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


Channel log from 31 October 2008   (all times are UTC)

00:01
<Ryan52>
warren: I don't have the kvm stuff set up yet, but once I do I know I'm gonna be wondering this. how do I do switch between "SCREEN"s (as lts.conf calls it)
00:02
<warren>
Ryan52: ctrl-alt-2 for kvm's debug console
00:02
Ryan52: sendkeys ctrl-alt-f2
00:02
ctrl-alt-1 to get back to the real console
00:03
<Ryan52>
okay, thanks.
00:10indradg has quit IRC
00:20indradg has joined #ltsp
00:41
<Ryan52>
warren: this may sound really stupid, but how do I get out of caps lock mode? :P
00:41
I don't even know how I got into it. I don't have a caps lock key. (it's a control key; I use emacs :))
00:42
nevermind, I figured it out.
01:00alkisg has joined #ltsp
01:53mistik1 has quit IRC
02:00mistik1 has joined #ltsp
02:10alkisg has left #ltsp
02:45indradg has quit IRC
02:58indradg has joined #ltsp
03:34indradg_ has joined #ltsp
03:41indradg has quit IRC
03:56Q-FUNK has joined #ltsp
04:01tjikkun_work has joined #ltsp
04:08tjikkun_work has quit IRC
04:26Egyptian[Home] has joined #ltsp
05:01Nubae has joined #ltsp
05:04Q-FUNK has quit IRC
05:06mikkel has joined #ltsp
05:27highvoltage has quit IRC
05:27highvoltage has joined #ltsp
05:29cyberorg has joined #ltsp
05:32highvoltage has quit IRC
05:33highvoltage has joined #ltsp
05:35gate_keeper_ has joined #ltsp
05:39highvoltage has quit IRC
05:40highvoltage has joined #ltsp
05:48Nubae has left #ltsp
06:14pmatulis has joined #ltsp
06:16Guaraldo has joined #ltsp
06:20Guaraldo has quit IRC
06:24Guaraldo has joined #ltsp
06:27exodos has joined #ltsp
06:30Guaraldo has quit IRC
06:30Guaraldo has joined #ltsp
06:33Guaraldo has joined #ltsp
06:50warren has quit IRC
07:00jammcq has quit IRC
07:13gate_keeper_ has quit IRC
07:15Guaraldo has quit IRC
07:19yanu has quit IRC
07:19yanu has joined #ltsp
07:19Q-FUNK has joined #ltsp
07:25Guaraldo has joined #ltsp
07:42six2one has joined #ltsp
07:59chrisinajar|home is now known as chrisinajar
08:09rahuld has joined #ltsp
08:10
<rahuld>
hi
08:11
can someone point me to a link so that I can authenticate TC in Kiwi-LTSP against OPenLdap
08:16Q-FUNK has quit IRC
08:20Gadi has joined #ltsp
08:27warren has joined #ltsp
08:30
<Gadi>
warren: morning
08:30
<warren>
morning
08:30
<Gadi>
you must be feeling better - you've been very busy hacking
08:30
:)
08:30
<warren>
The latest changes to Xsession in ldm really slow down the login
08:31
either that, or moving the setxkbmap stuff into a shell script did it...
08:31
not sure
08:31
<Gadi>
easy thing to test
08:33bjohnson has joined #ltsp
08:37
<ogra>
we should start working on integrating with hal instead of using setxkbmap
08:37
<warren>
ogra: which would mean the old config options would no longer be valid?
08:37
<ogra>
adapting the hal keymaping tools not so hard
08:38
no, they would be used as well
08:38
in fact they would override hal's defaults
08:39
<cyberorg>
rahuld, if your server can authenticate against ldap, TCs will do that automatically
08:39
<ltsppbot>
"ogra" pasted "hal keymap script" (23 lines) at http://pastebot.ltsp.org/82
08:39
<ogra>
warren, have a look
08:39
<cyberorg>
use yast2 ldap module to set up server to authenticate against ldap server
08:39
<bjohnson>
Hi. I've got a KDE thin client running off a ltsp 4.2 server (Centos 5.2). Within the last couple of weeks Firefox starts to drag on a forum that I frequent. I usually have lots of tabs open but the slow computer responsiveness continues even when I close all other tabs. From trials, I think the problem might be from large graphic files (I've turned of flash). Both the server load (CPU and RAM) is low and running "free" on the client shows 45
08:39
0MB of 500MB free. I wonder if the problem is LAN related but don't know what else to check.
08:39
<ogra>
we just need to call that script before X starts up
08:39generalsnus has joined #ltsp
08:40
<ogra>
with soe distro specific adaption indeed
08:40
err, without i mean
08:40
. /etc/default/console-setup > /dev/null 2>&1 || exit 0
08:40
<bjohnson>
I can't think of what might have schanged on the server that coincides with the slowdown (it seemed to work fine before). The same forum site on other machines (non-thin client) seems to be fine
08:40
<ogra>
needs to be replaced with a check if the variables are set
08:41
(and needs to fall back to a set of defaults ... i.e. "us", "pc105")
08:43
warren, its dar easy and should be a lot faster than setxkbmap
08:43
*darn
08:43
<warren>
did you test it?
08:43
<ogra>
it is what ubuntu uses by default
08:44
and i adapted it to touchscreens with some modification
08:44
i know it will work, gimme a week and i'll have something committable
08:45
just needs some variable checks at the top
08:45
i know fedora uses the same
08:46
<generalsnus>
ok, so i have installed edubuntu 8.10 in vmware, with 2 nic's 1 for thinclients and 1 for internet ... the thinclients work ok, but i am not getting any internet connection. ifconfig tells me that i have IP 192.168.0.254 on both nic's , how can i change one to dhcp? in the network manager it only says "ifupdown"
08:47rahuld has quit IRC
08:47
<ogra>
generalsnus, look in /etc/network/interfaces
08:49
<generalsnus>
pimary net interface: auto eth0
08:49cyberorg has quit IRC
09:03jammcq has joined #ltsp
09:03
<jammcq>
g'morning friends
09:04
<warren>
hmm
09:04
with the latest ldm and ltsp, I have a weird problem where GNOME startup gets stuck part way
09:04
I see a background, but no window manager
09:04
it just gets stuck
09:05
<ogra>
anything in ~/.xsession-errors ?
09:05
<johnny>
warren, that happens to me sometimes
09:05
i have to pkill -u $USER
09:05
usually ends up being some stuck process or other
09:06
warren, this happens with the one released with hardy (so.. 6 months ago)
09:06
and the one before that :)
09:07
<warren>
ogra: complaints about scim
09:07
<ogra>
shouldnt anymore with intrepid, gnome got a lot more evil in killing everything with a sledgehammer this release
09:07
warren, hmm, try uninstaling scim and check if it works then
09:09Armagnac has joined #ltsp
09:09
<johnny>
ogra!
09:09
you survived!
09:10
<Armagnac>
Hi all
09:10
can anybody show me how to allow read/write access to an NBD LTSP image
09:10
???
09:10Eeyore-Jr has joined #ltsp
09:11alkisg has joined #ltsp
09:13
<ogra>
Armagnac, that wont work with squashfs images, they are readonly by design
09:13
<Armagnac>
of course, i dont remember that fact
09:14
what can I do instead?
09:15
<sbalneav>
Morning all
09:15
<ogra>
there is an option to nbd-server that creates a writable overlay, but that will be gone after reboot of the client
09:16
<warren>
hmm, it isn't scim
09:16
<ogra>
another option would be to use ext3 images instaed of squashfs, but that requires tons of extra diskspace and you would have to hack up ltsp-update-image
09:16
<sbalneav>
ogra: Hey, so I'm dressed up here at work with my gasmask for halloween, which is german. What does "Zivilschutzfilter" translate to? Civil something filter?
09:16
<ogra>
warren, is debus running ?
09:16
*dbus
09:16pmatulis_t60 has joined #ltsp
09:17
<warren>
ogra: which dbus?
09:17
and where?
09:17
<ogra>
sbalneav, "Civil Protection Filter"
09:17
sbalneav, doesnt really make sense as a word though
09:17
<sbalneav>
So it's a police gas mask then
09:17
Not a military model
09:17
<ogra>
is it really written like that ?
09:17
<sbalneav>
Yup, all one work
09:17
word
09:17
<ogra>
weird
09:18
<warren>
oh boy. the budget for me to go to BTS is gone.
09:18
<ogra>
but well, its german
09:18
<warren>
i'm going anyway
09:18
<ogra>
warren, the session dbus ... does it get started properly ?
09:18
<warren>
what is the process name?
09:19
<ogra>
dbus-launch i think
09:19
or dbus-daemon (should have --session somewhere) ... thats the child
09:21
<warren>
session meaning on the terminal server side?
09:22
<ogra>
right
09:22
gnome-settings-daemon needs it nowadays
09:23
<jammcq>
warren: glad you are coming, without budget. i think i've got the big dinner covered, and the the workshop covered
09:24
<warren>
i'll drive straight in one go, no need to stop part-way
09:24
<ogra>
jammcq, make sure to have the big dinner the first day then, so he can put some fat on to live from it the next days :)
09:24
<Armagnac>
ok, I have been reading about NBD
09:25
it is posible to change the ltsp-update-image script
09:25
to make ext2 o ext3 filesystems instead of squashfs???
09:25
<johnny>
why exactly did you want to change it..
09:26
debian has a plugin to make ext3, you could find their changes compared to ubuntu and integrate them...
09:26
but i don't see why it matters, since you can't save the changes anyways
09:26
<Armagnac>
which debian package? ltstp-server???
09:26cliebow has joined #ltsp
09:26
<Armagnac>
why?
09:27
<ogra>
ubuntus ltsp-server package installs the debian plugins alongside
09:27
you should just need to copy it
09:27
or even link it
09:27
<warren>
hmm, KDE and XFCE are fine
09:27
<Armagnac>
ok, what's the name of the plugin?
09:27* ogra didnt know there was an ext3 debian plugin
09:27
<warren>
this is a GNOME specific problem that began between ltsp-5.1.26 and ltsp-5.1.30
09:28
<Armagnac>
I will try to change the script from mksquashfs to mkext3fs
09:28
<ogra>
ogra@osiris:~/Devel/branches/kourou/kourou$ apt-cache show ltsp-server|grep Version
09:28
Version: 5.1.29-0ubuntu3
09:28
warren, works here with .29
09:29
warren, might be related to the fact that you use a new gdm infrastructure
09:29
<warren>
this is still F9
09:30
unchanged from prior ltsp
09:30
<ogra>
weird
09:31
<johnny>
Armagnac, why are you trying to do this??/
09:31
<Armagnac>
look
09:31
i am using fat clients
09:31
<johnny>
even if it's nbd with ext3, the changes won't be saved
09:32
afaict
09:32
<Armagnac>
they are working fine
09:32
with NBD
09:32
but I need at least one client to have read-write access over some specific image
09:32
<johnny>
even if they don't get saved?
09:32
<Armagnac>
let say with other nbdport=2001
09:32
well
09:33
the problem is with alsa
09:33
<johnny>
what problem is that?
09:33
<Armagnac>
i am trying to compile latest alsa packages
09:34
I think that if I have a client with read/write access
09:34
<johnny>
then you could do it on the server
09:34
in the chroot
09:34
<Armagnac>
really??
09:34
<johnny>
yes
09:34
and then ltsp-update-image
09:34
<Armagnac>
didnt alsa checks for existing hardware and such
09:34
<johnny>
no
09:34
<jammcq>
warren: i'm gonna pick up a couple access points for BTS. you had some comments earlier about what a good setup would be. care to talk about it?
09:34
<Armagnac>
modprobe the modules and such
09:34
<johnny>
no
09:34
<ogra>
not if you build a proper package
09:34
<johnny>
it doesn't
09:35
<Armagnac>
I see
09:35
I need to try it
09:35
<ogra>
is that hardy or intrepid ?
09:35
<Armagnac>
hardy still
09:35
<johnny>
intrepid might have the fixes you need
09:35
<Armagnac>
anything better for intrepid??
09:35
yes
09:35
<johnny>
sure.. whatevers newest for alsa
09:35
<ogra>
well, i'D suggest grabbing the source package of alsa from intrepid then
09:35
<johnny>
since hardy :)
09:35
<Armagnac>
how should I upgrade the chroot
09:35
<warren>
jammcq: two WRT54GL's setup in WDS repeater mode should theoretically work
09:35
<ogra>
and just roll that
09:35
<johnny>
Armagnac, the same way you upgrade any apt distro
09:36
except one command to tell it to use teh chroot
09:36
<Armagnac>
well bandwith is a problem here
09:36
<jammcq>
ogra: btw, I loaded intrepid on my laptop last night. it gives me the same problems that hardy did. messed up screen sometimes and trackpoint mouse nub not able to move windows around
09:36
<warren>
jammcq: although it was just barely enough range last time, so we might need one side with a directed antenna instead of the omni it comes with
09:36
<johnny>
jammcq, broken hardware? :)
09:36
<jammcq>
warren: I figured I'd get the longer antennas
09:36
<warren>
longer doesn't help
09:36bjohnson has quit IRC
09:36
<johnny>
jammcq, might wanna try it in some other distro cd
09:36
<warren>
you need a directed antenna
09:36
<jammcq>
johnny: not likely, my daughter has the exact same model a same problems
09:36
<johnny>
oh
09:36
<ogra>
jammcq, intresting ... what graphics HW was that (note that both, amd and nvidia didnt make proper drivers in time for xorg 1.5)
09:37
<johnny>
jammcq, since you have so many, you could give me one :)
09:37
<jammcq>
intell everything
09:37
<Armagnac>
just to make sure i understand
09:37
<johnny>
hehe
09:37
<jammcq>
and I'd expect the intel drivers to be some of the best
09:37
<Armagnac>
can I chroot into the /opt/ltsp/fati386 folder
09:37
<johnny>
warren, GIVE ME FEDORA 10!
