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


Channel log from 3 February 2009   (all times are UTC)

00:16alkisg has joined #ltsp
01:05
<alkisg>
!lts.conf
01:05
<ltspbot>
alkisg: "lts.conf" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtsConf
01:38ogra has quit IRC
01:47wwx has joined #ltsp
02:03_Shane has joined #ltsp
02:09wwx has quit IRC
02:14ogra has joined #ltsp
02:17wwx has joined #ltsp
02:34pasmen has quit IRC
02:39_Shane has left #ltsp
03:11ogra has quit IRC
03:14ogra has joined #ltsp
03:27dirigeant has joined #ltsp
03:32zirconiumks has quit IRC
03:41alkisg has quit IRC
03:44dirigeant has quit IRC
04:22ogra has quit IRC
04:25ogra has joined #ltsp
04:49cyberorg has quit IRC
04:49daduke has quit IRC
04:49cyberorg has joined #ltsp
04:49daduke has joined #ltsp
04:54pasmen has joined #ltsp
04:59Egyptian[Home] has quit IRC
05:01Egyptian[Home] has joined #ltsp
05:16ogra has quit IRC
05:18ogra has joined #ltsp
06:26F-GT has quit IRC
06:27F-GT has joined #ltsp
06:30pmatulis has joined #ltsp
06:32pmatulis has joined #ltsp
06:37Yggdrasill_ has joined #ltsp
06:39
<Yggdrasill_>
hi all
06:45Appiah has quit IRC
06:55Appiah has joined #ltsp
06:59warren has quit IRC
06:59F-GT has quit IRC
07:01F-GT has joined #ltsp
07:01GandalfXx_ has joined #ltsp
07:10alkisg has joined #ltsp
07:11polytan has joined #ltsp
07:12
<polytan>
hi
07:27BrunoXLambert has joined #ltsp
07:30Egyptian[Home] has quit IRC
07:32Egyptian[Home] has joined #ltsp
07:37
<polytan>
is pxelinux.cfg/default rewritten all the time when building a new image/kernels ?
07:38
<alkisg>
With ltsp-update-kernels, I just reported a bug yesterday (ubuntu)
07:38
<polytan>
ok
07:38
because I did a nice default file and everything has been removed :(
07:39
<alkisg>
Copy it to /opt/ltsp/i386/boot/pxelinux.cfg/default so that it doesn't get overwritten
07:39
<polytan>
hum
07:40
and ity will be read there by the tftp daemon ?
07:40
<alkisg>
No, but ltsp-update-kernels copies this file to tftp dir, so if they contain the same things, even if it's overwritten, it doesn't matter
07:41
https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/324637
07:41
But the "quick hack" is not good for other distros that don't have this problem (e.g. fedora).
07:41
<polytan>
thanks :)
07:42
I have to do it again so :(
07:43elisboa has quit IRC
07:43elisboa has joined #ltsp
07:46try2free has joined #ltsp
07:47polytan has quit IRC
07:47nicoAMG has joined #ltsp
07:49polytan has joined #ltsp
07:59din_os has joined #ltsp
08:05try2free has left #ltsp
08:25bobby811 has joined #ltsp
08:25
<bobby811>
k anyone awake
08:25
need some help here
08:25
ltsp problems
08:25
<laga>
!ask
08:25
<ltspbot>
laga: "ask" is Don't ask to ask a question, simply ask it, and if someone knows the answer, they'll respond. Please hang around for at least 15 minutes after asking a question, as not everybody constantly monitors the channel.
08:26
<bobby811>
2.6.16.1-ltsp-2
08:26
cannot detect the network card
08:26
kernel panic
08:27
dhcp is running on server machine
08:27
but cannot get other workstations to boot up
08:27
with pxe
08:28
<ogra>
sonds like ltsp4 ... not many people that can help with that anymore ... just wait and with uck someone comes by who knows the old code
08:28
*luck
08:28try2free has joined #ltsp
08:28warren has joined #ltsp
08:28try2free has quit IRC
08:28Gadi has joined #ltsp
08:30ScorpKing has joined #ltsp
08:31
<ScorpKing>
hi guys. i've been trying all day to get a different usplash theme for the ltsp clients but no matter what i do the usplash theme does not change. any ideas?
08:33six2one has joined #ltsp
08:33
<Gadi>
ScorpKing: well, what exactly did you do?
08:34
:)
08:34
<ScorpKing>
Gadi: a shorter answer will be what i didn't do ;)
08:34
<Gadi>
you need to update the splash in the chroot, as well as update initramfs and update kernels
08:34
<bobby811>
any ideas for me Gadi
08:34
<Gadi>
usplash starts in initramfs
08:35
bobby811: here's an idea: tell me what ur question is
08:35
I just got here
08:35
:)
08:35
<bobby811>
k
08:35
cannot detect the network card
08:35
dhcp is running on server machine
08:35
pxe network boot for ltsp thin client
08:35
<Gadi>
what error do you get?
08:35
<bobby811>
broadcom network card
08:35
kernel panic
08:35
could not auto detect the network card
08:35
<Gadi>
what distro and ltsp version?
08:36
sounds like 4.2
08:36
<bobby811>
fedora core 6
08:36
2.6.16.1-ltsp-2
08:36
<Gadi>
yeah, you're on LTSP 4.2
08:36
old stuff
08:36
<bobby811>
oh
08:36
<Gadi>
if its broadcom, it is either b44 NIC
08:36
<ScorpKing>
Gadi: i installed a different usplash theme in chroot and made sure /etc/alt.../usplash... points to it, then update kernel image in chroot and exit chroot. ltsp-update-kernels + rebuild the image and then back to square one.
08:37
<warren>
bobby811: http://k12linux.org
08:37
<Gadi>
bobby811: you can pass NIC=b44 on the kernel command line
08:37
and see if that works
08:37
<bobby811>
how Gadi
08:37
im new at this
08:37
<warren>
Gadi: just point those old users at k12linux.org
08:37
<Gadi>
ok
08:37
well, some fear change
08:37
:)
08:37
<ogra>
Gadi, very unlikely b44 was in the old 4.2 kernel :)
08:37
<Gadi>
bobby811: choices:
08:38
<bobby811>
well im in my linux class here at college
08:38
<Gadi>
1. Upgrade to LTSP5 (preferred) as LTSP 4.2 is no longer supported
08:38
<bobby811>
even the instructor doesnt know the answer
08:38
<warren>
bobby811: LTSP 4.2 stopped development in 2004
08:38
<Gadi>
2. Pass a kernel parameter and hope it works
08:38
here's how:
08:38
<warren>
bobby811: get the modern version from k12linux.org, currently based on Fedora 10. New release yesterday.
08:38
<Gadi>
copy down the MAC address of your thin client
08:38
<bobby811>
i dont know the mac address
08:39
<ScorpKing>
Gadi: i'll be back some other time. have to run
08:39ScorpKing has quit IRC
08:39
<Gadi>
then, create a file in: /tftpboot/lts/2.6.16.1-ltsp-2/pxelinux.cfg/01-aa-bb-cc-dd-ee-ff
08:39
where MAC would be: AA:BB:CC:DD:EE:FF)
08:39
that file that you create should be a copy of the default file in that same directory
08:40
then, edit the line with the kernel arguments adding: NIC=b44
08:40
save
08:40
and reboot the client
08:40alkisg has quit IRC
08:40
<Gadi>
bobby811: to find the MAC address, boot the client with the cable unplugged
08:40
it should report to you the MAC
08:40
<din_os>
yeah but how do you put a smiley in the mac address? lol :) hey Gadi thx for yesterdays tips everything works fine now
08:41
<Gadi>
din_os: cool
08:41
<cyberorg>
hi, i have issue with sound playing from localapps, works fine normal ltsp session, locally PULSE_SERVER is set correctly, but still get (snd_pcm_dmix_open) unable to open slave
08:42
checked permissions of /dev/dsp, /dev/snd/* mixer etc, all looks ok
08:42
where can i look more?
08:42
<Gadi>
cyberorg: what application?
08:42
<cyberorg>
Gadi, i tested with firefox+youtube, and aplay xyz.wav
08:43
<Gadi>
and none work?
08:43
<cyberorg>
no
08:43
<Gadi>
does suse redirect alsa to pulse?
08:43
<cyberorg>
Gadi, yes
08:43
<bobby811>
Gadi it dont reportt the mac
08:43
get error
08:44
pxe-e61 media test failure
08:44
pxe-mof exiting broadcom pxe rom
08:44
<Gadi>
bobby811: ok, plug it back in and just after it gets a response from DHCP.... press the Pause key
08:44
<cyberorg>
Gadi, hmm, good hint, let me check that config
08:44
<bobby811>
k
08:45
<Gadi>
or check your syslog (/var/log/messages)
08:45
should be the last client to make a DHCPDISCOVER request
08:45
<bobby811>
k i got the mac address
08:45
<warren>
bobby811: seriously, get the modern version from k12linux.org
08:45
<Gadi>
ok, translate the : to -, make all lowercase, and prefix with 01-
08:45
<warren>
bobby811: your drivers have not been updated since 2004
08:45
<Gadi>
and that is the filename
08:46primeministerp has joined #ltsp
08:47
<Gadi>
warren: he's in a college - this is edumucational
08:47
:)
08:48etyack has joined #ltsp
08:49
<bobby811>
k that didnt work Gadi
08:49
still same error
08:49
made file
08:49
prompt 0
08:49
label linux
08:49
kernel bzImage-2.6.16.1-ltsp-2
08:49
append rw root=/dev/ram0 initrd=initramfs.gz
08:49
#option option-128 e4:45:74:68:00:00; # NOT a mac address
08:49
#option option-129 "NIC=ne IO=0x300";
08:49
#option option-128 code 128 = string;
08:49
#option option-129 code 129 = text;
08:49
NIC=b44
08:49
insiode is this
08:49
for filename default
08:50
<Gadi>
it needs to be inline with the append line
08:50
<bobby811>
../tftpboot/lts/2.6.16.1-ltsp-2/pxelinux.cfg
08:50
<Gadi>
should be: append rw root=/dev/ram0 initrd=initramfs.gz NIC=b44
08:51
<bobby811>
k let me try that
08:51
<Gadi>
and should be in filename: ./tftpboot/lts/2.6.16.1-ltsp-2/pxelinux.cfg/01-aa-bb-cc-dd-ee-ff
08:51
(replaced with your MAC address)
08:51
<bobby811>
k
08:52
<Gadi>
for edification, what you are doing is passing a kernel argument called NIC with value b44, which the initrd(initramfs) will use to select the driver
08:52
since it cannot select it on its own
08:52
<bobby811>
k now get a new error
08:52
<Gadi>
by creating that file, you make sure the setting only applies to a client with an ethernet interface with that MAC address
08:52
<bobby811>
could not find kernel image linux
08:53
<Gadi>
!pastebot
08:53
<ltspbot>
Gadi: "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.
