that m68k isn't in the current stable release or even the testing version
11:48
<vagrantc>
cliebow_: looking to get into some macintosh hell, yes.
11:49
<cliebow_>
why not step up to nubus 8~)
11:49
<vagrantc>
cliebow_: any idea what the faster m68k mac's are ?
11:49
<cliebow_>
not off top of the head
11:49
<vagrantc>
cliebow_: i tend to see quadra's and performa's coming through freegeek fairly often
11:49
<cliebow_>
i came in at 6100/5260
11:49
<vagrantc>
those were powerpc?
11:49
<cliebow_>
yeah..i think
11:49
<ogra_cmpc>
vagrantc, cluster some quadras :)
11:50
<vagrantc>
ogra_cmpc: i just need to get one for starters ...
11:50
then we'll look at clustering
11:50
<ogra_cmpc>
you just said you see them often
11:50
<vagrantc>
yes, but i just realized how this would be useful to me, so i haven't grabbed any yet
11:50
<cliebow_>
vagrantc:we have pretty much trashed anything we ever had from that vintage..
11:51
<ogra_cmpc>
cliebow_, everyone has ... apart from debian :P
11:51
<cliebow_>
and i always joked about a beowulf cluster..
11:51
<vagrantc>
ogra_cmpc: even debian has ... it's not part of etch officially, it's not supported by the security team, and it's not in lenny
11:51
ogra_cmpc: it's jjust hanging in there enough to get in my way
11:52
<alumno10>
vagrantc, it doesn't appear again vmlinuz or nbi.img , and clients still doesnt boot :S, it just reboot when they load initrd.img
11:52
<vagrantc>
alumno10: make sure the symlinks in /opt/ltsp/i386/boot/vmlinuz initrd.img point to the appropriate kernel, and make sure to run ltsp-update-kernels, and then try booting
11:53
i just need an m68k debian sid machine long enough to get one upload to debian in :)
11:54
and then i can forget it ever existed
11:54* vagrantc finds the situation kind of ridiculous
11:55
<vagrantc>
then i can start grumbling about mips(el)
11:56
seems like mips/mipsel has hovered between 150 and 450 in the buildd queues ... the frustrating part is it keeps going back up
11:58* ogra_cmpc has an ingigo2 here
11:58
<ogra_cmpc>
*indigo
11:58
but no linux on it
11:59
<vagrantc>
i've got an SGI o2, but it's got an r10000 proc which isn't supported by debian, an SGI octane2 which doesn't seem to respond to video or serial console, one indigo similarly unresponsive, and two indigo's i haven't yet tried, an a cobalt qube without a power supply
12:00
<cliebow_>
sounds like a school..
12:00
<ogra_cmpc>
i guess the cobalt is your best bet
12:00
<vagrantc>
i got part way through a debian install on the O2, but then some bug with the SCSI and/or network drivers kicks in
12:01
<ogra_cmpc>
it actually has some decent performance and runs linux easily
12:01
<vagrantc>
ogra_cmpc: well, the power supply is a strange wall-wort
12:01
<alumno10>
vagrantc: everything seems fine by now, wow... this things MUST be working at 15.00 localtime an this time is now , thanks!
12:01
<ogra_cmpc>
vagrantc, i bet you can get data and wiring diagrams for it
12:02
<cliebow_>
alumno10, so you got it working then....
12:02
<vagrantc>
ogra_cmpc: yeah, that was what one person said... we could build a custom power supply ... but kind of a lot of work without knowing that it even works
12:02
<cliebow_>
vagrantc: i can send you a couple netvistas to add to your collection..
12:02
<vagrantc>
cliebow_: heh. i've seen a few of those now and again
12:02
<alumno10>
sh*t! i got a lot of errors
12:03
<cliebow_>
someone has the ppc version more or less working
12:03
<vagrantc>
cliebow_: i'm mostly interested in the ones with bigger numbers: http://people.debian.org/~igloo/status.php?packages=ldm
12:04
<cliebow_>
isee!
