IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 12 December 2006 (all times are UTC)
| 00:06 | <tomcats> exit
| |
| 00:06 | \q
| |
| 00:06 | tomcats has quit IRC | |
| 00:33 | vagrantc has quit IRC | |
| 00:56 | GiantPickle has quit IRC | |
| 00:56 | GiantPickle has joined #ltsp | |
| 01:28 | MoL0ToV has joined #ltsp | |
| 02:06 | <shogunx> is lbus not being used in ltsp5/ubuntu muekow?
| |
| 02:06 | for localdev that is?
| |
| 02:07 | looks like it requires ldm's ssh keys method to work.
| |
| 02:08 | while I am using gdm...
| |
| 02:09 | jammcq?
| |
| 02:30 | Avatara has joined #ltsp | |
| 02:32 | Avatara has quit IRC | |
| 02:55 | weeswordy has left #ltsp | |
| 03:00 | MoL0ToV has quit IRC | |
| 03:07 | Avatara has joined #ltsp | |
| 03:08 | ogra has joined #ltsp | |
| 03:11 | psycodad has quit IRC | |
| 03:11 | psycodad has joined #ltsp | |
| 03:17 | MoL0ToV has joined #ltsp | |
| 04:04 | bobby_C has joined #ltsp | |
| 04:15 | Darke- has joined #ltsp | |
| 04:16 | <Darke-> Hi, I got my ltsp server up and running, got a working client. I'v mounted the remote filesystem.. X starts up with GDM and everyting works fine.
| |
| 04:16 | But, i cannot log in on the client.
| |
| 04:17 | I am running SLES10 on the server.
| |
| 04:17 | Any toughts ?
| |
| 04:18 | sshd is running and working like normal.
| |
| 04:21 | Avatara has quit IRC | |
| 04:33 | <shogunx> hey, ogra!
| |
| 04:34 | i have a edgy localdev question.
| |
| 04:36 | have the lbus daemons been abandoned or integrated into ltspfs and ltspfsd? i noticed that the udev scripts pass ltspfsmounter commands through a socket, but that only works with ssh-keys seup, correct?
| |
| 04:37 | i'm working a patch now that will work with non ssh tunneled(xdmcp) sessions.
| |
| 04:37 | <Darke-> shogunx: you seem to be the only person not afk here, you might be able to help me ?
| |
| 04:37 | <shogunx> its using the lbussd and lbuscd daemons, if you would like it.
| |
| 04:37 | whats up Darke-
| |
| 04:38 | <Darke-> read my questions above :) thin you can help ?
| |
| 04:38 | <shogunx> you are using ldm, correct?
| |
| 04:39 | <Darke-> hmm
| |
| 04:39 | <shogunx> oh... gdm
| |
| 04:40 | <Darke-> gdm yes
| |
| 04:40 | <shogunx> what happens when you log in?
| |
| 04:41 | <Darke-> Well, atm i dont use a "real" thin client. I boot a normal machine of a ROM floppy, it loads kernel from the server and mounts the filesystem (/opt/lts/i386)
| |
| 04:41 | and starts x with gdm
| |
| 04:41 | but i cannot log in
| |
| 04:41 | read something about ltsp-update-sshkeys, but i got no such package.
| |
| 04:42 | <shogunx> what happens when you log in?
| |
| 04:42 | <Darke-> well incorrect user/pass
| |
| 04:43 | so i cant log in.
| |
| 04:43 | <shogunx> are you trying to log in as root?
| |
| 04:44 | what happens when you log in from the console with the same user/pass? from a console x session?
| |
| 04:44 | <Darke-> tryed with every user that excists on the ltsp server
| |
| 04:45 | sorry, on the phone :P
| |
| 04:45 | sec
| |
| 04:45 | <shogunx> caps lock?
| |
| 04:47 | <ogra> shogunx, please coordinate with sbalneav on that, he was switching the whole communication layer to xatoms so the underlying system shouldnt matter anymore
| |
| 04:47 | but i dont know where he got with it
| |
| 04:49 | <shogunx> i was thinking just another else in the /lib/udev/ scripts that made the appropriate calls to /tmp/lbus.fifo
| |
| 04:49 | will do ogra
| |
| 04:50 | <ogra> great, thanks :)
| |
| 04:50 | he wanted to have something ready for feisty ...