08:53
<Gadi>
bobby811: use that link to paste your file
08:54litlebuda has joined #ltsp
08:55
<ltsppbot>
"bobby-school" pasted "thin client wont boot sits now at boot:" (11 lines) at http://pastebot.ltsp.org/231
08:55
<Gadi>
ok, delete the first 2 lines
08:55
and delete all of the commented lines
08:57
<bobby811>
k
08:57
done
08:58
still same error now
08:58
could not find kernel image linux
08:58
<Gadi>
can you paste again?
08:59
<ltsppbot>
"bobby-school" pasted "thin client wont boot sits now at boot:" (11 lines) at http://pastebot.ltsp.org/232
08:59
<Gadi>
uhh.... that looks the same
08:59
:P
08:59
<bobby811>
well i deleted the lines u told me to delete in conf
08:59
<Gadi>
no you didnt
08:59
you just pasted the same file
09:00
with no lines deleted
09:00
<bobby811>
kernel bzImage-2.6.16.1-ltsp-2
09:00
append rw root=/dev/ram0 initrd=initramfs.gz NIC=b44
09:00
is wat the file reads atm
09:00
<Gadi>
oh, compare that with: http://pastebot.ltsp.org/232
09:00
:)
09:00
<bobby811>
yea
09:00
diff
09:00
but conf is set to those 2 lines only
09:01
<Gadi>
ok
09:01
<bobby811>
now its sitting at boot:
09:01
could not find kernel image: linux
09:01
<Gadi>
can you look at the permissions on the file
09:01
make the permissions match the permissions of the default file
09:01
in that directory
09:02
<bobby811>
-rw-r--r-- 1 root root 90 Feb 3 14:54 00-1e-c9-46-6b-6b
09:02
<Gadi>
ls -l default
09:02
oh, and the filename is wrong
09:02
<bobby811>
-rw-r--r-- 1 root root 90 Feb 3 14:54 00-1e-c9-46-6b-6b
09:02dirigeant has joined #ltsp
09:02
<Gadi>
it should be: 01-00-1e-c9-46-6b-6b
09:02
<bobby811>
is the dwefault
09:03
<Gadi>
no, thats the same as what you just pasted
09:03
do this:\
09:03
pwd
09:03
<bobby811>
[root@localhost pxelinux.cfg]# pwd
09:03
err
09:03
.../tftpboot/lts/2.6.16.1-ltsp-2/pxelinux.cfg
09:03
<Gadi>
ok, good start
09:03
you should have 2 files in that directory:
09:03
default
09:04
01-00-1e-c9-46-6b-6b
09:04
<bobby811>
default i renamed to backup
09:04
<Gadi>
no, keep that default
09:04
or nothing will boot
09:04
<bobby811>
k
09:04
<Gadi>
default is the default config for all pxe boot thin clients
09:04
unless there exists one for a specific one
09:04
which you are creating
09:05
and which should be named:
09:05
01-00-1e-c9-46-6b-6b
09:05
note the 01- prefix
09:05
<bobby811>
k
09:05polytan has quit IRC
09:05
<Gadi>
"01" is the designation for ethernet
09:05
<bobby811>
ah ok
09:06
still same error
09:06
could not find kernel image: linux
09:07
<Gadi>
ls
09:07
<bobby811>
[root@localhost pxelinux.cfg]# ls
09:07
01-00-1e-c9-46-6b-6b backup2 default
09:07
[root@localhost pxelinux.cfg]#
09:07polytan has joined #ltsp
09:09
<Gadi>
ok, lets reset
09:09
cp default 01-00-1e-c9-46-6b-6b
09:09
and reboot the thin client
09:10alkisg has joined #ltsp
09:10
<SDuensin>
Greetings, all.
09:11
<bobby811>
k did that
09:11
rebooting client
09:11
<Gadi>
you should hopefully be back at the NIC error
09:11
<bobby811>
somewat yea
09:11
<Gadi>
somewhat?
09:12
<bobby811>
b44 errors now
09:12
<Gadi>
good
09:12
now, simply add NIC=b44 on the end of the append line in the 01-...
09:12
file
09:12
<bobby811>
i did
09:12
<Gadi>
and reboot the client
09:12
oh
09:12
<bobby811>
rebooting client
09:12
<Gadi>
ok
09:13
<bobby811>
now diff error
09:13
<Gadi>
pray tell
09:13
<bobby811>
running /linuxsrc
09:13
mouiting /proc
09:13
linuxsrc: installing b44 driver
09:13
modprobe b44
09:13
<Gadi>
so far so good
09:13shrek has quit IRC
09:14
<bobby811>
insmod /lib/modules/2.6.16.1-ltsp-2/kernel/drivers/net/b44.ko
09:14
running dhcpcd on port 67
09:14
error dhcpcd failed
09:14
kernel panic- not syncing attempted to kill init
09:15
<Gadi>
do you know what kind of broadcom card/chip it has?
09:15
<bobby811>
um
09:15
<Gadi>
broadcom has a few different drivers
09:15
you can try: tg3
09:15
<bobby811>
BCM5721
09:15
<Gadi>
ah, change b44 to tg3
09:15
I think
09:15
<bobby811>
k
09:16* Gadi faces the somber reality that he's been doing this way too long
09:16
<cyberorg>
no luck, can;t figure out, remote apps play fine, localapps not :(
09:16
<bobby811>
k rebooting client again
09:16
<Gadi>
cyberorg: try running: ltsp-localapps "padsp firefox"
09:16
(of course, you need pa-utils in the chroot)
09:17
<bobby811>
same error Gadi
09:17
dhcpd failed
09:17
gets to
09:17
running dhcpcd on port 67
09:17
then fails
09:19
<Gadi>
bobby811: according to this it is definitely tg3: http://www.broadcom.com/support/ethernet_nic/downloaddrivers.php
09:19
but, you may have old drivers
09:19
at this point, either replace the NIC, compile in the driver, or upgrade to LTSP5
09:19
<bobby811>
k how hard to upgrade to ltsp5
09:20
<Gadi>
easy
09:20
well, you would prolly be looking at a server upgrade to FC10
09:20
so, maybe not super easy
09:20
<bobby811>
oh
09:20
<Gadi>
can you upgrade your server?
09:20
<bobby811>
thats not an option here in class
09:20
as f10 not supported on these machines
09:20
<Gadi>
well, warren- what do you say to that?
09:20
:)
09:21
bobby811: is this a computer class?
09:21
<cyberorg>
Gadi, that didnt work
09:21
<bobby811>
yes
09:21
<Gadi>
bobby811: excellent! you get to learn about how the Linux kernel works :)
09:22cliebow has joined #ltsp
09:22
<Gadi>
(unless you take the easy way out and replace the network card in the client)
09:22
<warren>
Gadi: btw, how does the client know if a floppy is inserted?
09:23
I've never used that with localdevs
09:23
<cyberorg>
Gadi, i have xterm in client open via localapps
09:23
<Gadi>
warren: with floppies the drive is mounted
09:23
so, it accesses the media each time it is accessed
09:23
<warren>
bobby811: k12linux.org has a LiveUSB image, you can try Fedora 10 without installing it to the hard drive. You can even boot thin clients over the network from that LiveUSB demo mode.
09:23
<Gadi>
it does not know/care if the floppy is there or not until you try to look at it
09:24
<warren>
Gadi: but without a desktop icon that appears how do you know there is a floppy at all?
09:24
<Gadi>
there should be a desktop icon
09:24
<warren>
oh, if your client has a floppy the desktop icon ALWAYS appears?
09:24
<Gadi>
right
09:24
<warren>
oh neat
09:25
<Gadi>
because the *drive* is always there, not the media
09:25
it is different than CDs that way
09:26
cyberorg: do you have padsp on the client?
09:26
<warren>
nod
09:26
beware padsp
09:26
<cyberorg>
Gadi, yes
09:26
<warren>
oh wait
09:26
nevermind
09:26
<bobby811>
dunno if ltsp 5 will install on core 6
09:26
but will try
09:27
<warren>
bobby811: it wont
09:27
<bobby811>
oh
09:27
<warren>
bobby811: read k12linux.org
09:27
<bobby811>
guess this project needs scrapped
09:27
as its the final day for it
09:27
<warren>
bobby811: you can boot the LiveUSB on your server and see if it runs at all.
09:27
oh
09:27
too bad
09:28
<Gadi>
bobby811: all you need is a network card and a screwdriver
09:28
:)
09:28
<warren>
bobby811: what exactly is the project?A
09:28
<bobby811>
thin clients
09:28
<warren>
bobby811: K12Linux is SO EASY you could be booting thin clients within 15 minutes
09:28
<bobby811>
booting core6 from a 2nd workstation wioth no hdd
09:28
dont have any spare thumbdrives
09:29
<warren>
bobby811: can you burn a DVD? that works too.
09:29
<bobby811>
running from pxe
09:29
no burner
09:29
<warren>
I'm sorry then. It is too bad, you are very close.
09:29
<bobby811>
yea
09:29
oh well
09:29
teacher shoulda done his homework
09:29
<Gadi>
bobby811: can you pick a different client?
09:29
<bobby811>
nope
09:30
ill have teacher read this
09:30
<Gadi>
bobby811: wanna learn how to compile a driver?
09:30
<bobby811>
sure Gadi
09:30
<Gadi>
ok. take a deep breath
09:30
<warren>
Gadi: do you even know if b44 will work in that old kernel?
09:30
Gadi: doesn't he need the exact kernel source of that old kernel?