12:04
<alumno10>
after load nbi. img i got: "mount: mounting /dev/nbd0 on /rofs failed: no such device... and so, all filesystems
12:04
<vagrantc>
i haven't tried powerpc in a while ... in fact, i don't even remember if i ever got it working.
12:04
<alumno10>
im pretty scared
12:04
+o(
12:05
<vagrantc>
seems like some of the openfirmware implementations required manually specifying the network boot options ... it wouldn't take the filename from dhcp...
12:06
we've got a lot of those gumdrop looking powerpc's ...
12:06
<cliebow_>
a lot neede like a bootp..
12:06
static
12:07* vagrantc starts eyeballing hppa
12:08
<alumno10>
oh sh*t is working :S
12:08
i can believe it! thanks all!!
12:09
<vagrantc>
any idea what a HP PA-RISC is like?
12:09
<cliebow_>
not at all..
12:10
<alumno10>
thats weird , when i put kernel....-generic instead ...-386 the system doesn't mount any filesystem
12:10
<cliebow_>
my one advatnure was a solaris box..
12:18q_ has quit IRC
12:20vagrantc has quit IRC
12:33cliebow_ has quit IRC
12:38mhterres has left #ltsp
12:47daniQ has joined #ltsp
12:50vagrantc has joined #ltsp
12:50ogra_cmpc_ has joined #ltsp
12:50ogra_cmpc has quit IRC
12:51alumno10 has quit IRC
12:51alumno10 has joined #ltsp
12:52daya has quit IRC
12:54joebaker has joined #ltsp
12:56mcfloppy has quit IRC
12:57ogra_cmpc has joined #ltsp
12:58ogra_cmpc_ has quit IRC
13:03daniQ has quit IRC
13:07alumno10 has quit IRC
13:13
<milesd>
vagrantc, so why all this interest in multiple architectures?
13:15alekibango has joined #ltsp
13:16
<vagrantc>
milesd: because debian supports them, and some of the slower ones are interfering with migration from debian unstable to debian testing
13:16
<milesd>
aah, I see
13:16
vagrantc, is it not building on "some of the slower ones?"
13:17
<johnny>
when will debian drop them?
13:17
<vagrantc>
milesd: it has been since october that ldm was built on m68k
13:17
<milesd>
vagrantc, right, but are the builds failing, or are they just not being done?
13:17
<vagrantc>
milesd: they're just slow
13:17
<milesd>
...so it's been *building* since OCT?
13:18
<joebaker>
Slow to build or slow to execute?
13:18
<vagrantc>
it hasn't started a new build since october
13:18
<milesd>
vagrantc, right, but is that because there are no new builds scheduled, or because it's still clogged up with other from the October-era?
13:19
<vagrantc>
milesd: it's 400+ or so in the queue
13:19
<milesd>
wow
13:19
sounds like Debian needs some more machines in the build-farm
13:19
<joebaker>
Wow, so if I could get my old amiga on the network (m68k...)
13:19
<vagrantc>
it's a social problem as much as a technical one
13:19
<milesd>
...or that they need to drop m68k
13:19
<vagrantc>
i'm just looking at ways to work around the problem for ltsp
13:19
<milesd>
is anybody using it anymore?
13:20
<vagrantc>
m68k is dropped from everything but debian unstable officially
13:20* ogra_cmpc guesses that 90% of the software built on m68k wont even run on that low ressources
13:20
<milesd>
ogra_cmpc, that'd be my guess, too
13:21
<ogra_cmpc>
but hey, i'm spoiled ... ubuntu cares only for three arches atm
13:23
(and one of them is server only :) )
13:23
<Blinny>
Will Hardy have a method for flash sound on terminals booting from a 64-bit server?
13:23mistik1_ has joined #ltsp
13:24
<milesd>
ogra_cmpc, out of curiosity, which architectures are those?
13:24
<ogra_cmpc>
Blinny, if someone manages to build a libflashsupport64
13:24
milesd, i386, amd64 and sparc
13:24
<milesd>
...sparc, eh?