| |
| 04:50 | <Darke-> ok back
| |
| 04:50 | no its not a typo problem
| |
| 04:51 | <ogra> apart from that we're still discussing an unencrypted ssh tunnel that should behave like xdmcp but wouldnt need the overhead ... and you have at least the key handshake
| |
| 04:51 | <shogunx> ack. this is workign much better than my nasty kludge to make it work on dapper... the 2 second umount is actually working.
| |
| 04:51 | <Darke-> ogra: im using xdmcp
| |
| 04:51 | but i cant just log in, on the client.
| |
| 04:52 | <ogra> well, its not much supported in ubuntu ltsp5
| |
| 04:52 | (yet)
| |
| 04:52 | <Darke-> im not using ubuntu :)
| |
| 04:52 | <ogra> since its evil and incesure ...
| |
| 04:52 | right, but we were talking about ubuntu ;)
| |
| 04:52 | <Darke-> the ltsp server is running SLES10
| |
| 04:52 | ah
| |
| 04:52 | ok ;)
| |
| 04:53 | <ogra> suse didnt join the bandwagon yet (unlike gentoo and redhat) they will likely miss out on ltsp5
| |
| 04:54 | <Darke-> and what is this ltsp-update-sshkeys that i read about ppl using on the forums. I got ltsp-utils, but i only got ltspcfg ltspadmin ltspinfo
| |
| 04:54 | <ogra> its a part of the ltsp5 implementation
| |
| 04:54 | <Darke-> ah >.<
| |
| 04:54 | <ogra> you wont find it in 4.X
| |
| 04:54 | <shogunx> ogra i kind of figured... it was a bear on dapper, but getting there with edgy. as for suse, novell joined the wrong bandwagon for them, i think.
| |
| 04:55 | <Darke-> but i should be able to fix this still ? i mean, i got all working.. Should not be impossible to fix the login.
| |
| 04:55 | <ogra> right, but thats not ltsp related :)
| |
| 04:55 | <shogunx> ogra true
| |
| 04:55 | <ogra> we invited them to the ltsp hackfest in detroit in september where we planned the implementation for other distros than ubuntu/debian
| |
| 04:56 | <shogunx> they did not come? thats a mistake.
| |
| 04:57 | <Darke-> ogra: hehe ;) true, but you dont think you could give me a few toughts ?
| |
| 04:57 | could really use some help here ;P im to close to stop now :)
| |
| 04:58 | <shogunx> in any case, i'll have this lbus interface cleaned up in udev tonight or tomorrow... ill kick it over to scotty and see if he wants to use it.
| |
| 04:58 | thanx ogra
| |
| 05:00 | Darke- do you have a proper /etc/X11/Xsession or /etc/gdm/Xsession on the server?
| |
| 05:01 | or is it telling you bad password?
| |
| 05:01 | <Darke-> yes
| |
| 05:01 | i get the gdm login promt on the client
| |
| 05:01 | but it tells me either wrong user or pass or both ;P uno the standard stuff :)
| |
| 05:02 | <shogunx> are you sure you are connected to the right server via xdmcp?
| |
| 05:02 | <Darke-> O_O
| |
| 05:03 | if i was not ? would it work at all ?
| |
| 05:03 | <shogunx> yes
| |
| 05:03 | <Darke-> there is no other server to connect to
| |
| 05:03 | ogra has quit IRC | |
| 05:03 | <Darke-> hmm
| |
| 05:04 | <shogunx> i don;t know what to tell you, man. i've never seen that happen.
| |
| 05:04 | <Darke-> anything you want me to check ?
| |
| 05:06 | <shogunx> you can login via gdm from the console of the box?
| |
| 05:06 | the actual server that is?
| |
| 05:08 | if so, then perhaps /etc/gdm.conf might need a bit of tweaking.
| |
| 05:09 | <Darke-> hmm
| |
| 05:09 | you asking if i can login on the actual server ? ;P
| |
| 05:09 | thrue gdm
| |
| 05:09 | yes
| |
| 05:13 | MrMoo has joined #ltsp | |
| 05:13 | ogra has joined #ltsp | |
| 05:13 | Avatara has joined #ltsp | |
| 05:17 | <Darke-> works
| |
| 05:17 | ...