09:30
<Gadi>
warren: actually, he needs tg3
09:30
<warren>
oh
09:30
huh
09:30
that old kernel should have tg3
09:31
<Gadi>
warren: he just needs the headers
09:31
<warren>
maybe it isn't built into the image
09:31
<Gadi>
maybe
09:31
lets check the initramfs
09:31
bobby811: ready
09:31
<bobby811>
yep
09:31
<Gadi>
ok, lets make a working dir in /tmp
09:31
mkdir /tmp/work
09:31
<bobby811>
k
09:32
<Gadi>
if you are still in the pxelinuc.cfg dir, do: cp ../initrd* /tmp/work
09:32
<ltsppbot>
Someone pasted "localapps sound error" (6 lines) at http://pastebot.ltsp.org/233
09:32
<Gadi>
oops
09:32
cp ../initramfs* /tmp/work
09:32
<cyberorg>
Gadi, ^^ thats the error
09:33
<Gadi>
cyberorg: try: paplay
09:33
<cyberorg>
same plays fine from remote terminal
09:33
<Gadi>
instead of aplay
09:33
bobby811: still here?
09:33
<cyberorg>
Gadi, that works
09:34
<Gadi>
cyberorg: then, pulseaudio plumbing is fine
09:34
alsa redirection is not
09:36
cyberorg: try: pasuspender -- firefox
09:36
in the xterm
09:38
<cyberorg>
Gadi, sound server is not local, not suspending
09:38
<Gadi>
should be local
09:38
are you sure that is a local xterm?
09:38
<cyberorg>
PULSE_SERVER is tcp:127.0.0.1:4713
09:39
yes it is local :)
09:39
<Gadi>
ah
09:39
maybe it considers it not local because of the tcp:
09:39
nm
09:40
even if it did work, it wouldnt be a solution
09:40
you need to focus on getting the alsa redirection going
09:40
<cyberorg>
ok, where do i look for that?
09:40
<Gadi>
maybe talk to stgraber about how it works in ubuntu
09:40
you should have ~/.asound* files
09:41
<cyberorg>
no such files in user's home
09:43
<Gadi>
cyberorg: I have a .asoundrc with this line: </home/romm/.asoundrc.asoundconf>
09:43bobby-school has joined #ltsp
09:43
<bobby-school>
ok Gadi im backl
09:43
<Gadi>
and a .asoundrc.asoundconf with these:
09:43
pcm.!default { type pulse }
09:43
ctl.!default { type pulse }
09:44
bobby-school: hey, there
09:45
<bobby-school>
this is bobby811
09:45
school isp cutout for a few
09:47CAN-o-SPAM has joined #ltsp
09:47
<Gadi>
ok
09:47
<CAN-o-SPAM>
Gadi: get my e-mail?
09:47
<Gadi>
so, bobby-school, create /tmp/work and copy the initramfs file there
09:48
<bobby-school>
i did
09:48
<Gadi>
CAN-o-SPAM: I don't think so
09:48
lemme check spam folder in a sec
09:48
maybe it gobbled it up
09:48tjikkun_work has joined #ltsp
09:48
<Gadi>
bobby-school: cd /tmp/work
09:48
<bobby-school>
k
09:48
<Gadi>
then: file initramfs*
09:48
that will tell you the filetype
09:50bobby811 has quit IRC
09:53hanthana has quit IRC
09:53
<Gadi>
bobby-school: ?
09:54bobby-school has quit IRC
09:58bobby811 has joined #ltsp
09:58
<bobby811>
damn comcast
09:59
why in hell does a college use a shit connection
10:00
k ready Gadi
10:01
<warren>
Gadi: Ryan Niebur, Eric Harrison, Vagrant Cascadian, Gideon Romm, Oliver Grawert
10:01
Gadi: who else contributed significantly to LTSP upstream in the last 2 years?
10:01
(including Eric Harrison ceremonially)
10:01
<Gadi>
oh, and here I thought you were naming the fictional characters in your next novel
10:01
bobby811: cd /tmp/work
10:01hanthana has joined #ltsp
10:01
<Gadi>
file initramfs*
10:02
<bobby811>
k
10:02
<Gadi>
so, what filetype is it?
10:02
<bobby811>
.gz
10:03
can use gunzip filename.gz
10:03
<Gadi>
yup
10:03
then run file on it again and you'll see what it is uncompressed
10:03
<bobby811>
yep
10:03
now i see finitramfs file
10:04
<Gadi>
and what does file say?
10:04
<bobby811>
unreadable
10:04
15535 lines
10:04
unreadable code
10:04
<Gadi>
no
10:04
type: file initramfs*
10:04
<bobby811>
initramfs
10:05
initramfs: ASCII cpio archive (SVR4 with no CRC)
10:05
[root@localhost work]#
10:05
<Gadi>
there ya go
10:05
its a CPIO archive
10:05
you familiar with those?
10:05
<bobby811>
nah
10:05
<Gadi>
its like tar
10:05
<bobby811>
k
10:05
<Gadi>
so, we need to unpack it
10:05
<bobby811>
k
10:05
how?
10:06
<cyberorg>
Gadi, do you have .pulse folder?
10:06
<Gadi>
bobby811: cat initramfs*| cpio -i -d -H newc –no-absolute-filenames
10:07
that oughtta do it
10:07
<cyberorg>
it has stuff pointing to /tmp/pulse-XXXXX, which is broken on local term
10:07
<bobby811>
5731 blocks
10:07
[root@localhost work]# ls
10:07
initramfs
10:07
[root@localhost work]#
10:07
)
10:07
k i get this
10:08
<Gadi>
cyberorg: I have one but both files inside are empty
10:08shrek has joined #ltsp
10:08
<Gadi>
bobby811: no files?
10:08
<bobby811>
nope
10:09npman has joined #ltsp
10:09
<bobby811>
k i see stuff now
10:09
<cyberorg>
Gadi, 34973dd741cfeddab5dcb2b148845718:runtime -> /tmp/pulse-iLYLqq10A3ro
10:09
<bobby811>
[root@localhost work]# ls
10:09
bin etc initramfs mnt proc sys usr
10:09
dev init lib oldroot sbin tmp var
10:09
[root@localhost work]#
10:09
<Gadi>
bobby811: great!
10:10
<cyberorg>
that is in .pulse, that link is broken on the local shell as the /tmp/pulse-XXXX does not exist there, could that be the reason?
10:10
<Gadi>
the initramfs is a little filesystem that the kernel uses to set things up
10:10
before mounting the actual filesystem
10:10
<bobby811>
k
10:10
<Gadi>
now, do:
10:10
find lib/modules|grep tg3
10:10
and we will see if there is a tg3 kernel module in there
10:10
<bobby811>
[root@localhost work]# find lib/modules|grep tg3
10:10
lib/modules/2.6.16.1-ltsp-2/kernel/drivers/net/tg3.ko
10:10
[root@localhost work]#
10:10
<Gadi>
seems that there is
10:11
but, it must be old
10:11
or it must not recognize your card
10:11
so, that's what we will be replacing
10:11
<bobby811>
k
10:12
<Gadi>
to compile a new driver, you need two things:
10:12
1. driver code (found here: http://www.broadcom.com/support/ethernet_nic/netxtreme_server.php)
10:12
2. either the kernel source code for that kernel OR the kernel headers for that kernel
10:13
if FC6 is on the same kernel (2.6.16) you may be able to get #2 from the yum repository
10:13
go for the headers if/when available
10:13
<bobby811>
Linux localhost.localdomain 2.6.18-1.2798.fc6 #1 SMP Mon Oct 16 14:37:32 EDT 2006 i686 i686 i386 GNU/Linux
10:14
is the kernel
10:14
<warren>
would LTSP 4.2 use the same kernel as the host OS? probably not.
10:14
<Gadi>
maybe you can find the older kernel in the repository somewhere
10:14vvinet has joined #ltsp
10:14
<bobby811>
may just give up on it
10:14
or should we continue
10:14
which driver to dl
10:15
the linux 1?
10:15
<Gadi>
thats the spirit!
10:15
:)
10:16
hold on -maybe we can get something to help
10:16
<bobby811>
but it wont unzip
10:16
i tried
10:16
<Gadi>
http://wiki.ltsp.org/twiki/bin/view/Ltsp/CustomLtspKernels
10:17
http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42#ltsp_kernel_kit
10:17
there we go
10:17
grab that ltsp_kernel_kit
10:18
that should have what you need
10:18
<bobby811>
which 1
10:18
core4?
10:20
<Gadi>
ltsp_kernel_kit
10:20
<bobby811>
nah fuck it
10:20
class almost over
10:20
told instructor to scrap it'
10:21
<Gadi>
hehe
10:21
<bobby811>
class ends in 30mins
10:21
<Gadi>
fair 'nuff
10:21
<bobby811>
out of time
10:22
thanks for all the help though
10:22
<Gadi>
np
10:22
<cyberorg>
Gadi, i think that was it, deleting .pulse fixed it
10:22
<Gadi>
cyberorg: cool
10:22
everythings coming up Millhouse
10:23din_os has quit IRC
10:29
<cliebow>
as in Nixon?
10:29staffencasa has joined #ltsp
10:31
<polytan>
@+
10:31polytan has quit IRC
10:34alkisg has quit IRC
10:39alkisg has joined #ltsp
10:40spectra has joined #ltsp
10:53hanthana has quit IRC
10:53Blinny has joined #ltsp
10:56
<Blinny>
Can you specify scripts in /etc/X11/?? to run at user logout?
10:58
or is it in /opt/ltsp/.../usr/lib/ldm/rc.d/K##-scriptname ?