13:25
<ogra_cmpc>
yes, we're sun partner
13:25
<Blinny>
I do see an amd64 arch on http://packages.ubuntu.com/hardy/libs/libflashsupport
13:26
<ogra_cmpc>
ubuntu is fully certified on niagara
13:26
Blinny, you need one that hooks into lib32
13:27
flash will only run in a 32bit env ... linbflashsupport for 64bit would work with a 64bit flash plugin if there were one
13:28daniQ has joined #ltsp
13:28
<Blinny>
And right now (Gutsy) flash is run through a wrapper?
13:29
<ogra_cmpc>
yes, it will be the same in hardy
13:29
nspluginwrapper runs 32bit plugins in a 64bit env
13:29
<Blinny>
Gotchya.
13:30
That's the way I did it in FC6, but had to set it up by hand.
13:30
<ogra_cmpc>
the thing is that the 32bit pklugins need their libs to be sitting in /usr/lib32
13:31
<Blinny>
I've read about this lib32 lib64 naming convention differences. Seems like instead of keeping /lib it'd be wise to have both /lib64 and /lib32 and then link /lib to whichever arch your kernel is.
13:31
<ogra_cmpc>
so libflashsupport needs special packaging here which nobody did yet
13:31
<Blinny>
*Nod*
13:31
<ogra_cmpc>
i think ther are launchpad bugs about it open
13:31
<Blinny>
And it wouldn't hurt to have Adobe do some of the word too.
13:31
:s;ord;ork
13:32
<ogra_cmpc>
well, providing a 64bit build wouldnt hutr
13:32
then all that silly workarounds wouldnt be needed
13:32
<Blinny>
OK. Thanks.
13:33sonjagonz has quit IRC
13:33sgonzalez has quit IRC
13:33sgonzalez has joined #ltsp
13:38mistik1 has quit IRC
13:38mistik1_ is now known as mistik1
13:45daniQ has quit IRC
13:46
<warren>
ogra_cmpc, I wouldn't hold my breath about a 64bit build
13:46
<ogra_cmpc>
i know
13:47
but if i look at all the effort going into sall these workarounds the wasted manpower makes me cry
13:48
<vagrantc>
any hope from gnash or the other implementation?
13:48
other free implementation(s)
13:49
<ogra_cmpc>
i bet if you would pay 10ct for every mnhour that went into making flash on 64bit work you could pay adobe a fulltime dev who could take care of the 64bit port
13:49
*manhour
13:49
vagrantc, i have one user who claims to be happy with gnash on his gutsy ltsp
13:50
(amd64)
13:50
but i'm not sure he uses whats in the archive since he's the only one i heaqrd anything positive from
13:51
<milesd>
vagrantc, I've seen gnash work, but it chews up a *lot* of CPU time and on my P4-2.8Ghz laptop it can't play a movie without stopping every few seconds to let the CPU catch up
13:51
vagrantc, But it's never crashed on me, and so far I've not seen a flash file it couldn't play
13:52
vagrantc, so I'm hopeful, but it's not there yet
13:52elisboa has quit IRC
13:52elisboa has joined #ltsp
13:53
<ogra_cmpc>
vagrantc, debian live sadly doesnt have anything that would help me to create partitions inside an image, but the grub code looks worth to steal :)
13:54
<vagrantc>
ogra_cmpc: well, that's something at least
13:55
<warren>
unfortunately gnash even with unlimited manpower will never achieve 100% capability due to software patents =(
13:55
<ogra_cmpc>
pfft patents ...
13:56jammcq has quit IRC
13:59
<milesd>
warren, you have any specific patents in mind?
14:00
<ogra_cmpc>
milesd, there are surely some involved
14:00
<milesd>
warren, or did you just fall victim to a FUD campaign?
14:00bobby_C has joined #ltsp
14:01
<milesd>
ogra_cmpc, of course there are. Microsoft patented the "and not" logical operator and Amazon patented clicking a mouse
14:01
<ogra_cmpc>
but patents are a theoretical thread ... not dangerous unless someone enforces them ...