| |
| 05:17 | gdm.conf needed some work :)
| |
| 05:17 | Darke- has quit IRC | |
| 05:18 | Avatara has quit IRC | |
| 05:19 | Avatara has joined #ltsp | |
| 05:22 | Jenna has joined #ltsp | |
| 05:37 | Avatara has quit IRC | |
| 05:39 | Avatara has joined #ltsp | |
| 06:04 | jsgotangco has joined #ltsp | |
| 06:07 | egypcio has quit IRC | |
| 06:07 | FernandoMM has joined #ltsp | |
| 06:08 | egypcio has joined #ltsp | |
| 06:11 | egypcio has quit IRC | |
| 06:17 | vykarian has joined #ltsp | |
| 06:17 | <vykarian> hey guys
| |
| 06:17 | jammcq there?
| |
| 06:17 | does someone know where can I find a .config for 2.4 kernel series
| |
| 06:17 | ?
| |
| 06:19 | <cliebow> isnt in the kernel kit?
| |
| 06:20 | <vykarian> cliebow just 2.6 series =\ (at least the last version)
| |
| 06:22 | <cliebow> lemme look aaround..i must have one aaround somewhere
| |
| 06:22 | <vykarian> :):)
| |
| 06:22 | <cliebow> im off to school
| |
| 06:22 | <vykarian> the 2.6 has problems for dosemu ipx devices
| |
| 06:22 | <cliebow> jammcq should be around by aand by 8`)
| |
| 06:23 | <vykarian> already tried to hack out the dosemu but prefer downgrade the kernel
| |
| 06:23 | <cliebow> yeah..bbiaab
| |
| 06:24 | egypcio has joined #ltsp | |
| 06:33 | * msinhore is back (gone 11:27:07) | |
| 07:02 | yo has joined #ltsp | |
| 07:03 | <yo> bisa bahasa indonesia
| |
| 07:04 | yo has quit IRC | |
| 07:06 | xs4rg7tn has joined #ltsp | |
| 07:08 | jammcq has quit IRC | |
| 07:16 | Jenna has quit IRC | |
| 07:18 | xs4rg7tn is now known as help | |
| 07:18 | help is now known as xs4rg7nt | |
| 07:23 | <cliebow_> 2 :q
| |
| 07:23 | :q
| |
| 07:25 | <xs4rg7nt> hi every one im feel so cool this morning
| |
| 07:28 | vykarian has quit IRC | |
| 07:30 | Skarmeth has joined #ltsp | |
| 07:38 | bjohnson has joined #ltsp | |
| 07:44 | andresmujica has joined #ltsp | |
| 07:45 | <cliebow_> gonna be one of those days
| |
| 07:47 | jsgotangco has quit IRC | |
| 07:52 | J45p3r_ has joined #ltsp | |
| 08:04 | egypcio has quit IRC | |
| 08:04 | egypcio has joined #ltsp | |
| 08:07 | jammcq has joined #ltsp | |
| 08:07 | <jammcq> howdie
| |
| 08:13 | FaithX has quit IRC | |
| 08:13 | Faithful has quit IRC | |
| 08:14 | Faithful has joined #ltsp | |
| 08:37 | Avatara has quit IRC | |
| 08:38 | Avatara has joined #ltsp | |
| 08:39 | Sharmanshik has joined #ltsp | |
| 08:42 | Sharmanshik has quit IRC | |
| 08:42 | <chupacabra> yo
| |
| 08:44 | <xs4rg7nt> jimm
| |
| 08:44 | how are you
| |
| 08:44 | i have a ltsp network with about 20 clients
| |
| 08:45 | and 15 work fine but 5 clients restart X every 40 minutes
| |
| 08:45 | or so so
| |
| 08:47 | Stevec has joined #ltsp | |
| 08:48 | Faithful has quit IRC | |
| 08:49 | Faithful has joined #ltsp | |
| 08:52 | Fingahzz has joined #ltsp | |
| 08:58 | <Fingahzz> Morning! Anyone familiar w/ fstab ;-)
| |
| 08:58 | <-- has a really *silly* question.