11:00pimpministerp has joined #ltsp
11:02tjikkun_work has quit IRC
11:10bobby811 has quit IRC
11:15primeministerp has quit IRC
11:18alkisg has quit IRC
11:20ogra has quit IRC
11:25Lns has joined #ltsp
11:27GandalfXx_ has quit IRC
11:47chrisinajar has quit IRC
11:55Yggdrasill_ has quit IRC
12:20shrek has quit IRC
12:29makghosh has joined #ltsp
12:36quickslip has joined #ltsp
12:36etyack has quit IRC
12:41wxchat2 has joined #ltsp
12:43makghosh_ has joined #ltsp
12:48vvinet has quit IRC
13:01alkisg has joined #ltsp
13:04etyack has joined #ltsp
13:06makghosh has quit IRC
13:09makghosh_ has quit IRC
13:18bobby_C has joined #ltsp
13:27gate_keeper_ has joined #ltsp
13:36primeministerp has joined #ltsp
13:38pimpministerp has quit IRC
13:40etyack has quit IRC
13:48spectra has quit IRC
13:50pasmen_ has joined #ltsp
13:51etyack has joined #ltsp
14:07vvinet has joined #ltsp
14:14pasmen_ has quit IRC
14:15CAN-o-SPAM has quit IRC
14:15CAN-o-SPAM has joined #ltsp
14:15vagrantc has joined #ltsp
14:20
<warren>
vagrantc: ping
14:20
vagrantc: you ran an entire desktop session as local app before? how?
14:21
<vagrantc>
warren: i made a session on the server that called ltsp-localapps with the --wait option ...
14:21
warren: i.e. ltsp-localapps --wait your-favorite-session
14:21
warren: i don't remember the exact syntax ... it was kind of a weird idea, in some regards... but maybe not as crazy as it sounds.
14:21
<warren>
oh
14:22npman has quit IRC
14:22
<vagrantc>
so i added code to ltsp-localappsd to check for the value of LTSP_COMMAND_WAIT xprop, and if it's set, rather than backgrounding the process, it waits until the process finishes...
14:23Merri has joined #ltsp
14:23
<Merri>
any fedora k12 ltsp
14:26Blinny has left #ltsp
14:26RobertLaptop has quit IRC
14:30RobertLaptop has joined #ltsp
14:31
<Merri>
Anyone on using ltsp in Active Directory
14:35
<warren>
vagrantc: when?
14:36
Merri: you should talk to k12osn list
14:36
Merri: people there use it
14:36chrisinajar has joined #ltsp
14:36
<_UsUrPeR_>
Using ubuntu 8.04, trying to figure out why an external WD USB drive won't mount on my client. According to dmesg it's being found, and assigned to /dev/sda, but it doesn't go past that. The user I have is currently in the Fuse group, and USB flash drives work fine on the same client. I have also added "LOCALDEV = True" to the lts.conf
14:37
<vagrantc>
warren: ltsp-trunk 888 and 889
14:37
warren: october
14:38Q-FUNK has joined #ltsp
14:38
<Q-FUNK>
hm
14:38
<alkisg>
_UsUrPeR_: maybe it's formatted with ntfs?
14:38
<_UsUrPeR_>
alkisg: ubuntu should be able to read/mount that though shouldn't it?
14:38* _UsUrPeR_ thinks it is formatted with ntfs
14:39
<alkisg>
I don't think ntfs-3g is in the chroot... I'm not sure, really.
14:39
<_UsUrPeR_>
hmm
14:39* _UsUrPeR_ checks the filesystem
14:40
<alkisg>
sudo chroot /opt/ltsp/i386 dpkg -l ntfs-3g => nothing
14:40warren has quit IRC
14:40
<Merri>
thanks
14:41
<_UsUrPeR_>
ok, from tty1 on the client, the following worked fine: mount /dev/sda1 /tmp/disk/
14:41Merri has left #ltsp
14:41
<_UsUrPeR_>
I am able to open the drive properly
14:43warren has joined #ltsp
14:43
<warren>
vagrantc: Gadi: Ryan52: hmmm, if we want to start non-ldm clients, currently it unconditionally turns stuff on like pulseaudio, jetpipe and stuff
14:44
perhaps those should be instead added to the startup scripts of screen sessions like ldm?
14:44
keep the function in ltsp-client, but move the calling command to ldm
14:44
<vagrantc>
perhaps
14:52six2one has quit IRC
14:55
<Gadi>
warren: jetpipe is useful for all clients
14:55
*sessions
14:55
pulseaudio can be, too
14:55
they are non-ldm specific services
14:56
<stgraber>
vagrantc, Gadi: Hey, just had a fun bug at the office. We all are in the same primary group and plugging an usb key creates a directory in /media/ with permissions 750 making everyone to see it, even if it can't read the content (as it's mounted using fuse)
14:56
<warren>
Gadi: if you're running GNOME as a local app, it wants to start its own pulseaudio
14:56
<stgraber>
how difficult would it be to force the permission to 700 after the directory is created ?
14:56
<vagrantc>
stgraber: very.
14:57
<stgraber>
really ?
14:57sepski has joined #ltsp
14:57
<Gadi>
gnome as a local app?
14:57
<warren>
stgraber: I thought we came up with a solution to that?
14:57
<Gadi>
dude, if you run gnome as a localapp set SOUND=False
14:57
:P
14:57
<stgraber>
warren: well, the bug just appeared again here with one of the latest snapshot from bzr so ...
14:57
<vagrantc>
stgraber: not dificult to do, just difficult to do without giving write permission to a directory in /media
14:57cliebow has quit IRC
14:57
<warren>
stgraber: anyhow, the fix I theorized earlier but didn't code was to make it 700, but use fsacl's to grant an individual user access.
14:57
<stgraber>
vagrantc: can't we just chmod 700 the directory we create in /media ?
14:58
<warren>
stgraber: oh!
14:58
stgraber: what version of gvfs?
14:58
<vagrantc>
stgraber: that was my proposal ages ago, but people didn't want to give write permission to directories in /media
14:58
<warren>
stgraber: gvfs needs to be patched to hide it
14:58
<stgraber>
warren: 1.0.2
14:58
<warren>
stgraber: there were two levels of the fix
14:58
gvfs-1.0.3 here
14:59
<stgraber>
vagrantc: well, permission is already 750 so 700 isn't worse :)
14:59
<warren>
I wanted to use fsacl's on the /media/* directories created by ltspfs, but there was concern that some distros don't support fsacls.
14:59
and it might have weird interactions with fuse
15:00
<vagrantc>
stgraber: i don't remember the exact issues, but there was *something* wrong with 700, and i gave up fighting.
15:00Patina has quit IRC
15:00Patina has joined #ltsp
15:01Gadi has left #ltsp
15:01
<warren>
I documented what was wrong with 700
15:01npman has joined #ltsp
15:01
<warren>
hmm
15:01
in one of my commits
15:02Q-FUNK has left #ltsp
15:02
<warren>
I can't find it in bzr log
15:03
<johnny>
blame warren
15:03
hehe
15:03
there are multiple distros that don't use a fuse group
15:03
gentoo and opensuse are two of them
15:04
<vagrantc>
stgraber: there were also discussions on the list about it
15:04
<johnny>
nor do they put users in their own group
15:04
<warren>
crap
15:04
I just booted the latest ltsp, and it failed to write the group for localapps use in /etc/groups
15:04
<johnny>
fail.
15:05
<warren>
wasn't there talk about this recently?
15:06
<vagrantc>
warren: is /etc writeable for you?
15:06
<warren>
vagrantc: it wrote to /etc/passwd but not /etc/groups
15:07
<vagrantc>
warren: if /etc/group is writeable, but /etc/ itself is not, it will fail. i had to patch for debian as we use bind-mounted files.
15:07
<warren>
vagrantc: oh damn
15:07
vagrantc: what is the patch?
15:07
<vagrantc>
warren: usermod writes /etc/group$someextension
15:07
<warren>
vagrantc: i wish I knew this before we did that major new release
15:07
vagrantc: you had to patch usermod?
15:07
<vagrantc>
the patch is ugly...
15:08
warren: i'll give you a link in a moment
15:08
<warren>
vagrantc: I wont be allowed to change usermod
15:09
<vagrantc>
warren: not a patch to usermod, just to ltsp code
15:09
<warren>
ok good
15:09
<vagrantc>
http://bazaar.launchpad.net/~vagrantc/ltsp/ltsp-debian-experimental-packaging/annotate/head%3A/patches/localapps-edit-group-directly.dpatch
15:11
all that functionality to replace stupidity in usermod ...
15:11
assuming a writeable /etc ...
15:11
<warren>
vagrantc: we need this permanently in ltsp-trunk
15:12
If we can think of a shell cmd to check if you are allowed to write in a directory, this can be automated.
15:13
<vagrantc>
"touch /etc/" seems to work.
15:13
<warren>
without writing
15:13
vagrantc: ah this works
15:13
vagrantc: [ -w /etc ] && echo writable
15:13
<vagrantc>
warren: no, that'll jest test if the filesystem actually *thinks* it's writeable.
15:14
you actually need to try to write to it.
15:14
<warren>
are you sure?
15:14
<vagrantc>
about 99.5%
15:14
<warren>
I just tested [ -w /etc ] on my client
15:14
vs real system
15:14
seems to work
15:15* vagrantc is skeptical
15:15
<vagrantc>
last i looked, it just checked filesystem permissions, but maybe it's smarter now ... testing if the filesystem is actually read-only.
15:16
<warren>
it works here
15:17
vagrantc: I'm adding your patch upstream with [ -w /etc ] as the conditional
15:18
<vagrantc>
warren: alright :)
15:18
if it works for me, so much the better, and if it doesn't, i'll just maintain a tiny patch for the test
15:19
<warren>
- myGroups=$(echo ${myGroups},${myGroups1} | sed -e 's/^,//g' -e 's/,$//g')
15:19
+ myGroups=$(echo ${myGroups},${myGroups1} | tr ',' '\n' | sort -u | tr '\n' ',' | sed -e 's/^,//g' -e 's/,$//g')
15:19
vagrantc: what does this do?
15:20
<vagrantc>
warren: removes duplicates, which actually matters if you're looping through them.
15:20
doesn't matter with usermod, but doesn't hurt either.
15:20
<warren>
vagrantc: how would that look to remove duplicates within the conditional?:
15:21
vagrantc: after myGroups=$(echo ${myGroups},${myGroups1} | sed -e 's/^,//g' -e 's/,$//g') had run
15:21
<vagrantc>
myGroups=$(echo ${myGroups} | tr ',' '\n' | sort -u | tr '\n' ',')
15:21
i think.