09:37
<ogra>
jammcq, hrm ... i have a couple if intel devices and no prob at all on them
09:37
<Armagnac>
put a ubuntu 8.10 CD in the server
09:37
<johnny>
Armagnac, yes, but i'd suggest mounting /proc first in there
09:37
<warren>
johnny: you can have it for free
09:37
<Armagnac>
of course
09:37
and sys also
09:37
<ogra>
johnny, just not on CD :P
09:38
<jammcq>
it's hard to even describe the problems i'm having. I should just fly out to UDS in mountainview in december and show you guys :)
09:38
<johnny>
lol
09:38
<ogra>
jammcq, good idea !!!
09:38
<Armagnac>
whats the command to do that
09:38
<ogra>
bring scotte along :)
09:38
*scottie even
09:38
<Armagnac>
i know there is some CD scripts to do that
09:38
<jammcq>
no biggie, prolly only cost 600 or $700
09:38
<warren>
jammcq: how are things going there with the economy in MI?
09:38
<johnny>
jammcq, or take the train out that way :)
09:38
<ogra>
warren, houses are chep :P
09:38
*cheap
09:39
<jammcq>
warren: well, my business is as busy as it could possbly be. we'll prolly be adding another programmer soon. everybody else around me is dying
09:39
<Armagnac>
sh /cdrom/cdromupgrade
09:39
I see
09:39
<johnny>
ogra, huh?
09:39
<Armagnac>
I should test it in a diferent server not a production one
09:39
<ogra>
johnny, ?
09:39
what ?
09:39
<warren>
jammcq: any new industries moving in? I'd think somebody could use educated workers in a low cost of living area.
09:39
<johnny>
you said "just not on cd"
09:39tarzeau has quit IRC
09:40
<johnny>
i didn't know what you meant
09:40
<ogra>
johnny, yeah, no shipit for fedora ... they didnt ask us yet to distribute their CDs :)
09:40
<johnny>
ogra, fedora 10 isn't out yet
09:40
<jammcq>
warren: I wish. we're not at the bottom around here yet. I expect a big announcement about GM/Chrysler soon that should put another 25,000 people out in the street
09:41
<warren>
Or China buys the entire state
09:41
"We had all this excess US dollars"
09:41
<ogra>
yeah, in germany the car manufacturers stopped the production over christmas
09:41
<johnny>
warren, i'm thinking about trying out fedora 10 on my laptop.. WHEN IT COMES OUT
09:41
<ogra>
thats likely to coset a lot of jobs as well
09:42Armagnac has quit IRC
09:42
<jammcq>
at this point michigan owes more on the mortgage than it appraises for. China could buy Michigan, but we'd all have to add money to the deal :)
09:42
<warren>
johnny: PR is due out real soon, that is a good starting point
09:42
johnny: almost identical to final
09:42
<johnny>
how soon is real soon?
09:42
<warren>
a week or so?
09:42
I dunno
09:42
<ogra>
a wekk ?
09:42
final is in two iirc
09:42
<johnny>
i want something that will let me stay up to date relatively easily..
09:43
to newer packages
09:43
that's why i like gentoo.. none of this release nonsense..
09:43
i just upgrade ..
09:44
<jammcq>
wow, those wrt54g's are getting cheap. $45 each
09:44* warren downgraded to ltsp-5.1.26
09:44
<jammcq>
does it need to be a 'L' ?
09:44
<warren>
jammcq: yes
09:44
jammcq: non L are crap
09:44Guaraldo has quit IRC
09:44
<johnny>
jammcq, yes..
09:44
<Eeyore-Jr>
L is for linux
09:44
<jammcq>
ah, that's $49
09:44
<warren>
really? where?
09:44
<Eeyore-Jr>
then you put tomatoes on it
09:44
<johnny>
OH NO!!!!
09:44
<warren>
I paid $69 last time
09:44
<Eeyore-Jr>
and it works great
09:44
<jammcq>
$49.99 on amazon
09:44
<warren>
crap
09:44
it is something in ltsp-client
09:44
<Eeyore-Jr>
i get mine for free :)
09:45
<jammcq>
and you are syaing the 'Linksys High Gain Antenna Kit' won't help?
09:45
<warren>
jammcq: URL?
09:45
jammcq: are we leaving this at the motel?
09:45
<|Paradox|>
http://www.dailytech.com/Microsoft+Launches+Windows+Azure+New+Free+OS/article13301.htm <----looks like M$ is taking LTSP a step further :)
09:45
<jammcq>
http://www.linksys.com/servlet/Satellite?c=L_CASupport_C2&childpagename=US%2FLayout&cid=1169671130843&pagename=Linksys%2FCommon%2FVisitorWrapper&lid=JavaScript-1169671130843
09:46
I wasn't planning on leaving them
09:46
<johnny>
he was planning on giving them to me :)
09:46* johnny hypnotizes jammcq
09:47
<jammcq>
whatever I get, I pretty much need to order it today
09:47
<warren>
jammcq: I don't know about this antenna
09:47
<jammcq>
I see "high gain" and I think "longer range than normal"
09:47
where would I find a directional antenna?
09:48
<warren>
pringles can and wire
09:48
let me ask my radio friend
09:48
hold
09:49
<ogra>
jammcq, look for yagi
09:49
you need two of them though
09:49
<warren>
oh?
09:49
why two?
09:49Q-FUNK has joined #ltsp
09:49
<ogra>
they center on a beam
09:50
the beams shoud set up to match each other
09:50
*should be
09:50
that gets you a proper directional connection up to 500m
09:50
<jammcq>
ok, so 2 of these routers will get IP into the workshop. what do we do with it once it's in the workshop?
09:50Q-FUNK has quit IRC
09:50
<warren>
jammcq: you unplug one of the two antennas, point directional at each other, and turn off diversity mode
09:50
<jammcq>
seems like we need another AP
09:51Q-FUNK has joined #ltsp
09:51
<warren>
jammcq: i'll bring one of mine for the workshop
09:51
<jammcq>
point WHAT directional at each other?
09:51
one of the existing antennas?
09:51
<ogra>
you have two APs
09:51
<warren>
Four AP's...
09:51
<ogra>
each has one of the yagis
09:51
<jammcq>
ok, where do I get the Yagi's?
09:51
that's the question
09:51
<warren>
His existing one -> WIRE -> AP pointing at workshop -> WIRELESS -> AP pointing at office -> WIRE -> AP for workshop
09:51
<ogra>
you point the yagis at each other (bring a laserpointer, that helps with adjusting)
09:52
right, the yagis make up the "WIRELESS" part in the above scheme
09:52
<warren>
It *MIGHT* work without directional antenna, then we don't need the fourth AP
09:52
<jammcq>
and... I'm working with a very limited budget here
09:52
<warren>
but last year the signal was right at the edge
09:53
<ogra>
well, then build the yagis your own :)
09:53
<jammcq>
oh that's right, my time isn't worth anything
09:53
<warren>
jammcq: I can bring two WRT54GL's if they're temporary
09:53
<ogra>
the pringles setup is essentially a cheapo yagi
09:54
<warren>
http://www.netscum.com/~clapp/wireless.html
09:54
<jammcq>
at $50/ea, I don't mind buying a pair, but if we get into fancy antenna's and such, i'm worried the price will go up quite a bit
09:54
<ogra>
http://www.google.com/products/catalog?ie=UTF-8&oe=UTF-8&sourceid=navclient&gfns=1&q=yagi+antenna&um=1&cid=12698771761281978610&sa=X&oi=product_catalog_result&resnum=4&ct=result#ps-sellers
09:54
<jammcq>
heh
09:54
<ogra>
$194 each here
09:54
<jammcq>
yeah, 2 of those plus the routers, we're at $500 and I haven't paid for shipping yet
09:55
<warren>
we don't want to run a CAT5e across the telephone pole?
09:55
<johnny>
regex peoplez
09:55
what does this mean to you?
09:55
/([^.]+)$
09:55
i'm confused about the part in the []
09:55
<jammcq>
1 or more NOT dot characters at the end of the line
09:56
<johnny>
hmm.. really.. you don't have to escape the . ?
09:56
<jammcq>
^ inside [] means NOT the next char
09:56
no escaping when inside the []
09:56
<johnny>
aha..
09:56
sweet
09:56
<ogra>
HAHA
09:56
http://blog.makezine.com/archive/2006/08/diy_usb_wifi_antenna_made_from.html
09:56
<warren>
but . means any character
09:56
<johnny>
warren, thus the question :)
09:56
<ogra>
try that one, cheap and quikly put together
09:56
<warren>
[^.] means not any character?
09:56
<johnny>
warren, jammcq says no escaping in there :)
09:57
<jammcq>
normally '.' means any character, except when inside the []
09:57
ogra: that's awesome. you can read your email and cook your dinner with the same piece of hardware
09:57
<ogra>
lol
09:58
<warren>
jammcq: I have three WRT54GL's, so I might as well bring them if they are temporary
09:58
<ogra>
as long as its pasta
09:58
<_UsUrPeR_>
in Fedora, does adding "get-lease-hostnames true;" to dhcpd.conf do anything?
09:58
<warren>
jammcq: although I'm wondering if it is really a bad idea to just run CAT5e across the telephone pole
09:58
<jammcq>
warren: if you wanna bring them, that makes it much easier on me. then I don't have to worry about what to tell the customs folks as I enter canada
09:58
<warren>
jammcq: k, will do
09:58
<jammcq>
warren: if there's lightening anywhere in the area, then it's a REALLY bad idea
09:59* jammcq speaks from past experience on this
09:59
<Gadi>
yeah - and don't lick anything
09:59
<warren>
too bad those fiber to ethernet adapters are so expensive
09:59
<jammcq>
yeah, fiber would be awesome
09:59
i've always heard fiber is good for the diet
09:59
<johnny>
warren, get fios.. they give you free adaptor :)
09:59
hehe
09:59
<ogra>
Gadi, why are you alsways taking out all the excitement
09:59
:P
09:59
<Gadi>
sorry - lick away
09:59
<johnny>
lol
09:59
<ogra>
heh
10:00otavio has quit IRC
10:00* ogra is sooo missing to be there
10:00
<jammcq>
ogra: we are SOooo gonna miss having you there
10:00
<Gadi>
ogra: we'll do skype as a localapp and talk
10:00
;)
10:00
<ogra>
sniff
10:00
<Q-FUNK>
localtalk
10:00
<ogra>
ekiga ftw
10:00
<johnny>
and then Q-FUNK shows up :(
10:00
<warren>
ogra isn't coming?
10:01
<johnny>
hehe
10:01
<ogra>
warren, nope, some private issues that dont let me go
10:01
<Q-FUNK>
johnny: hm?
10:01
<johnny>
Q-FUNK, not serious.. :)
10:01
<jammcq>
ogra: just don't forget about Brazil next June
10:02
<ogra>
yeah, i surely wont
10:02
<warren>
Brazil is not looking likely =(
10:03
for me
10:03
<jammcq>
warren: think positive
10:03
<johnny>
i'll kidnap warren
10:03
err devnap
10:03
<jammcq>
having a fedora rep at FISL-10 would be very good
10:03
<Gadi>
jammcq: sounds like he's not positive :P
10:03
<warren>
that reminds me, time to renew passport
10:03
<johnny>
yeah.. me too..
10:03
<warren>
my passport photo is me at 17 years old
10:03
<johnny>
lol
10:03
you probably look the same
10:04
<Gadi>
warren: just because you're 18 now... ;)
10:04
<ogra>
18 1/2 !
10:04
<warren>
dog years
10:04
<jammcq>
how come nobody ever talks about Cat years?
10:05
<johnny>
cuz cats are the devil
10:05
<ogra>
they have seven lifes anyway
10:05
<johnny>
9 ogra
10:05
9
10:05
<jammcq>
ogra's talking metric
10:05
<ogra>
heh
10:05
<johnny>
metric lives?
10:05cliebow has quit IRC
10:05
<jammcq>
yeah, multiply by 9, divide by 5, add 32
10:15bobby_C has joined #ltsp
10:17
<sbalneav>
Mmmm
10:17
beer
10:21Guaraldo has joined #ltsp
10:22vagrantc has joined #ltsp
10:22wpgmb has joined #ltsp
10:23staffencasa has joined #ltsp
10:24
<wpgmb>
I have one winxp system on the ltsp client side, It's getting an ip via DHCP from the LTSP server, but it cannot get out to the 'net. "Regular" ltsp clients have no problem. What do I need to change?
10:24Q-FUNK has quit IRC
10:27
<sbalneav>
wpgmb: You've still got the linksys or dlink router?