| |
| 09:10 | Stevec has quit IRC | |
| 09:12 | jernst has joined #ltsp | |
| 09:13 | andresmujica has left #ltsp | |
| 09:14 | xs4rg7nt has quit IRC | |
| 09:16 | andresmujica has joined #ltsp | |
| 09:18 | <cliebow_> xs4rg7nt:how much memory in the ones that crash?
| |
| 09:22 | Avatara has quit IRC | |
| 09:26 | Avatara has joined #ltsp | |
| 09:32 | Avatara has quit IRC | |
| 09:33 | andresmujica has quit IRC | |
| 09:34 | andresmujica has joined #ltsp | |
| 09:41 | bricode has joined #ltsp | |
| 09:42 | <bricode> ogra: Got that workstation script to work quite nicely.
| |
| 09:42 | <ogra> cool
| |
| 09:42 | but mind you, mounting /home locally will quickly run you out of mem, you need an nfs mounted /home and something like ldap
| |
| 09:42 | <bricode> ogra: It's funny because you have to turn the sound off in the lts.conf file for it to work properly client side.
| |
| 09:43 | <ogra> yeah
| |
| 09:43 | thats something i need to fix in the ltsp-client initscript ...
| |
| 09:43 | <bricode> I've mounted everything on the NFS side.
| |
| 09:43 | <ogra> it should just ignore certtain things in a workstation env
| |
| 09:43 | <bricode> Also xorg was a little weird because I wanted a custom one for decent video drivers.
| |
| 09:44 | Took me forever to track down that I needed an lts.conf file, and where said file was hidden, and that it should be in /etc/
| |
| 09:44 | <ogra> thast right, the xorg autogeneration wont be touched
| |
| 09:45 | <bricode> But now I've got a "fat" client that does sound and video. Google maps runs like a champ. The only thing that sucks now is Flash.
| |
| 09:45 | <ogra> and we'll ship an empty lts.conf in feisty i think, even its totally moot to have it, people seem to get very confused if its not there
| |
| 09:45 | vagrantc has joined #ltsp | |
| 09:45 | <ogra> vagrantc, freeze-congrats :)
| |
| 09:45 | <bricode> That would be a good idea.
| |
| 09:46 | <vagrantc> ogra: hopefully we can backport ltspfs without requiring an ltsp backport... :)
| |
| 09:46 | <ogra> well, i'd rather get the info across that lts.conf should only be used to actually override defaults
| |
| 09:46 | <bricode> Hopefully Adobe will implement some decent video acceleration in Flash 10...like say, an xvideo overlay implementation.
| |
| 09:46 | <ogra> vagrantc, did you keep the socket in ldm ?
| |
| 09:46 | or at least the socket code ...
| |
| 09:46 | <vagrantc> ogra: if localdev is enabled, yes.
| |
| 09:46 | <ogra> cool
| |
| 09:47 | so backporting shouldnt be a prob
| |
| 09:47 | <bricode> Is there a place where the defaults for ltsp are defined?
| |
| 09:47 | <ogra> i'm just waiting for sbalneav to check who fixes the package
| |
| 09:47 | <vagrantc> i think there was some code added to edgy specifically for localdev that we didn't get into etch.
| |
| 09:47 | added to ldm
| |
| 09:47 | <ogra> bricode, they are autodetected by the OS (in ltsp5)
| |
| 09:48 | in ltsp4.x you have to define some pieces in lts.conf and soem are autodetected ... and afaik 4.x doesnt even boot if no lts.conf is there but i might remember that wrongly
| |
| 09:50 | MoL0ToV has quit IRC | |
| 09:50 | <bricode> ogra: I don't have a problem with autodetection, but if you have to change it, there isn't much guidance on what happens and when during the boot process...
| |
| 09:51 | <ogra> well, udev and the kernel are supposed to detect HW, alsa detects sound cards and the ltsp-client initscript calls the autoconfig routine from xserver-xorg
| |
| 09:52 | so there is not much "ltsp autodetection" involved
| |
| 09:52 | its all done by the bits and pieces of the OS
| |
| 09:53 | some additional things are done in the ltsp-client-setup initscript ... actually thats the core of muecow ...
| |
| 09:54 | <bricode> ogra: maybe I'm just irritated because it didn't do a good job with the bizarre vid card :)
| |
| 09:54 | <ogra> what kind of card is that ?