15:22
<warren>
i'll try that
15:23
<vagrantc>
that'll prevent the extra calls to tr and sort for those who are fortunate enough to not need them
15:24CAN-o-SPAM has quit IRC
15:24
<vagrantc>
warren: though i wonder if it wouldn't be good to get some patches upstream to usermod and other similar tools to not require a writeable /etc
15:24
<warren>
vagrantc: hmm
15:24CAN-o-SPAM has joined #ltsp
15:24
<warren>
vagrantc: I think readonly root is a dying breed, WAY too many little things to fix.
15:24
<vagrantc>
warren: it's quite possibly true ...
15:26
<warren>
OLDIFS="$IFS"
15:26
IFS=,
15:26
what does IFS do?
15:26
<vagrantc>
although it's no so insanely many ... having worked with read-only roots for so long ... i think it's actually a fairly small set of things
15:26
warren: it's what makes the for loop split on , instead of whitespace
15:27
<warren>
http://fpaste.org/paste/2777
15:27
i'm testing this.
15:27* vagrantc isn't a big fan of using IFS ...
15:27
<vagrantc>
but sometimes it's the only sane way of doing things...
15:28
warren: you're missing a "fi" i think.
15:28
<warren>
crap
15:30
hmm, that didn't work for one reason or another
15:30dirigeant has quit IRC
15:31
<vagrantc>
warren: you tested [ -w /etc/ ] on a read-only root ?
15:31
<warren>
yes
15:33
<alkisg>
vagrantc, you could leave the groups splitted by '\n' and use while read line instead of IFS...
15:33
<vagrantc>
warren: it's been working in debian lenny+experimental since january 9th, at least...
15:33
<alkisg>
*while read group
15:34
<vagrantc>
that would be nicer ...
15:34
<warren>
I just verified that [ -w /etc ] is owrking
15:34
the stuff in the conditoinal isn't
15:34Gadi_eeepc has joined #ltsp
15:34
<vagrantc>
"works for me"
15:35
<warren>
myGroups=$(echo ${myGroups} | tr ',' '\n' | sort -u | tr '\n' ',')
15:35
this is new
15:35
maybe it isn't right
15:36
<vagrantc>
oh, maybe that's actually an issue...
15:36
that's a little different...
15:36
<warren>
vagrantc: oh, something else is broken
15:36
myGroups contains:
15:37
test context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023,
15:37
<Gadi_eeepc>
lol
15:37
sorry
15:37
couldnt resist
15:38
<warren>
what is the format of what is supposed to be in that variable?
15:39
vagrantc: turns out it is broken in the part where it gets myGroups
15:39
<Gadi_eeepc>
warren: it should be a comma-delimited list of groups
15:39
that the user belongs to
15:41Egyptian[Home] has quit IRC
15:43alkisg has quit IRC
15:43
<warren>
warren@newcaprica etc]$ /usr/bin/getent group|egrep "[,:]warren(,|$)"|cut -d: -f1| tr '\n' ',' | sed -e 's/,$//g')
15:43
bash: syntax error near unexpected token `)
15:44
Gadi_eeepc: was this tested with bash?
15:44
<Gadi_eeepc>
warren: you have an extra parentheses at the end
15:44
take that off
15:44
<warren>
oops
15:44
/usr/bin/getent group|egrep "[,:]warren(,|$)"|cut -d: -f1| tr '\n' ',' | sed -e 's/,$//g'
15:44
this returns nothing
15:45
[warren@newcaprica etc]$ LANG=C /usr/bin/id | awk -F groups= '/groups=/{print $2}' | sed -e 's/[0-9]*(//g' -e 's/)//g'
15:45
warren context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023
15:45
Gadi_eeepc: is this supposed to print out only my group name?
15:45
Gadi_eeepc: and nothing else?
15:46
<Gadi_eeepc>
warren: yeah
15:46
well, no
15:46
all the roups you belong to
15:46
<vagrantc>
group names
15:46
<warren>
[warren@newcaprica etc]$ /usr/bin/id
15:46
uid=500(warren) gid=500(warren) groups=500(warren) context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023
15:46
<vagrantc>
apparently, works very differently on fedora...
15:46
doh.
15:46
<warren>
what does it look like on yours?
15:46
<Gadi_eeepc>
thats bizarre
15:47
<warren>
please show me the output?
15:47
<vagrantc>
uid=1000(vagrant) gid=1000(vagrant) groups=4(adm),20(dialout),24(cdrom),25(floppy),29(audio),44(video),46(plugdev),108(netdev),110(vboxusers),111(fuse),1000(vagrant)
15:47dirigeant has joined #ltsp
15:47
<vagrantc>
made the assumption that groups= was the last section...
15:48* Gadi_eeepc nods
15:48
<warren>
you need to look for groups= explicitly
15:48
<Gadi_eeepc>
yeah
15:49
<warren>
what's the best way?
15:49
<vagrantc>
it greatly simplified the code from using "groups" as it already included commas ...
15:50
<warren>
at least my groups= is the same format
15:50
<Gadi_eeepc>
id |awk '{print $3}'
15:50
<warren>
are we just going to trust that the third one is groups?
15:50
<Gadi_eeepc>
what does that give you, warren?
15:50Egyptian[Home] has joined #ltsp
15:50
<warren>
Gadi_eeepc: that works
15:50
<vagrantc>
works for me.
15:50
<Gadi_eeepc>
ok
15:50
<vagrantc>
that's, like, so much simpler.
15:51
hopefully we can rely on that.
15:51
we'll rely on it till we can't? :)
15:51
<warren>
sure
15:51
<Gadi_eeepc>
LANG=C /usr/bin/id | awk '{print $3}'|cut -d= -f2|sed -e 's/[0-9]*(//g' -e 's/)//g'
15:51
<warren>
[warren@newcaprica etc]$ LANG=C /usr/bin/id | awk '{print $3}'|cut -d= -f2|sed -e 's/[0-9]*(//g' -e 's/)//g'
15:51
warren
15:52
<Gadi_eeepc>
yeah
15:52
:)
15:52
<warren>
[test@newcaprica ~]$ LANG=C /usr/bin/id | awk '{print $3}'|cut -d= -f2|sed -e 's/[0-9]*(//g' -e 's/)//g'
15:52
mock,test
15:52
seems to work
15:52
I'll commit this.
15:52
<Gadi_eeepc>
cool
15:54
<warren>
http://fpaste.org/paste/2781
15:54
testing this now
15:54
suggest cleanups now
15:55
hm, that didn't work
15:57
<vagrantc>
warren: comment out 22, the myGroups de-dupe line
15:57
<warren>
vagrantc: the de-dupe line is leaving a comma at the end
15:58
<vagrantc>
then the user will get added to each group multiple times, if there are dupes, but i don't think that causes problems.
15:58
warren: of course.
15:58
:)
15:59
<Gadi_eeepc>
make it echo -n
15:59
in the de-dup line
15:59
<warren>
vagrantc: still failed to add anything to /etc/group
15:59
<vagrantc>
"echo -n" is not consistant across shells
15:59
<Gadi_eeepc>
really?
15:59
<warren>
really?
15:59
what shells?
15:59
<Gadi_eeepc>
hehe
16:00
<vagrantc>
bash behaves differently that dash, i think.
16:00
<Gadi_eeepc>
no way
16:00
<vagrantc>
or is it "echo -e"
16:00
<warren>
[warren@newcaprica ldm-rc.d]$ dash
16:00
$ echo -n "hello"
16:00
hello$ exit
16:00
[warren@newcaprica ldm-rc.d]$ echo -n "hello"
16:00
hello[warren@newcaprica ldm-rc.d]$
16:00
<vagrantc>
shells with an echo built-in vs. shells that fall back to /bin/echo
16:00* warren uses echo -n
16:00
<warren>
oh
16:01* Gadi_eeepc thinks you are thinking of echo -e
16:01
<vagrantc>
maybe
16:01
so many of these things...
16:01
<warren>
echo -n is in several other scripts
16:01
<Gadi_eeepc>
yeah
16:01
<warren>
including some of the debian specific scripts
16:01
<Gadi_eeepc>
warren: does adding -n fix the prob?
16:01
<warren>
Gadi_eeepc: no
16:01CAN-o-SPAM has quit IRC
16:02
<warren>
getting closer though
16:02BrunoXLambert has quit IRC
16:02
<warren>
hmm, echo -n didn't eliminate the trailing ,
16:02
<vagrantc>
the debian patch has a sed on the trailing de-dupe line ...
16:03npman has quit IRC
16:03
<warren>
I added sed 's/,$//'
16:03
that worked in removing the trailing , but stil didn't add to /etc/group
16:03
<Gadi_eeepc>
how does the de-dup line get rid of dups anyway?
16:04
seems it just sorts things
16:04
warren: I assume you do not have writable /etc, right?
16:04
<warren>
Gadi_eeepc: nod
16:05
Gadi_eeepc: hence we're adding the non writable /etc conditoinal
16:05
<vagrantc>
Gadi_eeepc: sorts duplicate groups from getent vs. id
16:06
i gotta run some errands ...
16:07
<warren>
oh
16:07
<Gadi_eeepc>
warren: and you don't see your user in the new /etc/group line?
16:07
<warren>
I don't have /var/cache/ltsp-localapps/
16:07
when was that added?
16:07
<Gadi_eeepc>
that been around for a while
16:07
<warren>
how long?
16:07
<Gadi_eeepc>
I think since we worte localapps
16:07
*wrote
16:08
<warren>
I had your earlier localapps code working here
16:08
and I never had that directory before
16:09
<Gadi_eeepc>
well, my earlier code added stuff to /etc/group and pruned dups
16:09
<warren>
in-place
16:09
?
16:09
<Gadi_eeepc>
copying only the pristine to var cache
16:09
<warren>
ooh
16:09
so my copy failed
16:09
<Gadi_eeepc>
and restoring pristine from var cache
16:09
<warren>
but it did succeed before
16:09
<Gadi_eeepc>
so, you prolly never had the pristine restored
16:09
:)
16:09
<warren>
that worked now
16:10
<Gadi_eeepc>
but, then vagrantc changed it so it doesnt edit /etc/group dirrectly
16:10
butt copies it to a working file
16:10
cool
16:10bobby_C has quit IRC
16:12
<warren>
Gadi_eeepc: what is supposed to restore it to pristine? I logged out and it is still there.
16:12
<Gadi_eeepc>
what is still there?