10:27
As I explained before, here's your problem:
10:28
<XP box and terminals> -> <switch> -> <ltsp server> -> <router> -> <internet>
10:28
<wpgmb>
actually: Internet > IPCop firewall > LTSP server > clients, one being a winxp box that's not a real client
10:28
yes, forgot the switch.
10:29
<sbalneav>
the TERMINALS get out because they've got their processes running on the LTSP server, which is connected to the firewall.
10:29
So, like I said before you either need to:
10:29
set up routing/forwarding/firewalling on the linux box (complicated)
10:29
OR
10:29
<wpgmb>
I understand that. Btw - the older issue has been resolved, this time, my kids want to be able to use ipod stuff from a winbox
10:30CAN-o-SPAM has joined #ltsp
10:30
<sbalneav>
simply move the XP box on the "internet" side of the LTSP server.
10:30
i.e.
10:30
<ipcop> -> <hub/switch> -> <ltsp server>
10:31
<warren>
uh... doesn't he only need iptables NAT rules on the LTSP server?
10:31
<sbalneav>
+
10:31
-> xp box
10:31
warren: sure.
10:31
<wpgmb>
physically impossible without major recabling. Three systems (2 ltsp clients, 1 winxp) plug into a switch on the 2nd floor of my house, and from there one line goes down to a another switch.
10:31
<sbalneav>
What's easier? Setting up IP nat rules, or moving a network connectio to a different hub? :)
10:32
<wpgmb>
see above...
10:32
<sbalneav>
Split the cable.
10:33
<wpgmb>
like, how?
10:33
<sbalneav>
How's the cable run? Does it have punchdowns at both ends, or just a cable hanging out of the wall with an end on it?
10:34
<wpgmb>
2nd flr house: 2 ltsp clients, 1 winxp box > switch in closet on 2nd flr. > cable to switch in basement > cable to ltsp box
10:34
<warren>
dude
10:34
the NAT rules are trivial
10:34
<alkisg>
iptables -t nat -A POSTROUTING -j MASQUERADE
10:34
echo 1 > /proc/sys/net/ipv4/ip_forward
10:34
:)
10:35
<wpgmb>
alkisg: that won't foobar ltsp clients access to the 'net?
10:35
<alkisg>
Yes, they
10:36
<_UsUrPeR_>
I'm having problems getting hostnames to show up on clients in Fedora
10:36
<alkisg>
'll be able to access the net throught NAT
10:36
More here: http://billauer.co.il/ipmasq-html.html
10:36
<ogra>
well, ubuntu uses ufw nowadays ... if thats ubuntu, use that instead
10:36
<wpgmb>
ogra: yes, Ubuntu 8.04 ltsp server
10:36
<ogra>
http://doc.ubuntu.com/ubuntu/serverguide/C/firewall.html
10:36
<_UsUrPeR_>
I have get-lease-hostnames true; in the dhcpd.conf... am I missing something?
10:37
<sbalneav>
The intenet isn't connected to the ltsp box directly, guys, it's already going through another firewall.
10:39
<wpgmb>
sbalneav is correct
10:39
<warren>
sbalneav: two layers of NAT works just fine for most things
10:39
I've used 3 layers of NAT for a few months once
10:40
it was fine except when one of the wireless repeaters would die
10:40
<sbalneav>
Breaks on a lot of things too.
10:40
<warren>
That's why I have 3 WRT54GL's... I was avoiding paying cable bills for a few months...
10:40
<wpgmb>
I seem to have read somewhere ufw isn't all that great, and not very stable?
10:40
<sbalneav>
Sources, please.
10:40
<ogra>
wpgmb, its the default FW software since hardy and developed by our security team ...
10:41
in intrepid is has a gui as well ...
10:41
<wpgmb>
ogra: I'll look into it in more detail later. thx for the link
10:41
<sbalneav>
All it is is just a frontend to iptables.
10:42
Me, personally, I'd just split the line. But whatever.
10:42
<ogra>
its limited in its function on purpose though ... if you want it massively powerful, go with iptables ... for 90% of the usecases it will be fine though
10:42
<wpgmb>
sbalneav: I like simple, but I don't really see how I can "split" the line
10:44
anyway - g2g, took the day off to get the yard ready for tonight.... better get to it, before I don't get to share in the loot tonight :-D
10:45wpgmb has left #ltsp
10:45
<sbalneav>
http://www.instructables.com/id/How-to-make-your-own-Ethernet-"splitter"/
10:45
Cripes, can't people hang around for two seconds?!
10:46
<alkisg>
Heh
10:46
<Appiah>
:)
10:46
<sbalneav>
I was going to tell him you can BUY PRE MADE SPLITTERS here in town at Active Electronics for like $15
10:47
Whatever.
10:48
<alkisg>
Ugh... gnome needs a better gpaint-like program... I hate to install all the kde*stuff just to get the (excellent) kolourpaint for the students... :(
10:48
<sbalneav>
When he pops in here 2 weeks from now asking the same question again (like he has for the past couple of months) I ain't bothering to answer.
10:48
<ogra>
alkisg, get touchscreens, use xournal :P
10:49
<alkisg>
ogra, I'll be very glad if I can convince the principal to give me money for structured cabling (is that how it's called?) this year... all the ftp cables are on the floor!!!
10:50
<ogra>
get a security guy to review that and let *him* convince the principal
10:58
<Appiah>
or just get another guy to pose as a security guy
10:58
costs less
10:58
<alkisg>
:)
10:58
Anyway... another thing: Wouldn't it be better if all the lts.conf directives were included (but commented out) in the default lts.conf, so that the ltsp admins just un-comment the ones they want? Something like http://pyramid.metrix.net/svn/Pyramid/dist/etc/dnsmasq.conf.example for dnsmasq? I can help on that if you like (gathering all directives from source etc), but my English is not so good for the comments...
10:59
<Appiah>
it should alkisg
10:59
I was thinking the same
10:59
or a lts.conf.example included
10:59
<alkisg>
...and as a side effect, we would know which directives are supported in each version ('cause they change... :))
10:59mccann has joined #ltsp
11:01
<sbalneav>
Appiah: Care to code one up?
11:02
I'm sure if you did it, we'd include it :)
11:02twinprism has quit IRC
11:04
<johnny>
alkisg, i suggested linking to specific bzr revisions of that file
11:04
per release
11:06twinprism has joined #ltsp
11:06
<alkisg>
johnny that would be great... But when the programmers change some directive, will they take the time to update lts.conf.example? I know that many programmers don't like documentation... :)
11:06
<johnny>
that's the point
11:06
link directly to the lts-parameters.txt
11:06
which is what people are supposed to update when they add new directives
11:07
instead of having a seperate copy
11:07Guaraldo has quit IRC
11:07
<johnny>
we already do have lts.conf.example.. it's called lts-parameters.txt :)
11:07Guaraldo has joined #ltsp
11:08
<johnny>
and in intrepid and other distro releases since hardy, lts-parameters is now in the server, and not in the ltsp chroot
11:08
so it is easier to find
11:08
<alkisg>
Wow... I hadn't see that
11:09
<johnny>
it's existed for a long time
11:09
there was a period where it was not accurate tho
11:09
but folks are much better at updating it now :)
11:09
it is not accurate in hardy, that is for sure
11:10
but will be for intrepid
11:10
<alkisg>
I think it's ok on intrepid
11:10
<johnny>
yes
11:10
<alkisg>
E.g. I was trying LDM_ALLOW_GUEST for some time until I found out about LDM_GUESTLOGIN...
11:10
<sbalneav>
One of the things I want to get fixed in the documentation is to import lts-parameters into the doco.
11:11
at the hackfest, I'm going to spend some time whipping that into some finalized, organized shape. I'd love for us to have an ltsp-server-docs package that could be installed.
11:13elisboa_ has joined #ltsp
11:13
<Gadi>
please to be updating the agenda on bts2008 wiki page
11:14
<alkisg>
There are some parameters missing from lts-parameters.txt though... E.g. LDM_SERVER, LDM_LANGUAGE, LDM_SESSION etc...
11:19
<johnny>
sbalneav, DONT IMPORT IT
11:19
link it!
11:19
link the raw bzr file
11:20
that way you can link per distro release
11:20
since debian packages supported different parameters than ubuntu, than fedora, than gentoo
11:20
etc
11:23
<vagrantc>
there are different supported lts.conf parameters?
11:23
ugh.
11:23
<johnny>
vagrantc, .. obviously..
11:23
hardy didn't have the same parameters that i or fedora had, since we were alot newer
11:23
<vagrantc>
that seems like a disaster.
11:23
oh, i see.
11:23
<johnny>
and you were putting out newer packages for debian
11:24
<vagrantc>
it's not so much they have different parameters, as there are different versions available
11:24
which have different parameters
11:24
<johnny>
yes
11:24Stonekeeper has joined #ltsp
11:24
<johnny>
like the timeout thing.. didn't exist for hardy. which alot of people will still be running
11:24
and the guest thing changed
11:25
etc
11:25
<vagrantc>
yes/
11:25
<Stonekeeper>
hi. I've been searching in vain to find out why my ldap-authed users, when pressing the logout button, have to wait around 60s for the logout panel to show (gnome). Anyone seen this? Local users are instant.
11:26
<Appiah>
sbalneav: not a bad idea
11:27
I could actully make an lts.conf.example , and show it and if everyone likes it
11:27
we can submit it?
11:27
<ogra>
well, that was the intention behind the doc team :)
11:27
<Appiah>
how do you submit stuff anyways?
11:27
<ogra>
Appiah, joun the doc tem
11:27
*team
11:27
<Appiah>
I just recently started coming here
11:27
<ogra>
https://launchpad.net/~ltsp-docwriters
11:29
<Appiah>
I'll look into it
11:29
<johnny>
Appiah, we don't need it.?
11:29
lts-parameters is the example ?
11:29
<warren>
ogra: hm, the problem is somewhere in ltsp-client
11:29
<Appiah>
johnny: there is one already?
11:29
<johnny>
yes
11:29
<warren>
changes between 5.1.26 and 5.1.30
11:29
<Appiah>
where is the file located johnny ?
11:29
<johnny>
it's in the chroot in hardy, but now is in the server on intrepid, and any releases per distro
11:29
since that time period
11:29
<warren>
KDE and XFCE have no problem, only GNOME
11:29
<ogra>
warren, well, all i can say i dont see issues in intrepid
11:29
<Appiah>
johnny: oh I c
11:30
no need for me to do it then!
11:30
<alkisg>
johnny, an lts.conf.example like dnsmasq.conf.example would be much better in my opinion (having the lts.conf format, not a tabular one, and also more comments for each parameter)
11:30
<Appiah>
I'll bet I find something else to contribute with
11:30
<johnny>
alkisg, just change that one
11:30
<ogra>
Appiah, well, see what sbalneav said above, he'd like to have it properly in the docs :)
11:30
<johnny>
use comment characters
11:30
<ogra>
so you could help with that
11:30
<Appiah>
oh sorry I was away and didnt read up
11:31
<johnny>
alkisg, most things should work out of the box, so if people start copying the example file.. i think there will be trouble
11:31
<ogra>
yeah
11:31
<alkisg>
johnny, but if all parameters are commented out, where's the trouble?
11:31
<ogra>
lts.conf should be a last resort only to override the autodeection
11:31
<Appiah>
and that should say in the lts.conf.example (if there should be one)
11:31
<ogra>
there is a reason we dont even have it existing by default :)
11:32
it does
11:32
<johnny>
alkisg, also.. i don't even support all the parameters
11:32
like any of the xorg ones
11:32Stonekeeper has left #ltsp
11:32
<alkisg>
x failing, load balancing, guest logins with specific usernames/passwords... There are many occations where I needed to modify it
11:32
<ogra>
alkisg, yes, but you shouldnt use/need it to just get up the clients
11:33
<Appiah>
wait does ltsp support load balacing yet?
11:33
<ogra>
if X is failing thats a bug in X that needs fixage
11:33
<johnny>
alkisg, none of the x parameters except X_CONF will work for me
11:33
<ogra>
i'm merely upset about people just suggesting to use lts.conf and not to file a bug against the X server
11:33
<alkisg>
ogra, I'm not saying it should go to /var/lib/tftpboot/..., but if it was *somewhere* (even on a wiki), I'd like it to be in an lts.conf format, not in tabular format
11:34
johnny, yeah, it should be distro specific...
11:37
(or each parameter should be followed by a "Supported in: Debian >= xxx, Ubuntu >= xxx, Fedora >= xxx" etc line)
11:38
<sbalneav>
ogra: Heh, I'm not even a member of the "docwriters-team", just sent in an application
11:38
<ogra>
sbalneav, you are
11:38
ltsp-upstream is a member
11:39
<sbalneav>
Ah, guess I'm an "indirect" member.
11:39
<ogra>
right
11:39
<alkisg>
ogra, and all the members have access to a specific wiki or something?
11:40
<ogra>
warren, all i see beyond your own fedora commits that could affect you are commits 899 and 900
11:40
alkisg, no, the the bzr tree holding the upstream documentation
11:40
team membership is the ACL for the tree
11:41
<Appiah>
is the ltsp teams listed anywhere?
11:41
<alkisg>
I see... ty
11:41
<ogra>
Appiah, there are only two
11:41
<Appiah>
ltsp and ltsp-docs ?
11:41
<ogra>
ltsp-upstream and ltsp-docwriters
11:41
<Appiah>
https://launchpad.net/ltsp is this upstream?