| |
| 09:54 | <bricode> ogra: Everything else was pie.
| |
| 09:54 | ogra: It's the Geode LX on-chip stuff. I had to roll my own xorg drivers to get the video overlay stuff to work...
| |
| 09:55 | <ogra> ah, right, i'll care that they are updated until release ...
| |
| 09:55 | dunno where rodravus stands with the xorg updates atm
| |
| 09:55 | jernst has left #ltsp | |
| 09:56 | <ogra> but we discussed that already ... for unichrome and geode
| |
| 09:56 | <bricode> ogra: I've been in light contact with Jordan Crouse who said that he was too busy with Geode Gx stuff for the OLPC.
| |
| 09:56 | <ogra> well, doesnt that also use X ?
| |
| 09:56 | <bricode> ogra: So I'm thinking about adding to the amd drivers myself :)
| |
| 09:57 | ogra: The GX does, yes, but it's different enough that it requires it's own code tree...the LX branch has been rotting.
| |
| 09:57 | <ogra> hmm, thats bad
| |
| 09:57 | <bricode> ogra: Agreed.
| |
| 09:58 | ogra: So, I sort of want to talk to all the stakeholders, roll something, get feedback from them, and go from there.
| |
| 09:59 | <ogra> cool
| |
| 09:59 | <bricode> The LX chip is pretty powerful as long as you give it the right software to eat.
| |
| 10:00 | sbalneav has joined #ltsp | |
| 10:00 | * bricode wonders if these IRC sessions are recorded. | |
| 10:00 | <Megant__> surely
| |
| 10:00 | <ogra> sadly no
| |
| 10:01 | <sbalneav> Morning all
| |
| 10:01 | <ogra> sbalneav, is to lazy to set up logging for ltspbot :P
| |
| 10:01 | <sbalneav> Hey ogra
| |
| 10:01 | <bricode> sbalneav: 'morning.
| |
| 10:01 | <ogra> hey sbalneav !!
| |
| 10:01 | <sbalneav> what?!? It's logging! :)
| |
| 10:01 | It's just that we're not posting the logs anywhere :)
| |
| 10:01 | <ogra> sbalneav, vagrantc has some issues with ltspfs for debian ... we should clean up the package so he can backport it :)
| |
| 10:01 | ah :)
| |
| 10:02 | <sbalneav> I am getting bitten HARD by an ubuntu bug :(
| |
| 10:02 | <ogra> ouch
| |
| 10:02 | what is it ?
| |
| 10:02 | lts bite back ;)
| |
| 10:02 | *lets
| |
| 10:02 | <sbalneav> 62308
| |
| 10:02 | * vagrantc bares teeth | |
| 10:02 | <sbalneav> NFS server doesn't work :(
| |
| 10:02 | Fingahzz has quit IRC | |
| 10:02 | <vagrantc> bout time to do root over NBD, eh?
| |
| 10:03 | <sbalneav> It's not LTSP stuff. It's shared home dirs.
| |
| 10:03 | <jammcq> Scotty !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 10:03 | <sbalneav> Hey jammcq
| |
| 10:03 | <bricode> sbalneav: Shared home dirs in what sense?
| |
| 10:03 | <sbalneav> ok, lemme get bzr set up here on my box at work.
| |
| 10:04 | <ogra> meh, thats an evil one
| |
| 10:04 | <sbalneav> I've just got two LTSP servers, one serves /home to the other via nfs.
| |
| 10:05 | ogra: yeah, I tried moving to a 2.6.18.5 kernel earlier today. LDAP stopped authenticating. I'm gonna have to come in tonight and hack it through, see what's going on.
| |
| 10:05 | bricode has left #ltsp | |
| 10:05 | <ogra> did you try the feisty kernel (in an vmware or something for a test)
| |
| 10:06 | (thats 2.6.19)
| |
| 10:07 | <sbalneav> No I didn't. Can I safely install a feisty kernel on a edgy box?
| |
| 10:07 | I'm certainly willing to try.
| |
| 10:08 | <ogra> #57543 seems to have a patch for 2.6.17-1
| |
| 10:10 | <sbalneav> argh. I'm an idiot. How do I locally check out a copy of http://bazaar.launchpad.net/~sbalneav/ltspfs/ltspfs-upstream again using bzr?