14:01
<milesd>
ogra_cmpc, you're probably violating a patent every time you breathe
14:01
ogra_cmpc, warren: the question is, are there any *enforcible* patents to worry about?
14:01
<ogra_cmpc>
keeping away from coding something because it could be patented is like not going out because a car could run you down
14:02
<milesd>
I even saw one patent for tying your shoe laces tied
14:02* vagrantc suspects you're more likely to get run down by a car
14:02
<ogra_cmpc>
yeah
14:02
<milesd>
"frictive fluid application for footware fastening devices" or something like that
14:03
It had a diagram of a shoe-lace tied in a nice-looking bow with a bottle of glue aimed at it... very professionally done :)
14:04
I'm not saying it can't be a problem
14:05
and I'm not saying we don't have a responsibility for due diligence
14:05
but SW patents aren't nearly so hairy as everybody seems to dread
14:07alumno10 has joined #ltsp
14:11
<alumno10>
when client boots, it says " module-detect.c: failed to detect any sound hardware module.c:failed to load module-detect.c"
14:11
but when i do lspci in tty1 terminal i got "... audio controler: sis 7019....."
14:12
what does it mean?
14:13
vagrantc, are you there?
14:15cdealer has joined #ltsp
14:16
<cdealer>
hi ... just a question ... why even after all clients had logout I still have some (about 3 to 6) users showing in top ?
14:17
and their process arent idle...
14:17
<alumno10>
hummm, sometimes it happened to me
14:18
sometimes bonobo still running
14:18
with an user which isnt logon
14:18
<cdealer>
this happens everyday .. everybody goes home and there are ... 6 users connect, and their processes like nautilus are active and consuming about 10% of cpu
14:18
<alumno10>
(had logged off minutes before)
14:18
humm... thats weird
14:19
<cdealer>
yep
14:19
<alumno10>
check logs
14:19
when they logged in
14:19
and when and from where
14:20
<cdealer>
Hmmmm I will take a look ... do you think that an incorrect shutdown may let they lock for sometime in the box ?
14:21
<Blinny>
Woah. My server's monitor just blanked after sitting idle at the gnome login window... and it killed my thin client session sitting right beside.
14:21
<alumno10>
could be but its difficult it happen everyday
14:23
<cdealer>
Hmm
14:24
dbus-daemon and gnome-vfs-daemon are two process that normally persists to stay up
14:25sgonzalez has quit IRC
14:25
<cdealer>
I have a user that had gone about 1 hour ago and there are 5 dbus and 5 gnome-vfs process idle
14:25
I will start killing these users
14:25
<Blinny>
NO!
14:25
Kill their processes
14:26
You'll end up in serious trouble if you start killing your users.
14:26
<cdealer>
Blinny, welll ... to late for some users O_o
14:26
Blinny, what could happen ?
14:26
just to get prepared...
14:26
<ogra_cmpc>
you could go to jail or worse i guess
14:27
ddepending where you live
14:27
<cdealer>
hahaha good one ogra ... ;P
14:27
<Blinny>
Seriously.
14:27
<vagrantc>
alumno10: lspci just tells you what cards are in the system, which doesn't mean that your kernel can support that hardware
14:28
<alumno10>
humm ok, thanks, that's exactly what i though
14:28
lol!
14:28
<cdealer>
Blinny, now talking seriously... you was joking or are you saying killing direct the users with killall -u user may cause problem?
14:28
<ogra_cmpc>
you need the full oss stack for that driver ... and likely an alsa-oss plugin installed
14:28
<alumno10>
if you kill users processes still will run
14:29
<Blinny>
cdealer: Just joking :)
14:29
<ogra_cmpc>
alumno10 has a point
14:29
<cdealer>
ah ok
14:29
ogra, but a killall -u user dont close all user process stances?
14:29
<ogra_cmpc>
use pkill rather
14:30
that will kill all processes, yes
14:31
<cdealer>
I thought that was the same then killing processes one by one ...