11:42
<ogra>
thats the upstream product
11:42
<warren>
what is the difference between LTSP_COMMAND and LTSP_COMMAND_WAIT?
11:42
<ogra>
the team is https://launchpad.net/~ltsp-upstream/
11:42
the documentation team is https://launchpad.net/~ltsp-docwriters
11:43
warren, ask vagrantc, his commit
11:43
<warren>
vagrantc: ^^
11:44otavio has joined #ltsp
11:45
<ogra>
Appiah, while we try to keep the ltsp-upstream team small so there is a contolling person for each distro in place that channels code commits, ltsp-docwriters is open for everyone who wants to help out
11:46F-GT has quit IRC
11:46F-GT has joined #ltsp
11:46
<Appiah>
I c
11:47
<vagrantc>
warren: LTSP_COMMAND is the actual command, and LTSP_COMMAND_WAIT=true will pause the ltsp-localappsd rather than continue to wait for more LTSP_COMMAND values ... i implemented it to run the whole session as a localapp.
11:47
<warren>
oh
11:47
<ogra>
sbalneav, i'm evil .... rejected :)
11:47
<Appiah>
I'll jump on ltsp-docwriters someday then
11:47
<warren>
vagrantc: that seems like a really roundabout way to run the entire session as localapp
11:48
<vagrantc>
warren: maybe so, but it works.
11:48
warren: keeps the authentication simple
11:48
<ogra>
sbalneav, btw, what made you think you are no team member, you committed to the branch for at least the last month regulary :)
11:50
<sbalneav>
When I go to the page, it says I'm not a member :)
11:50spectra has joined #ltsp
11:50
<ogra>
oh, and i see you didnt commit directly but through nubae
11:51
sbalneav, https://code.edge.launchpad.net/~ltsp-docwriters/ltsp/ltsp-docs-trunk does that offer you a push location ?
11:51
<sbalneav>
I'm always nervous about updating to any trunk. I screwed you guys up one time, now I'm gunshy
11:51
<ogra>
pfft
11:51
nothing we cant roll back
11:52
<sbalneav>
No, no push location
11:52
wait
11:52
<ogra>
sbalneav, i drove vagrantc and warren insane already with my broken tagging ... and my head still stits on my sholders
11:52
<sbalneav>
yes, duh
11:52
<ogra>
ah :)
11:52
so all fine then
11:52
<vagrantc>
messing with revision history is kind of annoying ... but if you set append_revisions_only=true in ~/.bazaar/bazaar.conf it should prevent that from happening.
11:53
<sbalneav>
sbalneav@feniks:~$ cat ~/.bazaar/bazaar.conf
11:53
email=Scott Balneaves <sbalneav@ltsp.org>
11:53
append_revisions_only = True
11:53
launchpad_username = sbalneav
11:53
I'm a good little doobee
11:54
<ogra>
better than me :)
11:54
so dont be a sissy ;)
11:55
the whole point of revision control is that you can roll back :)
11:59Lns has joined #ltsp
12:05
<warren>
you can always commit to your own tree
12:05
look at it
12:05
before you push
12:06K_O-Gnom has joined #ltsp
12:15
<vagrantc>
warren: the LTSP_COMMAND_WAIT thing is a bit roundabout, but the current implementation immediately moves on to look for more LTSP_COMMAND's to execute ... i couldn't figure out any simpler way to implement it.
12:15K_O-Gnom has quit IRC
12:15
<warren>
vagrantc: something between 5.1.26 and 5.1.30 is causing trouble for GNOME here...
12:16
login works, but GNOME startup gets stuck before the window manager
12:16
<vagrantc>
hm.
12:16
<warren>
if I downgrade to ltsp-client-5.1.26 the problem goes away
12:18
<vagrantc>
warren: the only one that you didn't commit specifically for fedora that would likely affect you is 900: the module-suspend-on-idle pulseaudio thing ...
12:21DonSilver has joined #ltsp
12:23cliebow has joined #ltsp
12:23generalsnus has quit IRC
12:29
<warren>
vagrantc: yikes
12:30
vagrantc: the "Add module-suspend-on-idle to the list of PulseAudio modules (so running audacity as localapp works)" broke it
12:31
stgraber: ogra: module-suspend-on-idle broke it
12:32
<_UsUrPeR_>
!pastebot
12:32
<ltspbot`>
_UsUrPeR_: "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.
12:33
<warren>
stgraber: ogra: vagrantc: On F9 it prevents GNOME session from starting. On F19 GNOME session starts, but pulseaudio daemon did not successfully run
12:34
<_UsUrPeR_>
could someone take a look at the following files and tell me what I am missing? I am having trouble assigning hostnames to clients, and I believe everything has been entered correctly
12:34
<ltsppbot>
"_UsUrPeR_" pasted "/etc/ltsp/dhcpd.conf" (62 lines) at http://pastebot.ltsp.org/83
12:35
"_UsUrPeR_" pasted "/etc/hosts" (258 lines) at http://pastebot.ltsp.org/84
12:36
<Lns>
_UsUrPeR_: I'm not sure what the issue is, but I'm assuming you know your IP addresses are all public addys, right?
12:37
<_UsUrPeR_>
yeah, it's a customer's setup...
12:41* Lns has always wondered what, in LTSP, besides actually assigning static DHCP leases to the clients, causes the clients to *always* get the same IP after putting entries in /etc/hosts...
12:41
<_UsUrPeR_>
I thought it was the server caching the mac address.
12:45
<vagrantc>
warren: well, glad i found it :)
12:45
<warren>
vagrantc: what is the permission on your /dev/snd/* ?
12:45
<vagrantc>
sounds like a big price to pay for audacity as a localapp
12:45
<Lns>
_UsUrPeR_: is the server caching of MAC addresses causing them to act like static leases, or just a normal dhcp server that will give the same one if it found it before and another station hadn't taken it up yet?
12:46
<vagrantc>
warren: read-write for root and group audio
12:46cliebow has quit IRC
12:46
<ogra>
warren, rw for group audio
12:46
<warren>
no group audio here..
12:47
<ogra>
warren, likely the same for debian
12:47
<warren>
so any logged in user is part of group audio?
12:47
<ogra>
and audio is the the default groups of adduser
12:47
right
12:47
<vagrantc>
on debian, there aren't really any default groups
12:47
<warren>
I think I will need to use setfacl to add the logged in user to have rw permission
12:48
here's what's going on
12:48
our /dev/snd/* are root root crw-rw----
12:48
pulseaudio run as root is able to connect ONCE to the devices, then drops permissions
12:49
<ogra>
well, it should run permanently
12:49
oh
12:49
indeed
12:49
heh
12:49
<warren>
-L module-suspend-on-idle causes it to disconnect
12:49
so it cannot reconnect
12:49
<ogra>
yeah
12:49
<warren>
thus it crashes
12:49
I can fix this with facl's
12:49
but....
12:49
<ogra>
i have no objection if you drop that again, but ask stgraber, he surely had a reason to add it
12:49
<warren>
yes
12:49
<_UsUrPeR_>
lns: I am not for sure on either, but it seems like the first one is most probable. For instance, at my college on an open wireless network, I have gotten the same IP address for almost a week and a half. It's a higher IP address in the range (xxx.xxx.xxx.223) and I assume it's because the server has cached my MAC and has not needed to give that IP to someone else due to it's normally low amount of DHCP leases.
12:50
<warren>
he added it as an ugly hack for audacity
12:50
he knew it was a hack
12:50
even without the permission issue
12:50
<ogra>
oh, for localapp ?
12:50
<warren>
i dislike it
12:50
yes
12:50
because
12:50
if an app talks to the device directly, then pulseaudio cannot reconnect
12:50
<ogra>
cant audacity use esd mode ?
12:50
<warren>
and you have the same problem
12:50
no
12:50
audacity is broken
12:50
<ogra>
hrm
12:50
yeah, definately
12:51
well, the alsa oss compat modules should solve that
12:51
<warren>
nope...
12:51
audacity doesn't workk through that
12:51
<ogra>
as long as your card is dmix capable at leat
12:51
<Lns>
warren: ogra: remember debian/ubu doesn't have acl support by default, so setfacl won't be portable.. (i'm probably missing part of the conversation thou)
12:51
<stgraber>
warren: ah, /dev/snd/* is owned by root.audio and the users are usually in the audio group, so it works fine in Ubuntu
12:52
<ogra>
Lns, yeah, thats why i said revert it
12:52
<Lns>
ah ok
12:52
<stgraber>
ogra: audacity is really broken, the right fix would be to add pulseaudio support in portaudio but that's not something I can do myself :)
12:52
<warren>
stgraber: ignoring hte permission issue, I'm disliking -L module-suspend-on-idle because you are working around a buggy application
12:52
<ogra>
stgraber, right, and we have it with the module in intrepid, so i dont object if it gets reverted upstream
12:52
<stgraber>
warren: yeah, I know that and I'm happy to use any other way to make audacity and other buggy applications (skype ?) to work as localapp
12:53
<warren>
skype really doesn't work?
12:53
<ogra>
we have 6 months to annoy audacity upstream to fix it once and for all
12:53
<warren>
I use skype all the time on F9 with pulseaudio never letting go of the sound device
12:53
<ogra>
stgraber, but we're unlikely to use a new upstream before jaunty
12:53
so if warren rolls it back now that wont do any harm
12:54
we just need to make sure to not forget about solving it before jaunty goes final ;)
12:54
warren, skype can use alsa if alsa can provide dmix
12:54
its HW dependant
12:54
if the card doesnt do dmix it wont work
12:55
<warren>
no
12:55
I mean, yes you are right
12:55
but our alsa turned off dmix
12:55
I'm using skype through alsa-plugins-pulseaudio
12:55
<ogra>
in ubuntu we dropped dmix completely
12:55
<vagrantc>
warren: so removing g_free(host_list); from ldm fixed the problem, or just worked around it? will that be a memory leak or some such?
12:55
<ogra>
ah, good to know
12:55
<warren>
vagrantc: it might leak, but a few lines later is exit(0)
12:56
vagrantc: Ryan52 said he'll analyze further and do a proper fix if necessary
12:56
<ogra>
what is host_list ?
12:56
<vagrantc>
warren: so that should clean it up after the exit?
12:56
<warren>
vagrantc: yes
12:56
<ogra>
a g_string ?
12:56
<vagrantc>
ogra: think so ...
12:56
<ogra>
glib handles some vars automatically
12:56
i think strings are among them
12:56
so you shouldnt need to free it explicitly
12:56
<warren>
So Debian and Ubuntu don't have getfacl/setfacl utilities?
12:57
<ogra>
i think we have them in some of the hardening packages
12:57
but not by default
12:57
<vagrantc>
debian has them in the acl package, but i don't think it's typically installed
12:57
<warren>
I thought ConsoleKit sets facl's of stuff in /dev/*
12:57
<ogra>
oh, we do as well
12:57
just checked
12:58Guaraldo has quit IRC
12:58
<ogra>
hal depends on acl
12:58
so every distro shuld have it apparently
12:59
<warren>
So maybe we should add something like:
12:59
upon login, allow the logged in user rw access to the sound devices
12:59
<ogra>
well, there is no need for that on debian or ubuntu
13:00
<warren>
OTOH... there is no good reason why user apps should directly talk to those devices
13:00
<ogra>
right
13:00
<warren>
ok, i'm just backing out that parameter and tagging ltsp-trunk again
13:00
anything else you folks want in?
13:00* ogra doesnt care atm ... one day after release :)
13:02Guaraldo has joined #ltsp
13:02
<stgraber>
right, we have six months to think about that
13:02gonzaloaf_laptop has joined #ltsp
13:03Guaraldo has quit IRC
13:07
<warren>
what is the lts.conf option for numlock?
13:07
and what is the default?
13:09
<vagrantc>
warren: LDM_NUMLOCK boolean
13:09
default is unset, so false.
13:11* vagrantc hangs head in shame at another undocumented implemented feature
13:11
<warren>
hmmm
13:12Guaraldo has joined #ltsp
13:12
<warren>
did anybody test the setxkbmap change?
13:12
it isn't working now
13:12
<vagrantc>
i tested it on debian
13:12Guaraldo has quit IRC
13:12
<ogra>
worked in ubuntu
13:13
<vagrantc>
warren: it really was just a port of the C code into shell.
13:13
<ogra>
but as i said before i'd prefer a hal solution
13:13
<vagrantc>
warren: maybe setxkbmap changed?
13:13
<warren>
the C code was a port of my original shell version
13:13Guaraldo has joined #ltsp
13:13
<ogra>
but we need to wait for the rest of distros to switch to xorg 1.5 first
13:13
<warren>
why?
13:13
that particular code should work on older x.org
13:14
<ogra>
because i dont want both codepaths
13:14Guaraldo has quit IRC
13:14
<warren>
eh, both codepaths are in there now?
13:14
<ogra>
no
13:14
<vagrantc>
but ogra wants the hal stuff
13:14
<ogra>
hal doesnt read any env vars to set keymap stuff
13:15
but sets the default anyway if you have 1.5
13:15
as soon as everyone has 1.5 we should modify the ha values instead of having a second tool
13:15
*hal
13:15Guaraldo has joined #ltsp
13:16
<vagrantc>
that'll be a while.
13:16
<ogra>
as soon as lenny releases, no ?