| |
| 10:10 | * sbalneav hates this &(*!^*& project at work. | |
| 10:11 | <vagrantc> bzr get URL ??
| |
| 10:11 | <sbalneav> durr
| |
| 10:11 | * sbalneav slaps head with a herring | |
| 10:12 | <vagrantc> or if you prefer typing more, "bzr branch"
| |
| 10:13 | bricode has joined #ltsp | |
| 10:13 | <ogra> vagrantc, which kernel will etch use ?
| |
| 10:13 | <vagrantc> ogra: 2.6.18, i think.
| |
| 10:13 | ogra: just hit etch the other day.
| |
| 10:13 | <ogra> ah
| |
| 10:14 | <vagrantc> ogra: debian-installer still users .17 at the moment ...
| |
| 10:14 | * bricode wonders what etch is. | |
| 10:14 | Faithful has quit IRC | |
| 10:14 | * ogra hasnt even seen .18 ... ubuntu jumped to .19 right away | |
| 10:14 | <vagrantc> bricode: debian's next stable release
| |
| 10:14 | <ogra> (and will ship .20 at realease time)
| |
| 10:15 | Faithful has joined #ltsp | |
| 10:15 | <vagrantc> i've rarely seen huge differences in kernels. but that probably comes with using hardware that's 6-8 years old.
| |
| 10:15 | <ogra> lol
| |
| 10:15 | yeah
| |
| 10:16 | <vagrantc> though the .17 kernels in debian didn't work with serial consoles...
| |
| 10:16 | and the switch from devfs to udev was certainly annoying.
| |
| 10:16 | that
| |
| 10:16 | Envite has joined #ltsp | |
| 10:16 | <vagrantc> s about that only things i've noticed.
| |
| 10:16 | <Envite> Hello all from Spain
| |
| 10:16 | Hola a todos desde España
| |
| 10:17 | <cliebow_> Envite: Hello
| |
| 10:18 | <sbalneav> Okay. Checked out, configured, built. Whatdda I need to feeex
| |
| 10:18 | ?
| |
| 10:18 | <ogra> vagrantc, the switch from devfs to udev ?
| |
| 10:18 | <vagrantc> sbalneav: compare it agains the .orig.tar.gz from ubuntu
| |
| 10:18 | <sbalneav> okie
| |
| 10:18 | <ogra> thats more than a year ago for me :P
| |
| 10:18 | <Envite> Hello cliebow_
| |
| 10:18 | <vagrantc> ogra: yeah, that happened around 2.6.1(something)
| |
| 10:18 | <ogra> yep
| |
| 10:19 | <vagrantc> ogra: about the same here, but it wasn't somethig i would've expected in the 2.6 series ...
| |
| 10:19 | <ogra> sbalneav, drop the complete ltspfs postinst content please
| |
| 10:19 | <bricode> udev was a painful transition.
| |
| 10:19 | but worthwhile.
| |
| 10:19 | <ogra> fuse-utils cares for the module loading
| |
| 10:19 | <vagrantc> sbalneav: i also have a bzr branch for my changes to the packaging, though some things are debian-specific.
| |
| 10:19 | <sbalneav> OK, that's in the -packaging branch, I beleive, yes?
| |
| 10:20 | <ogra> and for ubuntu we should move the scripts we have no manpages for to /usr/share/ltspfsd
| |
| 10:20 | for the backport for vagranbt that would brak though
| |
| 10:20 | yeah, thats in the debian dir
| |
| 10:21 | scripts we have no manpages for -> see the lintian-override file
| |
| 10:21 | <vagrantc> sbalneav: http://llama.freegeek.org/~vagrant/bzr/ltspfs/vagrant-ltspfs-debian-packaging
| |
| 10:21 | efra has joined #ltsp | |
| 10:21 | <vagrantc> sbalneav: but don't just merge wholesale.
| |
| 10:21 | <efra> Hi everybody
| |
| 10:21 | * vagrantc goes to breakfast | |
| 10:21 | <ogra> vagrantc, if sbalneav fixes these issues you could sync the package via utnubu
| |
| 10:21 | <sbalneav> ok, give me a few minutes here, phone's rining off the hook :)
| |
| 10:22 | <ogra> they are supposed to do ubuntu->debian syncs
| |
| 10:24 | egypcio has quit IRC | |
| 10:25 | egypcio has joined #ltsp | |
| 10:26 | <ogra> so you wouldnt have to maintain it in two places ...