16:13RobertLaptop has quit IRC
16:13
<Gadi_eeepc>
it should restore the original /etc/group
16:13
<warren>
Gadi_eeepc: what script should restore it?
16:13
Gadi_eeepc: and when?
16:13
<Gadi_eeepc>
in the localapps scrript, there is a cleanup
16:13
the cleanup happens when localappsd dies
16:13
<warren>
oh
16:14
<Gadi_eeepc>
which should be when the Xserver restarts
16:14
does it not?
16:14
<warren>
nope
16:14
<Gadi_eeepc>
at one point vagrantc had said something about not seeing the stuff after ltsp-localappsd run at all
16:15
is localappsd still running?
16:15
even after logout?
16:15
<warren>
that's on which end?
16:15
running on client?
16:15
no
16:15
<Gadi_eeepc>
client
16:15
hmm...
16:15
oh...
16:16
I wonder if it is because the subshell is a child of the xinit process
16:16
and it dies along with X
16:16
hmm...
16:16
we need to get the subshell to dissociate then
16:16
<warren>
Gadi_eeepc: where is the code that runs when localappsd dies?
16:17
<Gadi_eeepc>
in the localapps screeen script
16:17
*screen
16:17
there is a subshell
16:17
in that code
16:17
that is backgrounded
16:17
<warren>
which is what?
16:17
<Gadi_eeepc>
what is what?
16:17
<warren>
what file is the local apps screen script?
16:17
<Gadi_eeepc>
ah
16:17
one sec
16:18
its in ldm/rc.d/
16:18
if you have the latest, it should be: X01-localapps or some such
16:18
the same script you were editing
16:18
just now
16:19
<warren>
Oh i see
16:19
yeah, it isn't getting there.
16:20
<vagrantc>
Gadi_eeepc: yeah, i still don't ever get *any* of the post-localappsd cleanup code executed
16:21
<warren>
# Clean up localapps menu
16:21
if boolean_is_true "$LOCAL_APPS_MENU" && [ -n "${TMP_XDG_MENU}" ]; then
16:21
ssh -S ${LDM_SOCKET} ${LDM_SERVER} rm -rf "${TMP_XDG_MENU}"
16:21
fi
16:21
Gadi_eeepc: how is this supposed to ssh if localappsd died only after logout?
16:21
vagrantc: I vote we not fix the cleanup yet, this is not a regression
16:22
<Gadi_eeepc>
1 sec - ph call
16:23
<vagrantc>
i've for the moment given up on fixing it ...
16:23
the post-localappsd cleanup code, that is
16:24
<Gadi_eeepc>
ah, I know what happened
16:24
<vagrantc>
i think most, if not all of the cleanup issues are double-checked in other ways, at least.
16:24
<warren>
I pushed my changes to that file
16:24
<Gadi_eeepc>
back before we ran ldm out of xinit, the rc.d files kept going
16:24
even after the Xserver dieed
16:24
because ldm (the parent) was still running
16:25
now that ldm is a child of xinit
16:25
and exits to let xinit exit
16:25
the rc.d scripts die with it
16:25
and dont continue to run
16:25
so we cannot rely on the same mechanism
16:26
<vagrantc>
sounds plausible
16:26
<warren>
what is /var/cache/localapps/ for? is that used anymore?
16:26
<Gadi_eeepc>
we need to run the cleanup as an ldm kill script
16:26
<warren>
I had to create this in ltsp-server sometime in the past
16:26
<Gadi_eeepc>
but, that doesnt address abrupt poweroffs
16:26
ie unclean logouts
16:27kaos01 has joined #ltsp
16:27
<vagrantc>
warren: i think early on in the localapps stuff it was renamed to ltsp-localapps ... to prevent namespace clashes.
16:27
<warren>
Gadi_eeepc: ldm startup can see the existence of pristine files and copy them over too
16:27
<Gadi_eeepc>
warren: right
16:27
only some things we cannot cleanup
16:27
<warren>
I want to release RIGHT NOW
16:27
<Gadi_eeepc>
like things involving ssh
16:27
<warren>
I just made the biggest release ever only to discover that localapps broke
16:28artista_frustrad has quit IRC
16:28
<Gadi_eeepc>
warren: if you want release this as-is
16:28
<warren>
/var/cache/localapps was never needed on the server side?
16:28
<Gadi_eeepc>
warren: no
16:28* warren wonders why he put it in ltsp-server
16:28* Gadi_eeepc shrugs
16:28sepski has quit IRC
16:28
<warren>
ok, /var/cache/ltsp-localapps/ right?
16:28
<Gadi_eeepc>
localapps requires no special server side stuff
16:28
<vagrantc>
in the client
16:28
<Gadi_eeepc>
right
16:29
only in client
16:29artista_frustrad has joined #ltsp
16:29
<johnny>
so.. i should be making /var/cache/ltsp-localapps in ltsp-client?
16:29
package*
16:29
<Gadi_eeepc>
johnny: yes
16:29gate_keeper_ has quit IRC
16:29
<Gadi_eeepc>
warren: if you want to tag what you have now,
16:29* johnny needs to track down why the ltsp-build-client creats /var/lib/nfs
16:29
<Gadi_eeepc>
I will work on cleaning up the cleanup code
16:30
<johnny>
maybe it's unnecessary..
16:30
<Gadi_eeepc>
and those changes can be tested and tagged later
16:30
<warren>
Gadi_eeepc: tagged and pushed
16:31
<Gadi_eeepc>
ok
16:31* Gadi_eeepc has more homework for tonight
16:31
<Gadi_eeepc>
:)
16:33nicoAMG has quit IRC
16:33
<warren>
Please review all the changes between .56 and .57 for sanity
16:34
vagrantc: could you please test 5.1.57 too since you're also read-only /etc
16:35
<vagrantc>
warren: can't test today, maybe tomorrow
16:35
<warren>
ok
16:35
well, this can't be worse I think
16:36epsas has joined #ltsp
16:36
<epsas>
hello peoples
16:37
<vagrantc>
epsas: hey!
16:37
<epsas>
I am the new network administrator at an office that runs entirely on LTSP
16:37
whoa! vagrant!
16:37
<vagrantc>
epsas: what distro?
16:37
<epsas>
CentOs
16:37
but we are investigating other platforms
16:38
<warren>
epsas: you might want to try k12linux.org
16:38
<epsas>
i guess i have questions about local sound / hardware
16:38
<vagrantc>
epsas: well, i'm biased with debian, of course :)
16:38
<epsas>
hi warren (togami i presume :)
16:38
<warren>
epsas: all that stuff works out of the box in k12linux.org
16:38
epsas: charles?
16:38
<epsas>
yup
16:38
<warren>
wow
16:38
epsas: where are you?
16:38
<epsas>
I am in Seattle
16:39
Moved from New Orleans to a place with more work... Seattle delivered
16:39
<warren>
New Orleans seems like not a great place for work
16:40
<epsas>
nope, not at all
16:40
it's good here - i have a full-time job with benefits (34 paid days off a year), and two side programming gigs already
16:40
within a month of getting here - so i am happy
16:40
<warren>
what kind of shop uses LTSP?
16:40
<epsas>
now I have to learn LTSP =)
16:40
The downtown emergency homeless shelter
16:41
<warren>
epsas: I'm about to spin a new version of k12linux.org
16:41
epsas: I could use testers of that new spin
16:41
epsas: you want 32bit or 64bit server?
16:41
<epsas>
the administrative office runs entirely off of ltsp, for windows there are a couple of win4lin instances
16:41
we could do both
16:41
one of our main concerns is unifying the filesystems between the 4 different LTSP servers we have
16:41
<warren>
you mean /home?
16:42
<epsas>
yeah
16:42
<warren>
you better run the same version of GNOME on all four LTSP servers
16:42
because it fucks up if you don't
16:42
<epsas>
we also need a network filesystem to seamlessly integrate remote file shares
16:42
hmm - okay
16:43
it looks like we are on CentOs and KDE at the moment
16:43
<warren>
epsas: LTSP on CentOS is very ancient
16:43
<Gadi_eeepc>
warren: can you test what I just pushed?
16:43
I fixed the cleanup code
16:43* warren looking
16:43
<Gadi_eeepc>
I think
16:43
<warren>
pushed where?
16:44
<Gadi_eeepc>
er, upstream
16:44
:)
16:44
after it was tagged
16:44
1166
16:44
<warren>
Gadi_eeepc: it only cleans up right during the next login?
16:45
<Gadi_eeepc>
well, there as an I* script to cleanup the things that can be cleaned up at init of ldm
16:45
(which is only /etc/group and friends and the cups client thing
16:45
then, there is a cleanup for after logout
16:45
for the things thatt require ssh
16:45
*that
16:45
<warren>
what kills localappsd now?
16:46
<Gadi_eeepc>
it dies along with X
16:46
<warren>
are you sure?
16:46
<Gadi_eeepc>
on logout
16:46
yup
16:46
just as before
16:46
<vagrantc>
i've never seen it survive when X dies.
16:46
<Gadi_eeepc>
it cannot
16:46
<warren>
ok, trying...
16:46
<Gadi_eeepc>
it is ldm's child
16:46
ill be back in 20 mins
16:46
gotta pick up one of my kids
16:47
<warren>
nod
16:47
<vagrantc>
we should move ldm's init thingy to X95, so we don't have to prepend all the remainging scripts with z
16:48
X99-run-x-session, that is
16:49
<warren>
well, at least most users wont notice that localapps is broken
16:49
vagrantc: just do it
16:49
<vagrantc>
i am
16:49
<warren>
shit
16:49
/etc/group failed again
16:50
oh crap
16:50
I forgot to add it to rwtab
16:50
since I was rushing
16:51
<vagrantc>
the hurrier we go...
16:52
<warren>
so might as well test gadi's cleanup
16:53etyack has quit IRC
16:53
<warren>
I don't see how this can possibly work.
16:53
but will test it anyway
16:54
<vagrantc>
i'm honestly amazed localapps works at all. it's crazy.
16:54
<warren>
If localappsd is killed by X dying, then we could use a similar ugly hack to what I added to screen-x-common
16:54
vagrantc: which script is sourcing these ldm-rc.d scripts?