14:31
but pkill may be usefull as well
14:33
ogra, now going in the other way ... our server with 30 users was using only 5gb of memory ... very nice I guess ... what I notice is that sometimes nautilus get stuck and I have to kill then run again to the desktop come to normal with wallpaper and things like that
14:35Blinny has quit IRC
14:35
<ogra_cmpc>
do you use any kind of network mounted homedirs ?
14:35
<cdealer>
ogra_cmpc, no I dont... all in the ltsp box
14:36
ogra, Im very excited with the Idea of putting our implementation as a case of success on the ubuntu site... ltsp help us to save a lot of money with windows licenses and robust boxes to handle local resources...
14:36* ogra_cmpc hasnt heard something like that yet
14:36toscalix has quit IRC
14:37Q-FUNK has joined #ltsp
14:37
<cdealer>
ogra_cmpc, here happens about 2 to 3 times a day, havent inspected the logs yet
14:39
<alumno10>
where i have to put modules if i want to modprobe it?
14:40
i got Module sis7019.ko not found.
14:41
<cdealer>
I was going home ... but I will whait for this answer ;)
14:53dmorar has joined #ltsp
14:54
<dmorar>
Hi, im trying to get ltsp 5 to work on slackware using debian tarballs
14:54
so far i get a ldm login screen, and i can go into ssh if i use one of the tty, but i cant go into the graphics mode, i suspect i have a problem with Xsession, how do i debug this?
14:55
i dont think i have a problem with my ssh_config
14:55
would you help me please?
14:55cdealer has quit IRC
14:56
<vagrantc>
dmorar: where are you getting the tarballs from?
14:56
and why use the tarballs at all?
14:57
<dmorar>
vagrantc, i dont think, it is a problem with the tarballs, i got them from ltsp site, it was working but i suffered a data loss, and had to restart all my system, fortunatelly i had a backup, but lost all the documentation i had done when i set ltsp to work, i remember i had to mess with xsession issues...
14:57
<vagrantc>
dmorar: well, the only tarball i know is *really old*
14:57
dmorar: it was created before debian etch was released, and probably has lots of security bugs
14:58
<dmorar>
vagrantc, im using the tarballs, since i dont know how to get it to work on slackware
14:58
<vagrantc>
ah.
14:58
s,probably,definitely,
14:59
<dmorar>
but im here to hear your advice, if you think there is a better solution...
14:59
i have a i386 directory, all of slackware 12 bits...
15:00
<vagrantc>
well, possibly updating your chroot, or i could release a new tarball...
15:00
but before i make a new tarball, i'll want to get the latest ltsp working on etch
15:01
<dmorar>
vagrantc, if you tell me which files i have to tweek in order to get ltsp 5 slackware only bits, i would do it gladly
15:02
<vagrantc>
dmorar: there's no tweaking to be done, that's a whole project unto itself
15:02
!integration
15:02
<ltspbot>
vagrantc: "integration" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/IntegratingLtsp
15:02
<vagrantc>
dmorar: have a look at that document to get started
15:04
<dmorar>
vagrantc, ok, i have already read this, im stuck on step 1 tough, i dont understand some things on step 2
15:06
<alumno10>
what does ltsp-update-image
15:08
<warren>
milesd, it isn't FUD. Simply look at what it does. Sorenson, MP3, H.264 and many other codecs are only a few examples.
15:09
<milesd>
warren, can you be more specific? There are free, OSS mp3 encoders and decoders and I don't know of anybody who has faced legal action for them
15:09
<vagrantc>
dmorar: if you don't understand step 1, you have to approach this from a perspective that you have a *lot* of stuff ahead of you to learn :)
15:09
<dmorar>
vagrantc, i already done step 1
15:09
<vagrantc>
dmorar: then how are you stuck on it?