13:16
gentoo builds upstream
13:16Guaraldo has quit IRC
13:16
<ogra>
and i guess suse will use 1.5 next release as well
13:17
<vagrantc>
ogra: i guess i can maintain a backport to lenny with patches, but i don't want to make it hard-coded.
13:17
i actually try to support stable releases with current ltsp.
13:17
<ogra>
well, that'll get hard with xorg 1.5 in place
13:17
<vagrantc>
managed to keep backports working throughout most of the etch cycle
13:18
<ogra>
i would really see us dropping all the input devices crap and write proper hal equivalents
13:18
<vagrantc>
i know. and i don't want to break backwards compatibility.
13:18
<ogra>
upstream breaks backwards compatibility
13:18
and i dont like to have everything slow down because we have to duplicate everything
13:19
<vagrantc>
i'm sure it'll bring everything to a grinding halt :P
13:20
<warren>
Hmm.... it isn't respecting XKB* options
13:21
<ogra>
well, you can drop inputattach, 90% of the xorg.conf parsing code and a lot more for two small hal scripts
13:22
<vagrantc>
warren: look at the code in ldm-trunk/rc.d/I05*
13:25Egyptian[Home] has quit IRC
13:26
<warren>
yeah... downgraded to ldm-2.0.13 and it works again
13:27elisboa_ has quit IRC
13:31
<ltsppbot>
"ogra" pasted "ltsp-setkb.sh" (31 lines) at http://pastebot.ltsp.org/85
13:31
<warren>
i'm a little confused...
13:31
<ogra>
theer you got the hal variant (untested, but i'm pretty sure it works)
13:31
<warren>
I added a debug echo print to I05 and it began working
13:31
<ogra>
needs to run before X is up
13:33
<Ryan52>
ogra: host_list is a GList
13:34
<vagrantc>
warren: that's just weird.
13:35
<ogra>
Ryan52, hmm, not sure about glists ... i'm pretty sure g_string gets freed automatically
13:36
.oO( why did they have to name a variable descriptor after female underwear ... /me never gets that ... or is only gernamy calling tehm g-stings)
13:36
<vagrantc>
ogra: i'm sure they got a kick out of it.
13:36
<ogra>
heh
13:37* vagrantc notes that it's not just female underwear
13:37
<warren>
back
13:37
<vagrantc>
heh
13:41
<warren>
so I don't know wtf is going on
13:42
<Gadi>
hey, ogra: does the initramfs have stat?
13:43
in ubuntu
13:43
or do I have to parse the output of ls?
13:43
<ogra>
no idea
13:43
<Gadi>
big help :P
13:43
<ogra>
pick the safe way, use ls if you know it works :)
13:43
<warren>
grrr
13:44
I logged out and back in, and now it is broken again
13:44
something else is setting the keyboard?
13:44
<ogra>
hal does
13:45
but that happens before X is up
13:45tino has joined #ltsp
13:46
<tino>
LTSP5 is up and running fine, but with my second WS, which I would like to set-up, above message is comming. MAC address, I've entered in the ltsp.conf file in the same way as with the "running" WS. Any idear how to fix that?
13:48
LTSP5 is up and running fine, but with my second WS, which I would like to set-up, this message "Valid Link not established". The MAC address, I've entered in the ltsp.conf file in the same way as with the "running" WS. Any idear how to fix that?
13:51
<sbalneav>
Valid link not established?
13:51
<alkisg>
Gadi: stat is there on intrepid initramfs
13:51
<sbalneav>
Where do you get that message?
13:51
<warren>
something is very screwy
13:51
It sets the keyboard only sometime
13:52
<tino>
The msg comes during boot of the WS
13:52
<ogra>
warren, sounds racy
13:52
get a slower client :)
13:53
<sbalneav>
tino: At what point during the boot? After the kernel's loaded? And is that the exact message?
13:55
<tino>
sorry hw can I see that the kernel is loaded, I don't know, but the msg is comming quite early, I would gues when the WS starts "talking2 to the DHCP
13:56
<sbalneav>
What kind of network card is in the thin client. Maybe the kernel can't find a driver for it.
13:57* ogra has never heard that error message
13:57
<sbalneav>
Neither have I, that's why I asked if that's the exact errror message.
13:58
<ogra>
yeah
13:58* warren tries the hal way
13:58
<ogra>
warren, make sure the script is executed before X comes up but if hal already runs
13:59
<tino>
Good point, I guess also it's the card, I'm not sure but I think the card was working in the network . But I think I will try a different one
13:59
<ogra>
it must sit between the two
13:59
<warren>
ogra: the hal way works only if you use evdev?
14:00
<sbalneav>
tino: Do you see the kernel downloading onto the client?
14:00
<ogra>
which is the default
14:00
<tino>
before the MAC adr. is comming and then this exact msg:"Valid Link not established"
14:00
<ogra>
yes
14:00
<warren>
ogra: not on older distros
14:00
<sbalneav>
That's the card that's issuing that.
14:00
<ogra>
warren, on all distros using xorg 1.5
14:00
<sbalneav>
You've got a hardware problem
14:00
check your wires/switches, etc.
14:01
<tino>
sbalneav; I don't think so
14:01
<warren>
ogra: F9 didn't use evdev despite xorg 1.5
14:02
<sbalneav>
Well, you're getting it before the mac address, before dhcp, right?
14:03
<alkisg>
tino: what do you use to boot the WS? etherboot?
14:03
<ogra>
warren, how could FC9 use 1.5 ?
14:04
it wasnt existent back then
14:04
<warren>
ogra: 1.49999999999
14:04
the ABI of 1.5 was set long ago
14:04
F9 later upgraded into 1.5
14:04
<sbalneav>
I thought 1.49999972549 was the best version.
14:05
They totally wrecked it in 1.49999972550
14:05
<warren>
something racy is going on =(
14:05* ogra waits for 3.1415938 :)
14:05
<tino>
alkisg:I use etherboot
14:05
<sbalneav>
22/7: Not Pi, but an incredible simulation!
14:06
tino: So, looks like Etherboot's not happy with something on the network.
14:06
<alkisg>
tino: I think either the etherboot driver of your card is bad, or you have wires/switches problem like sbalneav said
14:07
<ogra>
warren, shouldnt be hard to adapt the script to whatever driver is used in FC9 though
14:07
warren, i'm sure the option names will still be the same
14:07
<sbalneav>
http://www.virtualbox.org/browser/trunk/src/VBox/Devices/PC/Etherboot-src/drivers/net/eepro100.c?rev=1
14:07
754 * This allows for faster failure if there is nothing
14:07
755 * we can do.
14:07
756 */
14:07
757 if (!(mdio_read(eeprom[6] & 0x1f, 1) & (1 << 2))) {
14:07
758 printf("Valid link not established\n");
14:08
759 eepro100_disable(dev);
14:08
760 return 0;
14:08
761 }
14:08
<Gadi>
alkisg: thanks, man - I had to stuff awk in there for something else, anyway - so I'll be safe and parse
14:08joebaker has quit IRC
14:08joebaker has joined #ltsp
14:09
<tino>
alkisg:Thanks will change the card and let you know
14:09
<sbalneav>
So, according to the etherboot source, it's checking to see if the network cable's plugged in.
14:09
And not finding one.
14:09
<alkisg>
tino, you boot from a rom or a floppy/cd/hd?
14:10
<tino>
alkisg: from CD which works fine with the other computer
14:10
<warren>
I'm running ldm-2.0.13 over and over again to see if it was intermittent back then
14:10
<alkisg>
tino, just download the newer version, gpxe. You can download a CD with all the drivers from rom-o-matic
14:11
<warren>
OH!
14:11
when do the I* scripts run?
14:11
<ogra>
after the ssh tunnel is up
14:11
<warren>
that's the wrong place
14:11
<sbalneav>
tino: according to the etherboot source code, the network card doesn't see a link
14:12
<ogra>
no ?
14:12
<tino>
alksig: that is the CD I'm using and the network card is also very old (all my WS are old :=) ). I will change the network card I've lots of them
14:12
<warren>
You want the keyboard layout set even before someone logs in
14:12
<ogra>
warren, the I scripts should be identical to the former behavior of the S scripts
14:12
<warren>
otherwise they might not be able to type their name and password
14:12
No
14:12
ogra: look at where my C method call was
14:12
<ogra>
i know wher your C call was
14:12
<alkisg>
tino, http://www.rom-o-matic.net/gpxe/gpxe-git/gpxe.git/contrib/rom-o-matic/ - and choose "gpxe:all-drivers"
14:12
<warren>
ogra: I put it in that seemingly strange place because it had to be done immediately before the greeter is painted
14:13
<ogra>
right
14:13
<warren>
setting the keyboard layout any other time is unreliable
14:13
<sbalneav>
Wow, I go look up the exact place in the source code where the error is, and I'm kinda ignored in this conversation. Shutting up now.
14:13
<warren>
I ran into exactly this problem earlier and put it there for this reason.
14:13
<ogra>
warren, the hal way is reliable
14:14
<warren>
ogra: but only after everyone is using evdev
14:14
<ogra>
you can run it from the ltsp-client initscript
14:14
well, it wil work for FC10, ubuntu, gentoo and debian sid
14:14
<warren>
ogra: I think we should do it both ways
14:14
<vagrantc>
the I* scripts run before the greeter starts
14:14
<warren>
ogra: F9 and F10 LTSP are identical in version
14:14
vagrantc: yeah, wrong place
14:14
<ogra>
warren, ltsp, yes
14:14
<vagrantc>
warren: how's that?
14:14
<ogra>
Xorg no
14:15
vagrantc, fedora doesnt have stable releases
14:15
<vagrantc>
warren: before the greeter starts is before someone logs in.
14:15
warren: after X starts, before the greeter.
14:15
<warren>
vagrantc: I put setup_keyboard_layout() in that strange place because it was the only reliable place to put it
14:16
vagrantc: where in the code is running of the I* scripts?
14:16
ogra is correct that the hal way is proper
14:16
but we should do both until distros no longer need the old way8.
14:16
way
14:17
<ogra>
which will be extra slow :/
14:17
<warren>
get over it
14:17
<vagrantc>
warren: src/ldm.c line 474
14:17
warren: that's where the I* scripts get run
14:18
<warren>
vagrantc: yeah, that's why I am seeing racy behavior
14:18
<alkisg>
sbalneav, sorry if I got in the way... I didn't see your message, I was looking for the url
14:18
<warren>
vagrantc: the location where I set the keyboard made it stick without relying on timing
14:18
<vagrantc>
i don't see how this relies on timing ...
14:19
<warren>
it is failing to stick setxkbmap settings here on both F9 and F10 maybe 80% of the time
14:19* ogra hugs sbalneav
14:19
<warren>
vagrantc: ldm and ldmgtkgreet are separate processes
14:19
<sbalneav>
what'd I do?
14:19
<ogra>
sbalneav, you sounded unhappy
14:20
<warren>
fprintf(ldmlog, _("Launching Xorg\n"));
14:20
launch_x(argv); /* Launch X, pass any command line args */
14:20
rc_files("init"); /* Execute any rc files */
14:20
<ogra>
so i thought a hug would help :)
14:20
<vagrantc>
warren: and ...
14:20
<warren>
vagrantc: X is forked but not fully initialized at the time it runs the I* scripts
14:20
<sbalneav>
Thankee
14:20
<ogra>
vagrantc, x might still be in the process of launching
14:20
<vagrantc>
ah.
14:20
<warren>
vagrantc: if you set keyboard where the greeter would paint for the first time, then you're guaranteed for it to be ready.
14:21
<vagrantc>
what a mess.
14:21
<warren>
Yes, I ran into this problem before.
14:21
<ogra>
well, hal fixes all that
14:21
<warren>
I had setup_keyboard_blah() exactly there
14:21
that's why I put it into the greeter
14:21
<ogra>
sicne input devices are handled separately from X
14:21
<warren>
ogra: I'll put it back into the greeter, but later add a boolean to avoid doing it again if evdev.
14:22
<ogra>
yeah, sounds sane
14:22
<warren>
ogra: that should satisfy both old and new X
14:22
<sbalneav>
We could add an XOpenDisplay(); call there, which should block until the X display's ready.
14:22
<ogra>
mind you, i might not port it to the new greeter though
14:22* vagrantc hugs sbalneav
14:22
<warren>
OH!!
14:22
<ogra>
??
14:22
<warren>
That also explains why ldm crashes sometimes but works the second time
14:22
sbalneav is totally correct
14:23
sbalneav: just XOpenDisplay(); is all that is needed?
14:23
<sbalneav>
Even a stopped clock is right once a day.
14:23
:)
14:23
<vagrantc>
put an XOpenDisplay between launch_x and rc_files("init) ?
14:23* warren tries it
14:23
<sbalneav>
Ummm, gimme two secs, I'll give you a code snippet
14:23
<warren>
ok
14:23* ogra would put it into the greeter
14:23
<warren>
ogra: put what?
14:23
<ogra>
XOpenDisplay
14:23
<warren>
ogra: no
14:24
<vagrantc>
then we'd have to move the rc_files function into both the greeter and ldm
14:24
<warren>
ogra: the greeter was failing to run the first time because it was launched before X was ready
14:24
<ogra>
oh, wait, the greeter wouldnt come up without display anyway
14:24
<vagrantc>
and that
14:24
<warren>
sbalneav has the correct solution.