16:54
<vagrantc>
/usr/share/ldm/ldm-script
16:55
<warren>
not the same script as screen-x-common or screen scripts right?
16:55
<vagrantc>
no
16:55
<warren>
if gadi's cleanup doesn't work (I expect it wont)
16:55
I have an idea
16:55
<vagrantc>
/usr/share/ldm/rc.d/* is handled by ldm itself.
16:56
we should probably just go ahead and use trap from ltsp-localappsd
16:58
<warren>
trap the kill you mean?
16:59
<vagrantc>
yeah
16:59
wouldn't that be more likely to catch when ltsp-localappsd has actually died?
16:59
<warren>
vagrantc: it wont be able to do the ssh part there.
17:00
<vagrantc>
well, sure.
17:00
but it can handle all the other bits, at least.
17:01
<warren>
yep, failed to cleanup
17:02RobertLaptop has joined #ltsp
17:08
<warren>
vagrantc: ah, I didn't copy the new scripts into my ltsp-client package
17:08
it does work
17:11
<vagrantc>
heh
17:11
<Gadi_eeepc>
hey, back - whatd I miss?
17:11
<vagrantc>
good to know :)
17:11* Gadi_eeepc scrolls
17:11
<warren>
Gadi_eeepc: it works, although it seems fragile to me
17:11
<Gadi_eeepc>
fragile how?
17:11
<warren>
Gadi_eeepc: I'm reordering the cleanup lines a little...
17:12
Gadi_eeepc: I moved the ssh parts after the local cleanup
17:12
<Gadi_eeepc>
okey dokey
17:12
<warren>
Gadi_eeepc: aren't you racing the teardown in this script though?
17:13
<Gadi_eeepc>
racing the teardown?
17:13
<warren>
Gadi_eeepc: I00-localapps-cleanup what defined LOCALAPPS_CACHE?
17:13
<Gadi_eeepc>
oh, did I not set that?
17:13
grr
17:14
sorry
17:14
can you grab the define for that from X01-localapps
17:14
?
17:14
I meant to copy/paste it in
17:14
<warren>
install -m 0755 localapps/ldm-rc.d/X01-localapps $RPM_BUILD_ROOT%{_datadir}/ldm/rc.d/
17:14
install -m 0755 localapps/ldm-rc.d/X99-zkill-localapps-menu $RPM_BUILD_ROOT%{_datadir}/ldm/rc.d/
17:15
hmm, does install accept wildcards?
17:15
Gadi_eeepc: ok
17:15
<Gadi_eeepc>
thx
17:17
warren: there should not be a race condition anywhere that I see
17:18
<warren>
Gadi_eeepc: oh right, we don't do kill -1 anymore?
17:18
<Gadi_eeepc>
right
17:18
so, everything should be executed
17:18
the only thing is if someone does ctrl-alt-backspace or turns the client off without logging out, there will be extra /tmp folders on the server from localapps
17:19
but, those should be cleaned out by the server over time
17:19
so, we should be ok
17:20
<warren>
ok, see any reason not to tag?
17:21
<Gadi_eeepc>
you added that LOCALAPSS_CACHE thing?
17:21
*LOCALAPPS_CACHE
17:21
:)
17:23
<warren>
# Copy back passwd and group
17:23
+ LOCALAPPS_CACHE=/var/cache/ltsp-localapps
17:23
for i in passwd group; do
17:23
[ -e "${LOCALAPPS_CACHE}/${i}" ] && cp "${LOCALAPPS_CACHE}/${i}" /etc/${i}
17:23
done
17:23
<Gadi_eeepc>
perfecto
17:24
if it works for you, go ahead annd tag
17:24
did vagrantc move any of the filenames?
17:24
to get rid of the "z"s
17:24
?
17:24
it may make it easier on the packagers if those name changes happen all at once
17:25RobertLaptop has quit IRC
17:25
<warren>
can you go ahead with the file renames? my brain hurts
17:25
<Gadi_eeepc>
np - but I dont want to mess up the packagers
17:26
especially since I am not one
17:26
maybe I'll defer to vagrantc
17:26
:)
17:26
<warren>
i'im tagginng now
17:26
gotta start hte long spin processs
17:27
ok?
17:27
<Gadi_eeepc>
go for it
17:28vvinet has quit IRC
17:29
<warren>
done
17:30
<vagrantc>
Gadi_eeepc: i didn't do the renames in ltsp-trunk yet, just ldm-trunk
17:31
it'll require a versioned dependency on ldm, to make sure the cleanup stuff works...
17:31
<warren>
ok, not yet
17:31dirigeant has quit IRC
17:32
<vagrantc>
keeps the most backwards-compatibility that way...
17:41chrisinajar has quit IRC
17:49warren has quit IRC
17:57dirigeant has joined #ltsp
18:06vagrantc has quit IRC
18:12
<ltsppbot>
"Lns" pasted "updating chroot causes proc madness" (44 lines) at http://pastebot.ltsp.org/234
18:12
<Lns>
If anyone has time to look at my function, which obviously messes my chroot up, that would be so cool. =)
18:13
(forgot to mention, in the function, when trying to umount proc, it says it's busy)
18:18nicoAMG has joined #ltsp
18:23vvinet has joined #ltsp
18:24
<Gadi_eeepc>
Lns: depending upon what you install and what it runs post-install, something may very well make use of /proc
18:25
also, you might as well unmoount proc after the upgrade
18:25
not when updating the kernels
18:25
<Lns>
Gadi_eeepc: ok.. so i'm wondering how to force the umount
18:25
without wreaking havock on my chroot that is..
18:26
<Gadi_eeepc>
well, you could umount -l
18:27
<Lns>
lets see when i do that and then ltsp-update-image
18:29
cool, no errors =)
18:29
thx Gadi_eeepc
18:31
<Gadi_eeepc>
np
18:32
btw, using [ -n "$()" ] as a test uses less memory than (( $() ))
18:32
as long as you are looking at the return on stdout
18:32
not that you may care
18:32
:)
18:33
<Lns>
not that i fully understand ;)
18:33
<Gadi_eeepc>
in your if statement, you use: if (( $(.....) )); then
18:33
<Lns>
thank you, i'll keep that in mind, honestly that (( $() )) thing was something i was just trying to get to work, so optimization was a secondary goal ;)
18:33
<Gadi_eeepc>
what that tests is the return value of a subshell
18:34
wwhere the last command in the subshell is your command
18:34
<Lns>
oh, so the [ -n "$()" ] won't launch a subshell.. ?
18:34
<Gadi_eeepc>
right, that collects the command stdout in a string
18:34
and looks to see if the string is empty
18:34
<Lns>
oh cool. i get it
18:35
thanks!
18:35
<Gadi_eeepc>
or, you can just do:
18:35
if blah; then
18:35
without the $() entirely
18:35
which would directly evaluate the return value
18:35
so, your current way uses 2 subshells
18:35
-n "$()" uses 1 subshell
18:36
and if blah; then uses no subshells
18:36
subshells cost memory (tho, admittedly not much)
18:36
<Lns>
but how do you form something (what i think is) as complex as that pipe of commands through a simple 'if' statement?
18:36
<Gadi_eeepc>
and additional processes
18:37
well, if you just want the return value and dont care about stdout, then you can just do;
18:37
if sudo chroot /opt/ltsp/i386 apt-get upgrade -s | grep upgraded >/dev/null 2>&1; then
18:38
the return value will be the return of the last process in the pipe
18:38
in this case, the grep
18:38
<Lns>
right
18:38
<Gadi_eeepc>
so, if the grep for upgraded returns true, then it was upgraded
18:38
else, it was not
18:39
or there was another error along the way
18:39
:)
18:39
<Lns>
hrm...well the grep returns either 0 or another number
18:39
<Gadi_eeepc>
right
18:39
if will look for a zero return
18:39
for true
18:39
<Lns>
oh, built into bash
18:40
<Gadi_eeepc>
right
18:40* Lns hasn't looked into that stuff yet ;)
18:40
<Gadi_eeepc>
the if statement is basically:
18:40
if [ this stuff returns true ]; then
18:40
regadless of what test it is
18:40
you could do:
18:40
if true; then echo yes; fi
18:41
and it will always echo yes
18:41
<Lns>
so builtin is 0 = true, 1 = false, 2+ = some crazy return value thats also not true
18:41
<Gadi_eeepc>
0 = true
18:41
!0 = false
18:41
<Lns>
ah
18:42
<Gadi_eeepc>
to evaluate return codes more carefully, simply run the command and check $?
18:42
<Lns>
So what makes "if true" true all the time?
18:42
<Gadi_eeepc>
so; if [ "$?" = 255 ]; then "The error was this"; fi
18:42
true just returns 0
18:42
ture = /bin/true
18:43
or /usr/bin/true
18:43
<Lns>
oh its a command.. true - do nothing, successfully. ha!
18:43
<Gadi_eeepc>
right
18:43
there's also a false
18:43
:)
18:43* Lns loves *nix
18:43
<Lns>
Can't really get more logical than that
18:44
i always thought someon's shell being /bin/false was just a placemarker for not launching a shell at all
18:44
not an actual command
18:45
<Gadi_eeepc>
programmers like to call real commands and not rely on the lack of a command's existance
18:45
:)
18:45
usually
18:45
:P
18:45
<Lns>
hehe
18:45chrisinajar has joined #ltsp
18:45
<Lns>
it's crazy to see the bigger picture (from my perspective anyway).. working in the shell really is almost like a programming environment in itself
18:47MRH2 has joined #ltsp
18:47* Gadi_eeepc goes to veg' a lil' before doing more work
18:47
<Gadi_eeepc>
l8r
18:48
<MRH2>
hi can anyone point me to workarounds for left handed mouse issues with 4.2
18:50
<Lns>
bye Gadi_eeepc , thanks again
18:50
<MRH2>
using gnome if itmakes a difference
18:56warren has joined #ltsp
18:59staffencasa has quit IRC
19:02Lns has quit IRC
19:06kaos01 has quit IRC
19:11kaos01 has joined #ltsp
19:27
<Ryan52>
warren, wtogami has edited the ldm-2.0.33-1.fc9,ltsp-5.1.58-1.fc9 update for Fedora 9 pending
19:27
what did you change?