15:09
<warren>
milesd, a codec would need to be non-GPL in order to be legal outside of America
15:09
milesd, (due to patent & GPL incompatibility)
15:09
<dmorar>
vagrantc, after step 1, but before step 2 :P
15:10* vagrantc doesn't see anything between the two
15:10
<vagrantc>
dmorar: if you could be more specific, i *might* be able to help you
15:10
<warren>
milesd, in any case you still don't have a freely available and fully capable flash alternative
15:10
<dmorar>
vagrantc, lol ... i dont know how to install the LTSP supplied scripts/executables, i dont even have a copy of them
15:10
:)
15:10
<milesd>
warren, not yet, that's true... But flash is both younger and more complicated
15:11
warren, and how, exactly are patents incompatible with GPL?
15:11
<vagrantc>
dmorar: so you actually built a slackware chroot and configured it to boot to the network?
15:11
dmorar: the rest is just details :P
15:11
dmorar: if you got that far, why are you messing with debian tarballs?
15:12
<warren>
milesd, read the GPL and consult your own lawyer
15:12
<milesd>
warren, I'll admit I'm not up on GPLV3
15:12
warren, but GPLv2 has nothing to say about patents
15:12
<warren>
milesd, it might be technically feasible with fluendo and gstreamer I admit
15:12
<dmorar>
vagrant, i can boot with a kernel, and have a chroot enviroment... (i dont know if it will boot the nfs share),
15:13
<warren>
milesd, however you're paying ~40 euros and you still don't have full capability
15:13
<dmorar>
it boots the debian share
15:13
<warren>
milesd, difficult to tell users that's their only option
15:13
vagrantc, regarding update-kernels, I'm afraid we can't use it directly as is
15:13
<vagrantc>
dmorar: so it's a slackware server with a debian ltsp environment?
15:13
warren: figured that was a possibility
15:14* vagrantc never expected anything to work for other distros without *some* changes
15:14
<warren>
vagrantc, I really like eharrison's thing that takes whatever latest kernel installed in the chroot and copies it outside and symlinks it. That way the PXE config is untouched by the update tool and custom options persist, including the possibility of a PXE menu.
15:14
<dmorar>
vagrantc, yes, its a slackware server, but if you tell me i can make it to boot slack, then i will do it gladly :) , wait ill see if i can boot my slackware chroot enviroment, havent done so, cause i though it required some scripts in order to do the nfs mount
15:14
<milesd>
warren, well, lame is licensed under lgpl-V2. So far as I know, nobody has faced legal action about it so far and it's available all over the world
15:15
warren, am I wrong about the legal action?
15:15
warren, or about its availability?
15:15
<vagrantc>
warren: well, if you want persistance, have dhcp point to a custom pxe menu
15:15
<warren>
milesd, not being sued != legally safe. No major distributor of linux can distribute it without fear of lawsuit, and even their customers could be sued.
15:16
<milesd>
warren, ah, ok. Now I see where you're coming from
15:16
<vagrantc>
warren: i've long planned on making ltsp-update-kernels/update-kernels more flexible ... just haven't gotten around to it.
15:18
<warren>
vagrantc, for pragmatic reasons and the fact that none of us know really what is best yet, I'm going to write my own alternative, then after our next major releases we can assess what works best and work toward a common solution.
15:20
<vagrantc>
warren: sounds reasonable
15:21
<dmorar>
vagrantc, ok i cant boot into my slackware chroot, i dont know how to set up fstab... and it complains about it
15:21
<warren>
vagrantc, btw, you still go by "Vagrant Cascadian" right? I want to get your name right when I explain who is involvd.
15:21
<vagrantc>
dmorar: sorry, i don't have the time to help further
15:21
warren: sure :)
15:21
<dmorar>
vagrantc, ok, thanks
15:22* ogra_cmpc curses about nbd
15:22
<ogra_cmpc>
apparently there was a version once that supported partitioning ...