14:24
<ogra>
right
14:24
<warren>
holy crap, we're a team
14:24
<ogra>
yay
14:24
<warren>
Ubuntu breaks things, I figure out why, sbalneav fixes it.
14:24
=)
14:24
<ogra>
:P
14:25
debian broke it
14:25
<warren>
This does solve the long standing issue of ldm being racy
14:25
<ogra>
yeah
14:26
though i never saw these races here i must admit
14:26
ubuntus X must be faster :)
14:27
<warren>
or your systems are slower
14:27
<ogra>
or that
14:27
someone said that recently
14:27
though they didnt compare to other distros only to older ubuntu yet
14:28
warren, btw, whats your opinion about using squashfs for initrd ?
14:28
<ltsppbot>
"sbalneav" pasted "XOpenDisplay" (17 lines) at http://pastebot.ltsp.org/86
14:28
<ogra>
(assuming you read the lwn article)
14:30
<warren>
ogra: ?
14:30
<sbalneav>
Summat like that should suffice.
14:30
<ogra>
squashfs is in mainline now
14:30
with .28
14:31
<warren>
that hasn't stopped anyone from using it before...
14:31
<sbalneav>
Really, we wanted to move the starting of X outside of ldm per Gadi's suggestion.
14:31
<ogra>
and they had a comparison to cramfs up ... very impressing
14:31
no, but now you could use a squashed initramfs
14:31
<warren>
sbalneav: would that snippet work even if X was started before ldm?
14:31
<sbalneav>
What we COULD do, when we do that is do the XOpenDisplay right at the beginning...
14:31
yes, I'm just coming to that...
14:32
<warren>
sbalneav: ok, so I should insert this after it launches X?
14:33
<sbalneav>
So, we could leave the display structure open, and then we could have things like LDM read atoms, and clean up the atoms after a session ends, so that X doesn't need to be restarted, etc.
14:33
yeah, after X starts, but before you run the init scripts.
14:33
<warren>
wait... there was some other reason why restarting X was a good idea
14:33
sbalneav: GNOME explicitly restarts X every time for security reasons
14:33
<sbalneav>
Cleaning up X atoms
14:34
But, I'm beginning to wonder if we couldn't just enumerate all the atoms in the server, and any that DON'T get cleaned up, we clean them up manually.
14:35
Lets hash out that idear in main.
14:35
<ogra>
it wont work
14:35
<sbalneav>
No?
14:35
<ogra>
(leaving X running i mean)
14:35
no, because that means your localapp stuff stays
14:36
thats a major security issue
14:36
<warren>
there were other reasons
14:36
you can change the xauth cookie for the next user
14:36
<ogra>
i personally like the ised of keeping it running
14:36
<warren>
but existing clients can remain connected and snoop
14:36
<ogra>
*idea
14:36
yeah
14:37
<warren>
This really is not an important problem
14:37
<ogra>
xauth, sshfs remainings, loaclapp remainings etc
14:37
<sbalneav>
Meh, restarting's find for now.
14:37
<ogra>
warren, Gadi will make it one :) he is pushing for it since years
14:37
<warren>
to workaround buggy X drivers that fuck up at restarting?
14:37
<sbalneav>
But actually having LDM have access to X atoms might be useful.
14:37
<warren>
=)
14:37
<ogra>
(since the first python ldm)
14:38
heh
14:38
<sbalneav>
What we need to do is talk to KeithP, and get him to implement a "light" restart function that:
14:38
Kills all current connections
14:38
cleans out all existing Xatoms,
14:38
<ogra>
that cleans up the filesystem :P
14:39
<sbalneav>
anything else..
14:39
WithOUT restarting the core X and display.
14:39
<ogra>
ignores xauth
14:39
<sbalneav>
X soft reset.
14:39
<warren>
when we have kernel modesetting you wont notice flicker of restarting X
14:39
This is really a non-issue
14:39
<ogra>
sbalneav, i think thats what Xmove does
14:39
and i thing thats why Xmove was abandoned ... even by its dev
14:40* warren building with scottie's patch
14:41
<sbalneav>
Keep in mind, we might need to put it in a loop.
14:41
So we might need something like:
14:42
for i = 1 to 10; if (XOpenDisplay()) break; else sleep(1); rof
14:42
in pseudocode
14:42
We'll have to experiment. I THINK it blocks.
14:43
<warren>
sbalneav: ajax said this isn't good enough
14:43
sbalneav: xopendisplay will try twice then just fail
14:43
sbalneav: it doesn't block waiting for X
14:43
oh
14:43Guaraldo has joined #ltsp
14:44
<warren>
sbalneav: sleep(1) is one second?
14:44
<ogra>
eeek
14:44
<warren>
sbalneav: he said that X sends a signal to its parent process when it is ready
14:44
sbalneav: gdm has code for this
14:44* ogra is allergic to sleep
14:44
<vagrantc>
ogra: i thought you just released!
14:45
<ogra>
vagrantc, well, enforced sleep :)
14:46
<sbalneav>
ok, so let's steal^h^h^h^h^h^h borrow from gdm :)
14:46
brb, workping
14:46
<ogra>
sbalneav, as long as that still works :P
14:47
new gdm wont have anything thats worth stealing
14:51
<loather-work>
yeah, the new gdm as shipped with F9 and probably the latest ubuntu isn't all that great at all
14:51CAN-o-SPAM has quit IRC
14:51
<loather-work>
breaks well-defined behaviour, launches a bunch of stuff you don't necessarily want launched, etc.
14:52
<stgraber>
ok, we now have our LTSP-Cluster box installing. Core2Duo + 2GB of RAM and a 80GB 10k RPM
14:53ogra has quit IRC
14:55ogra has joined #ltsp
14:56tino has quit IRC
14:57pmatulis has quit IRC
15:00* vagrantc wonders what GDM is doing that went so wrong ...
15:00
<ogra>
gdm is a user session now
15:00
with the latest code
15:01
i.e. it also runs things like NM or gnome-power-manager
15:01
no mores way to use/set anything on the system level
15:01
<loather-work>
and gnome-screensaver
15:02
<ogra>
right
15:02
<loather-work>
yeah, you can't theme it, it's always got the click to select a user thing
15:02
<ogra>
it wont in ubuntu :)
15:02
<loather-work>
it's regressed into something only marginally useful
15:02
yeah, the more i'm using fedora, the more i think it's going in the wrong direction
15:02
<ogra>
i thnk MacSlow is basing his cool cairo code on the old greeter
15:04
<loather-work>
and i've been a redhat user for a very logn time
15:06* ogra points to http://macslow.thepimp.net/?p=163
15:06
<ogra>
likely to be in 9.04
15:07
<mccann>
so much misinformation here
15:09
<ogra>
educate us :)
15:09
isnt it running a user session ?
15:10
<mccann>
gnome-session is just a way to execute .desktop files. you can add whatever you like
15:10
it always was a "user session" just not a useful one
15:10
<ogra>
right
15:10
<mccann>
always had a windows manager etc
15:10
just didn't allow you to suspend correctly when you close your laptop lid
15:10
etc
15:11
does not run gnome-screensaver obviously
15:11
<ogra>
not the job of gnome-session imho
15:11
suspend resume is a kernel job
15:11
<loather-work>
ogra has that nailed 100%
15:11
<mccann>
not talking about gnome-session
15:11
<ogra>
even pm-utils are a big failure in my opinion
15:11
ah, your mean gdm
15:11
ok
15:12
right, if you base th whle system on the assumption that gpm is the thing responsible for power management, you need a user session to run and manage it
15:13
<warren>
sbalneav: OK, just talked to rstrode, he recommended looking at rhgb for simpler code to handle this.
15:13
<ogra>
same goes for network management ... if you assume NM should manage everything using a user session is the only proper way to go
15:14
<loather-work>
yeah, i think i'm gonna switch back to XDM and use xsession files.
15:14
IMO NM has absolutely no place on a server
15:14
<ogra>
i dont complain about waht you do with gdm, but i dont like the assumptions that bases on
15:15
from a gdm POV i might have been the right selection and will surely ease security maintenance
15:15
the while underlying concept is what i dont like there
15:16
<warren>
ogra: NM is growing the concept of system connections
15:16
<ogra>
*whole
15:16
<warren>
ogra: you can have connections as either user or system
15:16
<ogra>
warren, right .... *growing*
15:16
have you ever looked at intels network manager tool ?
15:16
<warren>
it is pretty fucked that we shipped it as default before it is ready
15:17
<ogra>
they have a conurrent product that gives a way saner impression
15:17
*concurrent
15:17
NM is getting there though
15:18
but its sad thet the intel NM didnt get any community attention ... if you would have merged both projects 1/2 year ago we would already be there
15:19
<loather-work>
well, the intel one isn't a python monstrosity that doesn't work right half the time :P
15:19* ogra hates NIH as much as he suffers from it
15:19
<warren>
sbalneav: ogra: OK, I know how to fix this properly, but it is going to take a significant amount of code refactoring
15:20
<ogra>
warren, many changes to the greeter ?
15:20spectra has quit IRC
15:20
<warren>
ogra: no, ldm.c
15:20
<ogra>
thats my main focus of ltsp work atm
15:20
fine with me then
15:22
<warren>
ogra: until I have time to refactor, I'm moving rc_files("init"); into the greeter
15:22
it is the only safe way to do this.
15:22
<ogra>
ok
15:23
i havent ported much to the new greeter yet
15:23
still playing with concepts
15:23
<warren>
damn...
15:23
<ogra>
??
15:23
<warren>
greeter lacks ldmlog, ldm_spawn, etc.
15:24
<ogra>
the new code has tons of debug options
15:24
<warren>
what new code
15:24
is this secret?
15:24
why has it never been mentioned on ltsp-developer?
15:24
<ogra>
scottie is working on an exportable ldm_log feature that can be switched to log to syslog
15:25
warren, i started working on it when i noticed that the old stuff doesnt suffice anymore
15:25
<warren>
ldm.c needs almost a complete rewrite
15:25
<ogra>
enhancing the UI always influences and interferes with other element
15:25
s
15:25
feel free
15:26
i'm doing the greeter now ... having each element in its own function (so you can add new elements by adding a new funtion, can switch elements on and off as you like)
15:26
and having all elements freely positionable ... even overlapping so nothing interferes with anything anymore
15:26six2one has quit IRC
15:27
<ogra>
thast my main focus ... i will put the code online as soon as i have some structure and the basic features ported
15:28
all lelements i have yet use theme debugging funtions for theme devs, that should be switchable as soon as i have scotties implementation
15:29
so theme autors should have an easy way to develop by switching on theme debugging
15:30
and it should easily be possible to add someting like a ... analogue clock
15:32
<Gadi>
have a wood geekend, kids!
15:32
:P
15:32
<ogra>
you too
15:32
<Gadi>
don't port anything I wouldn't
15:32
<ogra>
and have a beer for me
15:32
<Gadi>
:)
15:32
u bet
15:32
<ogra>
:)
15:32Gadi has left #ltsp
15:32
<ogra>
and hugs and greetings to ragnar
15:32
bah
15:32
<Appiah>
anyone know how much space do you need on the Ltsp client to use local firefox?
15:32
<ogra>
gone
15:33
Appiah, its all a matter of ram
15:33DonSilver has quit IRC
15:33
<ogra>
local apps are still sitting in the nbd image or nfs root on the server
15:34
<Appiah>
but if I run "local" firefox
15:34
<warren>
gahh
15:34
<Appiah>
it would take cpu and ram load of the server right?
15:34
<warren>
this is a LOT of work...
15:35
<ogra>
Appiah, right and put bth on the client
15:35
*both
15:35
<warren>
I'm going to use a different approach
15:35
ugly hack
15:35
<Appiah>
this wiki entry seams old...
15:35
<ogra>
well, if you rewite ldm.c at some point ...
15:36
<warren>
ogra: loop in the shell script, if setxkbmap exited 255, X is not ready yet
15:36
<Appiah>
http://www.ltsp.org/twiki/bin/view/Ltsp/LocalAppFirefox
15:36
<ogra>
bah
15:36
<warren>
I need SOMETHING that works now
15:36
<ogra>
Appiah, all wiki entries are
15:36
warren, well, use the hal script if hal is installed ... that fies it on FC10
15:37
<Appiah>
does the docu team update the wiki ogra ?
15:37
<ogra>
oh, you installed hal in FC9 already, right ?
15:37
Appiah, not yet
15:37
<warren>
ogra: I need to simultaneously fix it in RHEL5, F9 and F10
15:37
<ogra>
yeah
15:37
<warren>
so doing it this way
15:37
<Appiah>
well , they are the ones to update it right?
15:37
<ogra>
not sure
15:37
<warren>
The ugly loop can be removed when ldm.c is refactored to watch for the SIGUSR1 from X properly
15:38
<ogra>
i dont think there actually is a wiki team
15:38
warren, yeah, the only person actually caring about ldm atm might be waren
15:38
err
15:38
vagrantc,
15:38
since he is near a release
15:39alkisg has left #ltsp
15:40* ogra already had to much of that awesome Rioja
15:41alkisg has joined #ltsp
15:41jammcq has quit IRC
15:42
<warren>
while true; do
15:42
setxkbmap 2> /dev/null
15:42
if [ $? -eq 255 ]; then
15:42
sleep 0.1
15:42
else
15:42
break
15:42
fi
15:42
done
15:42
<ogra>
does sleep work without brackets there ?