19:28
<warren>
Ryan52: local apps was broken
19:28
Ryan52: I should have noticed this =(
19:28
respinning the ISO's now
19:28
<Ryan52>
oh.
19:28pmatulis has quit IRC
19:33
<warren>
Ryan52: http://alt.fedoraproject.org/pub/alt/ltsp/k12linux/f10/stable4/
19:33
I'm verifying that these ISO's work in every way...
19:33
Ryan52: can you download quickly?
19:33
<Ryan52>
k12linux-f10-i386-stable.iso 04-Feb-2009 01:22 813M
19:33
k12linux-f10-x86_64-stable.iso 04-Feb-2009 01:32 381M
19:33
why is the i386 so much bigger?
19:34
or does the x86_64 cd use a i386 chroot?
19:34
<warren>
Ryan52: uploading x86_64 now
19:34
<Ryan52>
oh :)
19:42
<warren>
oops, wrong file names, renaming....
19:50alekibango has quit IRC
19:51
<Ryan52>
but it's fine that I'm still downloading at the old link, right?
19:51
cause it's almost done ;)
19:52
<warren>
yeah
19:52
rename it to stable4
19:52kaos01 has quit IRC
19:52kaos01 has joined #ltsp
19:53* Ryan52 is at 96%
19:53
<Ryan52>
I think I need faster internet.
19:53
<warren>
you are downloading faster than me
20:04* Ryan52 booting it
20:04
<Ryan52>
what specifically should I test? do you have a checklist of ltsp (& friends) features that you usually test before a release?
20:06MRH2 has quit IRC
20:08
<kaos01>
can one have two installs like say /opt/ltsp-f10 and /opt/ltsp-f7 ?
20:13
<Ryan52>
warren, the email you just sent how has an error in the URL.
20:13
warren, Iceweasel can't find the server at delphi.bos.redhat.com.
20:15
<warren>
shit
20:16npman has joined #ltsp
20:21nicoAMG has quit IRC
20:36
<warren>
Ryan52: tried it?
20:38kaos01 has quit IRC
20:41
<Ryan52>
almost there..
20:44
the floppy disks that I rely on to boot my thin client are unreliable :)
20:44* Ryan52 is back to physical thin clients until he works out the two kvm guests thing
20:45
<warren>
you can get an old laptop from freegeek?
20:46
fine thin client tester
20:46
and quick
20:47
<Ryan52>
and small :). my thin client is horribly ugly looking...it's just the metal part of the case (with all of the plastic taken off), a mobo, a floppy disk drive lying on the bottom, and a power supply duck taped to the top. :p
20:48
*duct
20:51
hrm. the client says "nfsmount: error mounting /opt/ltsp/i386 on /sysroot as nfs: No such file or directory.".
20:52* Ryan52 wonders if that's a problem with his dhcpd configs, he isn't using the dhcp server on the live image..
20:52
<Ryan52>
I would, but it would break my network ... I *really* need to get this kvm thing figured out.
20:53npman has quit IRC
20:58
<warren>
I just tested the i386 version, both booting clients from live demo mode and after installing to hard drives.
20:58
installing x86_64 onto my usb stick now
21:09
Ryan52: brb, testing x86_64 version
21:09warren has quit IRC
21:14artista-frustrad has joined #ltsp
21:17RobertLaptop has joined #ltsp
21:18artista_frustrad has quit IRC
21:19warren has joined #ltsp
21:19
<warren>
Ryan52: x86_64 seems working
21:19
Ryan52: i'm putting stable4 on k12linux.org
21:19
<Ryan52>
okie doke.
21:24
<warren>
Ryan52: https://fedorahosted.org/k12linux/wiki/LiveServer updated page, please double check me
21:26
<Ryan52>
looks right.
21:40slashdotfx has joined #ltsp
21:46litlebuda has quit IRC
21:46litlebuda has joined #ltsp
22:02CaScAdE^FarAway has joined #ltsp
22:02litlebuda has quit IRC
22:07Gadi_eeepc1 has joined #ltsp
22:07Gadi_eeepc has quit IRC
22:08Gadi_eeepc has joined #ltsp
22:08Gadi_eeepc1 has quit IRC
22:19CaScAdE^1arAway has quit IRC
22:43nubae has joined #ltsp
22:46
<Gadi_eeepc>
nubae: ping
22:46
nubae: kick your log server when you get a chance
22:47Gadi_eeepc has left #ltsp
22:55ltsplogbot has joined #ltsp
22:56
<nubae>
bah.. logbot was killed by stupid godaddy sysadmins
23:00
test test test
23:01
<Ryan52>
The #LTSP IRC Logs
23:01
2009 MSGE000
23:01
0.o
23:02
<nubae>
bah not again...
23:02
<Ryan52>
nubae, is it closely attached to the server it's on? if not, you can move it to my vps, where the big bad godaddy sysadmins can't hurt it :). /me doesn't mind giving you an account
23:02
<nubae>
now I have to remember how I fixed that last time
23:02
<johnny>
Ryan52, what vps are you using?
23:02
<nubae>
it might be an idea...
23:02
<Ryan52>
johnny, slicehost
23:02
<johnny>
not linode?
23:02
how come?
23:03
i wonder how slicehost is doing these days
23:03
<Ryan52>
johnny, they were bought by rackspace, so they have lots of moolah to do new cool things...they got a new datacenter today :)
23:04
<johnny>
yet they are somewhat more expensive than linode
23:04
<Ryan52>
ya.
23:04
<johnny>
i'm spending $30.. and have a little less
23:04
than their $38 plan
23:05
<Ryan52>
johnny, I know, but is linode good?
23:05
slicehost has been *really* good.
23:05
<johnny>
yes
23:05
<Ryan52>
only downtime was because of ddos attacks, which were resolved in 10 minutes or so.
23:05
<johnny>
had my linode for a year now :)
23:06
i wonder how their admin panel compares
23:06
it doesn't say whether they host dns for ya or not
23:06
do they?
23:06
<Ryan52>
slicehost does dns.
23:06
<johnny>
good
23:06
do they generate zone files for ya?
23:06
<Ryan52>
no.
23:06
<johnny>
ones you can reuse elsewhere?
23:06
that is nice feature of linode dns manager
23:06
<Ryan52>
do you have an API, tho?
23:07
<johnny>
hmm.. yes, gimme link
23:07
<Ryan52>
we can use active resource to do all kinds of cool stuff.
23:07ltsplogbot has quit IRC
23:07
<johnny>
to slicehost api
23:07
i'll compare it
23:07
<Ryan52>
k, one sec.
23:07ltsplogbot has joined #ltsp
23:08
<Ryan52>
johnny: http://articles.slicehost.com/2008/5/13/slicemanager-api-documentation
23:08
<nubae>
ahhh thats better
23:08
<johnny>
seriously? pdf?
23:08
lol
23:09
api documentation in a pdf is not a good idea
23:09
<Ryan52>
hrm..."Febriary"
23:09
<nubae>
lots 3 days though
23:09
hehe, yeah that was the russian coders logs2html
23:09
and its compiled C
23:10
so not much I can do... but I can try
23:10* Ryan52 wonders if you can just edit binaries to change strings
23:11
<johnny>
ok, slice's api is a bit better
23:11
but, does slicehost let you get slices for even just a few minutes?
23:11
i can't tell from the information provided
23:11
it sounds like not, but it is kind of ambiguous
23:12
<Ryan52>
what do you mean for just a few minutes?
23:12
<johnny>
provison the linode.. do some intensive task.. destroy it
23:13
<Ryan52>
ya, you can do that.
23:13
<johnny>
word
23:13
it seems that for both of these, hard disk is the most expensivep art
23:13
i'm looking for a secure backup solution
23:14
altho.. i'd prefer not to use amazon or google
23:14
<Ryan52>
ya.
23:14
why not google or amazon?
23:15
if you want secure, you should be gpg encrypting it anyway ;)
23:15
<johnny>
i don't trust either for a second
23:15
too big
23:16
they are so automated, there is almost zero real customer service
23:16
which is surely why they can offer it so cheap
23:19wwx has quit IRC
23:21
<warren>
amazon is really expensive
23:21
<johnny>
warren, any suggestions?
23:22
<warren>
for what?
23:22
<johnny>
i've got less than a 100gs of real data
23:22
important stuff
23:22
<warren>
what is your goal?
23:22
<johnny>
secure data backup
23:23
more like a virtual hard drive
23:23
so i can access my stuff anywhere
23:23
altho.. i'd rather just put my stuff
23:23
in some friendly data center
23:23
but i can't do that yet
23:23
<warren>
Ryan52: many host providers don't allow any IRC of any kind because it is too often associated with DDoS attacks
23:25
<johnny>
the atlanta datacenter filters 6667
23:25
for linode
23:25
the rest seem fine
23:25
<Ryan52>
slicehost lets us do whatever we want to, as long as we don't break the law.
23:27
<warren>
yeah, I heard good things about slicehost.
23:30
<johnny>
warren, but all vps charge too much for disk space
23:30
<warren>
is it really too much?
23:30
johnny: why do you need 100GB?
23:30
johnny: Google is launching their own online drive service soon
23:30
<johnny>
sure.. but i do not want to use google services
23:31
too automated.. no real customer service
23:31
<warren>
johnny: just encrypt what you put there.
23:31
oh
23:31
yeah, but they dont tend to screw things up.
23:31
<johnny>
yeah.. for the current gtalk issues
23:31
except for*
23:31
and last year when they lost a bunch of people's mail
23:31
not permanently
23:31
but still..
23:32
google's support stuff is like a blackhole sometimes
23:32
plus they are generally too big
23:32
i don't really need to support something like that
23:32
and their standards support seems kinda weak in general..
23:33
like imap
23:33
and xmpp
23:34
google is no saint
23:38ltsplogbot has left #ltsp
23:39ltsplogbot has joined #ltsp
23:39ScorpKing has joined #ltsp
23:41
<nubae>
and we have liftoff
23:42
proper weekday names and month names...
23:44
yeah go daddy doesnt alllow irc of any kind, let alone bots...
23:44
but the are too dumb to stop people using other ports... go figure..
23:48
<johnny>
no l7 protocol filters
23:49
they are somewhat expensive to ru
23:49
run*
23:52alkisg has joined #ltsp