15:22
hrm
15:22alumno10 has quit IRC
15:28Q-FUNK has quit IRC
15:30Q-FUNK has joined #ltsp
15:54Q-FUNK has quit IRC
16:01Q-FUNK has joined #ltsp
16:07cesar_ has joined #ltsp
16:07
<cesar_>
hello people
16:08
I have a question
16:08
about local drivers for client ltsp
16:08
well
16:09
I am reading the next link
16:09
http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev#Troubleshooting
16:10
my server ltsp is "LTSP 4.2 Update 3"
16:13
in point Step 8
16:13
I found a problem
16:14J45p3r has joined #ltsp
16:14
<warren>
vagrantc, ogra_cmpc : is there any reason why the System.map-foo must be copied into the tftpdir?
16:15
<vagrantc>
warren: is there any reason not to?
16:15
<cesar_>
in a client ltsp
16:15
I run this
16:15
ps -ef | grep lbussd
16:15
but
16:15
it isn't running
16:15
and then
16:15
<warren>
vagrantc, unneeded redundancy
16:15
<vagrantc>
warren: originally, there was a list of files to copy, but it *greatly* simplified to the code just copy the whole boot dir
16:16
<cesar_>
in the TS I run
16:16
/usr/sbin/lbussd
16:16
how root
16:16
but
16:16
<vagrantc>
warren: and the additional cruft that gets copied is so minimal compared to the additional code ...
16:16
<cesar_>
I get a error
16:16
I get an error
16:17
<warren>
vagrantc, System.map is > 800K
16:17
vagrantc, I guess we have differing opinions on what minimal means
16:18
vagrantc, and that's OK.
16:18
<cesar_>
Use of uninitialized value in substitution (s///) at /usr/sbin/lbussd line 46.
16:18
Use of uninitialized value in concatenation (.) or string at /usr/sbin/lbussd line 62.
16:18
Couldn't establish a connection to :9202: IO::Socket::INET: Bad hostname ':9202'
16:19
this happend when I try run "/usr/sbin/lbussd"
16:19
what happend?
16:19
<johnny>
cesar_, i personally have never ran ltsp 4.x
16:19
<cesar_>
ok
16:19
<johnny>
and many others haven't either
16:19
<vagrantc>
warren: in the grand scheme of how big an LTSP chroot is ...
16:19
<johnny>
so you might be waiting awhile
16:20
<cesar_>
I can see "Bad Hostname"
16:20
but
16:20
hostname is the "hostname" for the ST?
16:20
or client terminal server?
16:24
<warren>
vagrantc, my LTSP chroot is currently 412MB, how large is Debian's?
16:24
vagrantc, I haven't optimized it yet
16:25
<cesar_>
mmm..
16:25
<johnny>
the img ? or /opt/ltsp/whatever ?
16:26dmorar has quit IRC
16:26
<warren>
the chroot itself
16:28
<cesar_>
may be "lsbussd" need a param
16:28
may be "lsbussd" need get a param
16:29
<johnny>
du -csh says 438
16:29
but this is for a gutsy chroot
16:30
at least you know you're smaller than that :)
16:31
<warren>
well, I haven't confirmed that it contains everything we need yet
16:31
I might have stripped out too much
16:32
<vagrantc>
warren: there were some changes to dependency resolution which caused it to get up to around 500MB
16:35
<warren>
btw
16:35
where is $LTSP_NFSROOT defined?
16:36
or maybe upstream doesn't use that
16:36
vagrantc, where do you get the values to pass to your initramfs tool so it knows where to find the NFS or NBD server?
16:38dmorar has joined #ltsp
16:42
<dmorar>
Hi, again, ok, i give up trying to set slackware, i know i have a dumb problem, can you please help me to fix this?
16:46
<johnny>
slackware doesn't have the best ltsp support afaik
16:46
it's still ltsp4
16:46
i imagine?
16:47
you're better off going with something with ltsp5
16:47
<dmorar>
johnny, it was working, im using debian tarballs, i only need to fix ldm, and thats it
16:48
<warren>
vagrantc, like I did for the initscript, I'll try to split out functions that could be common in a separate library
16:49
<cesar_>
ufff.
16:49
in the list ltsp spanish
16:49
I found the same question
16:49
<warren>
vagrantc, ogra_cmpc: hmm, does your LTSP not hardcode any server address in the initrd and instead it gets it from tftp?