15:43
or are you in shell ?
15:43
oh, you are
15:43
right
15:43
<vagrantc>
warren: does it only exit 255 if it can't reach an X server ?
15:43
<warren>
vagrantc: I dunno, testing now
15:43
<Appiah>
Should I send a question to the mailing list to find out whos updating the wiki?
15:44
<ogra>
Appiah, yeah, that might help
15:44
<vagrantc>
warren: what about something like a call to "xwininfo -root" instead ? maybe there's not really any need.
15:44
<ogra>
though it wuld mae sense if the ltsp-docwriters would do t
15:44
it
15:44
<warren>
This hack doesn't come anywhere near the awesomeness of Gadi's hacks.
15:44
<Appiah>
oh my looks like sbalneav is on the ltsp-localapps for ubuntu wiki
15:45
<ogra>
warren, i'd say about 50% near though
15:45
Appiah, thats ancient
15:45Guaraldo has quit IRC
15:46* vagrantc is still using ldm 2.0.6 for release
15:46
<Appiah>
oh
15:46
<vagrantc>
so changes in ldm-trunk don't worry me really.
15:46
<Appiah>
you're on the list too;)
15:46
editoer 2008-08-06 was localhost
15:46
damn
15:46
<ogra>
Appiah, ltsp development was sponsored by canonical for two years, we worked out a lot of specs we didnt implement during that timeframe
15:46
but thats about a year ago
15:46
<Appiah>
canoical ?
15:47
<ogra>
the company sponsoring ubuntu
15:47
<Appiah>
oh
15:47
so it is not being developed anymore?
15:47* ogra points Appiah to http://wiki.ltsp.org/twiki/bin/view/Ltsp/Ltsp5
15:48
<Appiah>
oh ofcourse
15:48
:D
15:48
<ogra>
it moved from being developed by me paid for it fulltime to a multi distro project
15:48
<Appiah>
Ye I understand now
15:48
<warren>
vagrantc: seems like it is 255
15:49
vagrantc: I had it print to ldm.log, it printed 3 times before it was ready.
15:49
<ogra>
a year go i moved on to new duties in canonical and ltsp is now maintained by a team of devs from all major distros
15:49
as a team effort across distro borders
15:49
<warren>
All?
15:49
Where is Mandriva?
15:49
or SCO?
15:49
<ogra>
dead ?
15:49
<Appiah>
haha
15:49
<ogra>
haha
15:49
<Appiah>
sco :D
15:50
<vagrantc>
warren: i'm wondering if we shouldn't implement a I01-wait-for-x or some such rather than implementing it in each plugin
15:50
<ogra>
isnt Mndriva in the process to rename itsefl to "the distro formerly known as ..." ?
15:50
<vagrantc>
or even implementing it in ldm-script
15:51
<ogra>
vagrantc, better in the C code
15:51
before you parse any of the scripts
15:51
<vagrantc>
that's fine too.
15:51
<otavio>
ogra: formely known as ...?
15:51
:-D
15:51
<vagrantc>
otavio: hi! :)
15:52
<ogra>
while ! getenv(DISPLAY) sleep(1) ....
15:52
<warren>
vagrantc: OK, I'll do that.
15:52
<otavio>
vagrantc: :-D
15:52
<vagrantc>
ogra: i think DISPLAY is set, you just can't access it
15:52
<ogra>
hey otavio :)
15:52
<otavio>
vagrantc: nice to see you around
15:52
<ogra>
vagrantc, from ldm.c ?
15:52
that would be bad
15:52
<otavio>
ogra: we use xwininfo to detect when a display is available here in company
15:52
ogra: DISPLAY fails
15:52
<ogra>
oh, wow
15:52
it shouldnt imho
15:53
<otavio>
ogra: xwininfo really waits to the display to be connectable
15:53
<ogra>
but if xwininfo is proven to work, lets take that
15:53
<warren>
vagrantc: you're actually correct, any X client will work
15:53
<vagrantc>
warren: i just choose xwininfo because i think it's pretty core to X ... but maybe i'm wrong.
15:53
<ogra>
otavio, the ldm greeter is spawned by ldm.c though ... would be bad if there were no display
15:54
<otavio>
ogra: xwininfo -root
15:54
ogra: this does the trick
15:54
<warren>
xorg-x11-utils isn't guaranteed to be installed I think
15:54
<ogra>
yeah, no objections
15:54
<Appiah>
Guess I'll start playing with local app firefox and see if I can write a more uptdate howto
15:54
<ogra>
warren, make it a dependecy
15:55
<warren>
xwininfo blocks?
15:55
<otavio>
ogra: if it is ldm.c; a simple while look that uses XOpenDisplay does it
15:55
warren: yes, it does
15:55
<Appiah>
then get whoevers in charge of the wiki to accept my howto
15:55
<otavio>
warren: and fail with $? too
15:55
<warren>
wait
15:55
<ogra>
otavio, that was the initil suggestion from sbalneav
15:55
<warren>
fail with $? and blocking are not thes ame
15:55
<ogra>
not sure why warren and vagrantc dropped it
15:55
<warren>
dropped what?
15:55
<ogra>
XOpenDisplay
15:55
<otavio>
warren: it has a small timeout; and fail too. So you need to use it in a look
15:55
<ogra>
in ldm.c
15:56
<otavio>
I'd more to use XOpenDisplay in ldm.c
15:56
<vagrantc>
XOpenDisplay in a loop sounds reasonable.
15:56
<ogra>
yeah
15:56
<warren>
I'm in favor of a I01 script because it is less code, and this has to be redone properly anyway
15:56
XOpenDisplay is not proper
15:56
<ogra>
i'm pretty sure even getenv(DISPLAY) should work from ldm.c
15:56
else the greeter wouldnt come up
15:57
<warren>
We need to be launching X in a different way, watching for the SIGUSR1 signal when it is complete
15:57
with signal handlers and such
15:57
<vagrantc>
otavio: how's debian-installer coming along? :)
15:57
<warren>
I'm doing the I01 script because it is simplest
15:57mikkel has quit IRC
15:57
<ogra>
oh, otavio id working on d-i nowadays ?
15:58
*is
15:58
<vagrantc>
otavio: release manager for d-i, no?
15:58
<ogra>
wow
15:58
cjwatson gave that away ?
15:58
cool
15:58
<otavio>
ogra: http://paste.debian.net/20405/
15:58
ogra: this does it
15:58
vagrantc: yep
15:59
<ogra>
warren, ^^^
15:59
<otavio>
warren: yes. It signals SIGUSR1 when it is ready to accept connections
15:59
warren: however if you have multiple X running you can't wait for it
16:00
warren: so XOpenDisplay is the safest way of checking it. So you can choose which display to /wait/ for
16:00
<warren>
umm
16:00* otavio goes back to work :P
16:00* vagrantc pushes for ltspfs 0.5.5 in debian lenny
16:00
<warren>
isn't it ok to use any X client with DISPLAY set to check for the same thing?
16:00
<ogra>
.5 ?
16:00
<vagrantc>
ogra: maybe we'll have the same upstream after all :)
16:00
<ogra>
yay
16:01
well, your release seems to take a while
16:01
<vagrantc>
ogra: you've got 0.5.5-0ubuntu?
16:01
<otavio>
warren: it is; but if you have an app that launches ldm why not check it?
16:01
<vagrantc>
ogra: they sure do.
16:01
<warren>
I rather redo ldm.c properly with signal checking
16:01
<ogra>
yep. 0.5.5-0ubuntu1
16:01
no ubuntu modifications even
16:02
<vagrantc>
ogra: we've got 0.5.5-1 in experimental, but i'm trying to get a -2 into lenny.
16:02
<otavio>
warren: as I said, it will work for most cases but fail in few ones. If you look at GDM/KDM code they also check for XOpenDisplay due that
16:02
<ogra>
stephane will be able to sync right away in jaunty
16:02
<otavio>
warren: in our multiseat app, we do it too
16:02
<vagrantc>
ogra: you should be able to pull 0.5.5-1 whenever you do that sort of thing.
16:02
<ogra>
two to three weeks from now
16:02
depends on the toolchain
16:02
<warren>
if DISPLAY is set to the desired display, how can it fail to check the proper X display?
16:02
also ldm doesn't support multiple concurrent ldm
16:02
<otavio>
warren: the signal
16:03
warren: signal will be send but you don't know which starting X is sending it
16:03
<ogra>
otavio, doesnt matter atm
16:03
<warren>
otavio: are you sure? the signal is sent to the parent process by the child
16:03
<ogra>
we only have one ldm
16:03
<otavio>
warren: humm ...
16:04
warren: we had a cases in the multiseat that cause us to drop the signal ...
16:04
warren: /me tries to remember why
16:04
<warren>
that sounds like a bug?
16:04
<ogra>
multiseat was a bug all over :)
16:04
<otavio>
warren: X has a lot of ugly things.
16:04
ogra: hehe
16:04
ogra: people likes it
16:04
<ogra>
ubuntu finally dropped it
16:04
<otavio>
ogra: and it is somewhat interesting
16:05
<ogra>
yeah, the idea is
16:05
<vagrantc>
multiseat thin clients make a lot of sense.
16:05
<ogra>
the implementation always was a pain
16:05
<otavio>
ogra: it works fine; but is rather complex
16:05
<ogra>
yeah
16:05
which makes it painful
16:05
<otavio>
ogra: mostly due the badness of VGA drivers and like
16:06
<ogra>
i remember daniel stone spending 1/3 of his time of X maintanance on it
16:06
<otavio>
X is also not friendly with multiple VGA cards
16:06Guaraldo has joined #ltsp
16:06
<ogra>
the other 2/3 were for the rest of Xorg
16:07joebaker has left #ltsp
16:07
<otavio>
So you can use a single instance, that works fine for VGA cards with multiple internal GPU or multiple X instances that works somewhat better but required a XDMCP server locally to wrap it out
16:07
so; ugly and complex to manage :P
16:07
<ogra>
yeah
16:07
and dstone even tried to automate it for us :)
16:08
<otavio>
ogra: we've written a multiseat manager; it is a pain but works for most distros now
16:08
<warren>
is multiseat a production thing?
16:08
or abandoned?
16:08
<ogra>
abandone
16:08
<otavio>
warren: depends
16:08
warren: in ubuntu, gone
16:08
<ogra>
upstream at least
16:09
well, ubuntu uses upstream and enhances it
16:09
<loather-work>
multiseat; this being two physical consoles on the same box?
16:09
<ogra>
where necessary for integration
16:09
loather-work, or more
16:09
<otavio>
ogra: upstream will get multiseat going; people is going to do it right in future but it depends on kernel support
16:09
<ogra>
yeah
16:09
if it is upstream agai it will return in ubuntu as well
16:10
+n
16:10
<loather-work>
i see. i can see the utility of it, but thin clients are probably a better, easier choice IMO
16:10
<otavio>
loather-work: yes. TC are easier to work with
16:10
loather-work: but MS are cheaper
16:10
<ogra>
loather-work, imagine a work cube with four workplaces
16:10
<otavio>
loather-work: it depends on the use
16:10
<ogra>
you only have one achine to maintain
16:11
<otavio>
loather-work: MS has distrance as a limit
16:11
loather-work: distance
16:11
<loather-work>
definitely. when i worked at the naval ocean systems center we had a bunch of HP J210s that did it natively
16:11
up to four consoles per machine
16:11
which was nice, since the machines were $24K a pop
16:11
<ogra>
HP had the 441 .,, which tried to do multiseat on a HW level
16:11
<otavio>
loather-work: yes; exactly it. But done with regular machines
16:11
<ogra>
failed as well ... sadly
16:12
<loather-work>
yeah, these J210s were awesome in their time. dual 180MHz PA-8200c with a whopping 1GB RAM
16:12
<ogra>
yep
16:13
but it wasnt adopted by customers
16:13
not enough at least
16:13
<loather-work>
i miss PA-RISC. it was a great architecture.
16:13
I don't miss HP/UX though :)
16:13* ogra loved MIPS
16:13
<ogra>
and i liked IRIX
16:13
<loather-work>
MIPS was also a good one.
16:13
irix had some neat features
16:13
<ogra>
*and* the SGI design
16:13
<otavio>
heheh
16:14
nostalgy :-)
16:14
<loather-work>
yea... it's a shame they all jumped on the itanic boat
16:14* ogra still has some SGIs in his office
16:14
<warren>
tagging
16:14
<loather-work>
i got rid of the last of my SGIs a few years back
16:14
the last one i had was an O2 with a godawful slow R4400
16:14* otavio waiting compilation to finish; go beast!
16:14
<ogra>
i would never give them away
16:15captain_magnus has joined #ltsp
16:15
<loather-work>
i hadn't turned it on in a couple of years
16:15
<ogra>
me neither
16:15
but thats no reason :)
16:15
<loather-work>
had been collecting dust, so i put it on craigslist and in like 15 minutes i had a guy saying, "GIMMEEE!!"
16:15
<vagrantc>
it's not even a question of multiseat or thin-client, you can actually have multi-seat thin clients ...
16:15
<loather-work>
ended up trading it for a 6-pack fo beer
16:15
<ogra>
pffft
16:15
<loather-work>
multiseat thin clients would be interesting
16:16
<otavio>
vagrantc: indeed an interesting thing
16:16
<ogra>
yeah
16:16
<loather-work>
i miss serial terminals. :)
16:16
green screen baby!