16:49
vagrantc, ogra_cmpc: or dhcp?
16:50
i'm not sure how it works
16:50
<cesar_>
what does "lbussd" do?
16:50
but anybody response
16:51
and.... I don't find a "man page" for this script
16:51
:-(
16:55
<dmorar>
ok, i solved it, it was a problem with ssh
16:57
:)
17:01cesar_ has quit IRC
17:02dmorar has quit IRC
17:02rcy` has joined #ltsp
17:03
<vagrantc>
warren: yes, usually the server is grabbed from dhcp
17:05
<warren>
vagrantc, where is the initramfs code and dhcp syntax for that?
17:07
<vagrantc>
warren: ltsp-trunk/client/initramfs/* (except init-bottom, which should probably just be removed)
17:07
warren: scripts is what gets run at boot time, hooks is used during image generation ...
17:07
warren: though it leverages heavily on initramfs-tools ... we just plug into initramfs-tools mostly
17:08mikkel has quit IRC
17:11
<warren>
dhcpd.conf syntax?
17:14moquist has quit IRC
17:14BadMagic has quit IRC
17:14Topslack has quit IRC
17:14jim`` has quit IRC
17:14oh207 has quit IRC
17:18BadMagic has joined #ltsp
17:19moquist has joined #ltsp
17:21
<vagrantc>
warren: typically pulled from next-server or can be included as part of nfsroot
17:21
i guess next-server is technically a hack, but if you're actually hosting your boot images on a different server than the root filesystem ... you've got a mess anyways.
17:22bobby_C has quit IRC
17:24spectra has quit IRC
17:26vagrantc has quit IRC
17:26K_O-Gnom has quit IRC
17:31Topslack has joined #ltsp
17:31oh207 has joined #ltsp
17:32jim`` has joined #ltsp
17:37klausade has quit IRC
17:37Topslakr has quit IRC
17:37dberkholz has quit IRC
17:37GiantPickle has quit IRC
17:39GiantPickle has joined #ltsp
17:39dberkholz has joined #ltsp
17:41jammcq has joined #ltsp
17:41
<jammcq>
!seen sbalneav
17:41
<ltspbot>
jammcq: sbalneav was last seen in #ltsp 2 weeks, 2 days, 8 hours, 42 minutes, and 3 seconds ago: <sbalneav> What are we looking at?
17:41
<jammcq>
hmm
17:43Topslakr has joined #ltsp
17:51klausade has joined #ltsp
18:11staffencasa__ has joined #ltsp
18:13lns has quit IRC
18:20lns has joined #ltsp
18:27staffencasa_ has quit IRC
18:28
<warren>
What is scottie up to these days?
18:30twinprism has joined #ltsp
18:30staffencasa__ has quit IRC
18:31
<jammcq>
warren: I wish I knew
18:31
i've tried contacting him several times
18:37elisboa_ has joined #ltsp
18:46Q-FUNK has quit IRC
18:51staffencasa has joined #ltsp
18:52elisboa has quit IRC
18:59staffencasa has quit IRC
19:10
<lns>
warren, jammcq, sbalneav (sp)? is he not around anymore?
19:10
<johnny>
he's busy
19:11
very busy
19:11
<lns>
ahh
19:13elisboa_ has quit IRC
19:37milesd has quit IRC
19:49cliebow has joined #ltsp
20:17gu_ has quit IRC
20:17gu_ has joined #ltsp
20:54
<rcy`>
im having problems on *some* clients timing out on tftp when my dhcp server is a different machine
20:54
some work fine, however
21:07J45p3r has quit IRC
21:07
<steph_>
With LTSP, does it automaticaly install pulse audio with esound module?
21:31ogra has quit IRC
21:32ogra has joined #ltsp
21:49mccann has quit IRC
22:05cliebow has quit IRC
22:06
<johnny>
steph_, that depends on the distro prolly
22:25oblek has joined #ltsp
22:27oblek has left #ltsp
22:36joebaker has quit IRC
23:10
<steph_>
johnny: are you still there?