16:17
<alkisg>
Brazil uses a combination of multiseat and ltsp (tweaked)
16:17
I tried multiseat on my laptop, I got it working with 2 screens/mice/keyboards, it was very fast but somewhat unstable
16:18F-GT has quit IRC
16:19
<alkisg>
In theory in should be much faster than ltsp (video etc), but it has to be working first :)
16:19
<otavio>
alkisg: yes; for multiseat to work a small set of well tested hardware needs to be used.
16:19
alkisg: you can't just buy a new hardware and expect it to just work
16:19
alkisg: people fail to understand it
16:20
alkisg: and I'm Brazilian :P
16:20
<alkisg>
Heh... :) Are there any improvements now with the new xorg and xephyr?
16:20* ogra curses brazil ...
16:20* ogra is trying to hire some new mobile devs
16:21F-GT has joined #ltsp
16:21
<ogra>
but no way in brazil where the cleverest mobile guys seem to sit
16:21
since you godeammed cant get the HW past customes
16:21
<otavio>
alkisg: the current solutions usually uses xephyr since it is avoids a lot of complexity
16:21
alkisg: but still depends on hardware stability and drivers quality
16:21
ogra: :-D
16:22
<ogra>
your country is shooting itself in the foot :(
16:22
<otavio>
ogra: i'm in brazil :P I'm not that clever but :P
16:22
<stgraber>
yeah, new chroot built, took an hour thanks to overloaded archives :)
16:22
<ogra>
otavio, i would take you :)
16:22
but still
16:23
<otavio>
ogra: hehe
16:23
<ogra>
your customs office stops me from doing o
16:23
*so
16:23
<alkisg>
otavio, I've read that the new xorg has "GLX and DRI passthrough support for Xephyr", so it supports hw acceleration on all screens now?
16:24
<warren>
oh really?
16:24
<otavio>
alkisg: I've didn't try it
16:24
<warren>
how much performance is lost?
16:24
<otavio>
alkisg: it would be awesome
16:24
<ogra>
not much
16:24
i heard it works quite well at least for composite
16:24
<alkisg>
otavio, too bad that nvidia doesn't support more than 2 accelerated surfaces since 8800... Damn directx 10! :(
16:24
<otavio>
ogra: if it works for composite it should work for most, if not all, other usages
16:25
alkisg: hahah
16:25
<ogra>
yeah
16:25
<alkisg>
otavio, do you have a page/wiki/something with proposed hardware?
16:25
<otavio>
alkisg: nvidia sucks!
16:25
<ogra>
++
16:25
<otavio>
alkisg: no; I don't
16:25* ogra recently thinks *intel* sucks
16:25
<otavio>
ogra: really? why?
16:25
<ogra>
at leat since i wokr with mobile stuff
16:25
<otavio>
ahahah
16:26
<ogra>
their puolbo driver is as closed as nvidia
16:26
<alkisg>
Ah, talking about graphic cards, which cards do you guys propose for ltsp? I thought intel would be a right choice...
16:26
<ogra>
*poulsbo
16:26
and only works with DRI for kernel 2.6.24 and xorg .4
16:26
*1.4
16:27
<stgraber>
alkisg: Intel works fine here (Intel chipsets + Atom CPU)
16:28
<ogra>
everything the intel driver supports is unbeaten under linux
16:28
<alkisg>
stgraber, can you e.g. run decently googleearth with ltsp (and gigabit network and descent clients etc)?
16:28
<warren>
In Japan they had HP thin clients... AMD Sempron with some kind of ATI Radeon cards
16:28
<ogra>
definately
16:28
<warren>
They were fully supported by the open drivers
16:28
<ogra>
warren, free drivers ?
16:28
<warren>
3D and everything
16:28
yes
16:29
<ogra>
cool
16:29
<warren>
I was told that that particular ATI card was NOT supported in 3D with an Intel cpu
16:29
only AMD
16:29
<ogra>
yeah, the latest free one is already supporting a bunch
16:29
heh
16:29
so that's the path they go now ?
16:30
<warren>
I'm not sure why
16:30
<ogra>
keeping your revenue up ?
16:30
<warren>
I don't know
16:30
<ogra>
would be my guess
16:31
<stgraber>
alkisg: I didn't try googleearth but if there is a problem with it it won't be the graphic card
16:31
<alkisg>
stgraber, ty
16:31
<ogra>
i mean ... you wont find an intel GM965 on a AMD board either
16:32chrisinajar is now known as chrisinajar|work
16:34Egyptian[Home] has joined #ltsp
16:34twinprism has quit IRC
16:49F-GT has quit IRC
16:52alkisg has left #ltsp
16:56warren has quit IRC
17:02FuriousGeorge has joined #ltsp
17:02
<FuriousGeorge>
hey all
17:03
is there a preferred vnc sever to use with ltsp?
17:04F-GT has joined #ltsp
17:06
<FuriousGeorge>
i need some means to share the desktop of a logged in user on a client
17:09twinprism has joined #ltsp
17:11
<Lns>
FuriousGeorge: iTALC is pretty much the standard these days for LTSP
17:12F-GT has quit IRC
17:16
<Eeyore-Jr>
how does ltsp comply with CIPA for schools and libraries?
17:16
<Appiah>
CIPA as in?
17:16
(means several things to me)
17:16
<Eeyore-Jr>
Child Internet Protection Act of the USA
17:17
<Appiah>
oh
17:17
<Eeyore-Jr>
the ablity to block pron sites, chat, im, etc.
17:17
<Appiah>
Well since im not in the USA
17:17
well I dont see how blocking websites is a ltsp thing really
17:17
you can block sites in Linux
17:17
like any other OS
17:18
I mean dont your firewall handle that kind of thing?
17:23
<Lns>
Eeyore-Jr: Appiah is right - LTSP has nothing to do with content filtering, it's a thin-client architecture that sits on top of an existing Linux distribution. You external tools to do content filtering.
17:23
Commonly a firewall/proxy does this task.
17:24
<Appiah>
I do understand if you want the LTSP to act as a firewall (in a extreme small envoirment)
17:24
<Eeyore-Jr>
ur right. i was thinking of edubuntu
17:25
<Appiah>
still linux in general is capable of blocking sites
17:25
there are several firewall guis are available to do this
17:25
In a bigger network , I dont think it would be any problem
17:26
<Eeyore-Jr>
yes, i'm aware of this and am implementing this
17:26Guaraldo has quit IRC
17:26
<johnny>
people use dansguardian as well
17:27* Lns hearts ipcop + advproxy + urlfilter
17:27F-GT has joined #ltsp
17:28Lns has left #ltsp
17:28Lns has joined #ltsp
17:36Q-FUNK has joined #ltsp
17:39Blinny has joined #ltsp
17:42Blinny has quit IRC
17:42F-GT has quit IRC
17:57bobby_C has quit IRC
17:57staffencasa has quit IRC
17:57leio_ has joined #ltsp
17:58leio has quit IRC
17:58leio_ is now known as leio
17:59F-GT has joined #ltsp
18:13jammcq has joined #ltsp
18:13
<jammcq>
howdy
18:16
<FuriousGeorge>
Lns: delayed thanks for the suggestion
18:17
<Lns>
FuriousGeorge: no prob.
18:20mccann has quit IRC
18:27sbalneav has quit IRC
18:27sbalneav has joined #ltsp
18:27
<jammcq>
Scotty !!!!!!!!!!!!
18:30Q-FUNK has quit IRC
18:30Q-FUNK has joined #ltsp
18:39Faithful has quit IRC
18:40
<Lns>
!s
18:40
<ltspbot`>
Lns: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
18:40
<Lns>
:p
18:41
<Ryan52>
!r
18:41
<ltspbot`>
Ryan52: Error: "r" is not a valid command.
18:41
<Ryan52>
:(
18:41* Ryan52 feels so left out :P
18:42
<Lns>
lol
18:42
!moo
18:42
<ltspbot`>
Lns: Error: "moo" is not a valid command.
18:43
<Lns>
aww, come on
18:43
<|Paradox|>
moo: os: Microsoft Windows XP Professional - Service Pack 2 (5.1.2600) up: 3days 5hrs 46mins 44secs cpu: AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ at 2204MHz (7% Load) gfx: NVIDIA GeForce 6150 LE 512MB res: 1280x960 32bit 72Hz ram: 3452/3454.48MB (99.93%) hdd: C:\ 4.98GB/217.7GB Free E:\ 21.98GB/372.61GB Free F:\ 8.2GB/111.79GB Free G:\ 6.82GB/465.76GB Free H:\ 1.92GB/74.13GB Free I:\ 909.87MB/6.26GB Free net: NVIDIA nForce Netwo
18:43
<Lns>
uhh
18:43
huh?
18:43* Lns runs
18:43
<Ryan52>
heh
18:44
<Lns>
looks like |Paradox| has a lot of hdd space here on irc...i wonder why.....hmmmmmmm...
18:47johnny has left #ltsp
18:47Q-FUNK has quit IRC
18:49johnny has joined #ltsp
18:52pleed_ has quit IRC
19:02* Lns waves goodnight and happy Halloween to #ltsp
19:02Lns has quit IRC
19:23Nubae has joined #ltsp
19:49Nubae has left #ltsp
20:14otavio has quit IRC
20:22
<sbalneav>
Evening all
20:37
<stgraber>
evening sbalneav
20:39
<jammcq>
Scotty !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
20:39* vagrantc grumbles about a typo in the Debian plugins ...
20:52ogra has quit IRC
21:06
<vagrantc>
i'm still getting the free errors on ldm 2.0.16
21:12ogra has joined #ltsp
21:12
<Ryan52>
vagrantc: that's warren's fault. I told him to test that that actually fixed it.
21:12
"21:10 < Ryan52> if that doesn't work, then just comment out the g_free that's removed in the patch, test to make sure that works, and be done with it."
21:13
<vagrantc>
heh
21:13
and nowhere to be seen, this warren.
21:20kobpzzd has joined #ltsp
21:20|Paradox| has quit IRC
21:20kobpzzd is now known as |Paradox|
21:36shogunx has quit IRC
21:42shogunx has joined #ltsp
22:15captain_1agnus has joined #ltsp
22:17Eeyore-Jr has quit IRC
22:17Eeyore-Jr has joined #ltsp
22:18captain_magnus has quit IRC
22:19captain_1agnus is now known as captain_magnus
22:26gonzaloaf_laptop has quit IRC
22:31FuriousGeorge has quit IRC
22:35davidj has joined #ltsp
22:35
<jammcq>
davidj: well.... ?
22:35
<davidj>
jammcq: I'll be coming.
22:35
<jammcq>
AWESOME !!!!!
22:35
<davidj>
I get into Logan at noon.
22:35
<jammcq>
hmm
22:35
have you had any contact with Ragnar?
22:36
<davidj>
I'll get a car. I haven't heard from Ragnar yet.
22:36
<jammcq>
hmm
22:36
hey, don't forget your drivers license
22:36
<davidj>
:)
22:36kleewyck has quit IRC
22:36
<davidj>
My flight out is at 6am Monday, I'm staying at one of the airport hotels.
22:37
..Sunday night.
22:37
<jammcq>
wow
22:37
6am is awful early
22:37
well, I'm very glad you were able to arrange the travel so you can be there
22:38
<davidj>
Yeah, well, $69 for a 6am flight that gets me home for lunch, or $2,355 for a 10am flight that get me home after supper.
22:38
<jammcq>
hmm
22:38
I know that was a tough decision
22:38
<davidj>
I still think the airlines are yanking my chain.
22:38
heh
22:38
I like my sleep.
22:38
<jammcq>
and there are people who will actually pay 2355
22:38
<davidj>
Obviously, or they wouldn't be asking it.
22:39
The worst part is that $2355 is for coach, first class is more.
22:39
plus you have to pay $20 for an exit-row seat.
22:39
<jammcq>
I can fly to brazil first class for not much more than that
22:39
<davidj>
I'll bet any of us could.
22:40
<jammcq>
yep
22:40
<davidj>
Right now, short internal flights are probably tough on the airlines' budgets.
22:40
<jammcq>
so if you are gonna have a car, then maybe ragnar can ride with you, eh?
22:40
<davidj>
ISTR that it takes more fuel to take off and get to altitude than it does to fly for two hours.
22:41
Yes, I figured I'd wait for Ragnar.
22:41
Any idea when his flight gets in?
22:41
I'd hate to say I'll wait, and then find out he gets in at midnight :)
22:41
<jammcq>
seemed like he said around 3pm, but i'm not sure
22:42
<davidj>
3pm sounds like when he got in the last time.
22:42
<jammcq>
I think he was coming from florida or NYC
22:42
i'm pretty sure he's already in the states
22:42
<davidj>
but I don't know for sure that he's coming from the same place.
22:42
<jammcq>
and there was some chance that his plans would change
22:43
<davidj>
k
22:48
'night, everyone
22:48davidj has quit IRC
22:53gonzaloaf_laptop has joined #ltsp
22:57gonzaloaf_laptop has quit IRC
22:59gonzaloaf_laptop has joined #ltsp
23:01CaScAdE^1arAway has joined #ltsp
23:19CaScAdE^FarAway has quit IRC
23:52keithclark has joined #ltsp