| |
| 10:26 | bobby_C has quit IRC | |
| 10:35 | <sbalneav> ogra: well, we don't want to drop the chmod 4755 of lbmount, do we?
| |
| 10:35 | <ogra> nope
| |
| 10:35 | <sbalneav> ok
| |
| 10:36 | <ogra> but all the module related stuff
| |
| 10:36 | <sbalneav> k, gotcha
| |
| 10:36 | <vagrantc> ogra: utnubu, eh?
| |
| 10:37 | <ogra> vagrantc, yes that project that exists since 18months ...
| |
| 10:37 | <vagrantc> ogra: yes, i'm aware of it.
| |
| 10:37 | <ogra> nobody in debian seems to know about it ...
| |
| 10:37 | or to use it
| |
| 10:37 | <vagrantc> ogra: i recall there was a lot of controversey when it first started.
| |
| 10:37 | <ogra> well, they are donig some packages ...
| |
| 10:38 | <vagrantc> canonical didn't like that the name was so similar to ubuntu
| |
| 10:38 | <ogra> dmraid and teatime applet are two from the top of my head :)
| |
| 10:38 | right
| |
| 10:38 | <vagrantc> bzrtools, as well.
| |
| 10:38 | <ogra> but they kept it and we stopped complaining :)
| |
| 10:40 | <vagrantc> we'll see how many debian-specific differences it requires for ltspfs ... if it's just maintainer/uploaders fields ... i don't really see any reason to maintain it in another project
| |
| 10:40 | <ogra> right
| |
| 10:40 | i'd like to get it there
| |
| 10:41 | <vagrantc> we've got a long ways to go for ltsp...
| |
| 10:41 | <ogra> maintainer fileds are automatically rewritten by LP now
| |
| 10:41 | not really
| |
| 10:41 | the differentces get less and less :)
| |
| 10:41 | <vagrantc> i've only see the difference increase since breezy
| |
| 10:41 | <ogra> i just didnt have time in edgy to sort them all
| |
| 10:41 | yes, so lets fix that ;)
| |
| 10:41 | <vagrantc> ok! :)
| |
| 10:42 | <ogra> i.e. send me that list with changes you want and i'll make them :)
| |
| 10:42 | apart from the control file we should be on par with the feisty release i hope
| |
| 10:43 | <vagrantc> i'm hoping the spanish crew gets it's act together and submits a new spanish translation. it would be a shame to have 3/4ths of a translation done and not make it into etch.
| |
| 10:43 | <sbalneav> Oh, Lordy. 2.6.19-7 bigiron kernel panics right away :(
| |
| 10:43 | Woe is me
| |
| 10:44 | <bricode> sbalneav: I'm starting to think that you like it when things don't work...gives you something to do :)
| |
| 10:44 | <vagrantc> ogra: i still don't like the changes to ltsp-client-builder... it's soooo... hard-coded. :P
| |
| 10:44 | <ogra> make me that list :P
| |
| 10:44 | i'll change what i can :)
| |
| 10:45 | ltsp-client-builder is in the process of being partially rewritten, so i can look into it
| |
| 10:45 | <sbalneav> bricode: While I'm working on this NFS issue, I'm *NOT* working on LTSP, so, no, I don't like these kinds of problems :)
| |
| 10:46 | <ogra> it depends *where* things stop working :)
| |
| 10:46 | <sbalneav> Who's the kernel guy in devel?
| |
| 10:46 | <ogra> if its the server that generates your breadrolls, thats never funny
| |
| 10:47 | BenC and kylem
| |
| 10:47 | <sbalneav> Can I disturb him with a question? I know this is kind of a "supporty" thing :)
| |
| 10:48 | <ogra> sure
| |
| 10:48 | refer to me pointing you to him since its important for edubuntu that NFS /home works in feisty;)
| |
| 10:49 | <jammcq> breadrolls ?
| |
| 10:50 | <ogra> sure :)
| |
| 10:50 | and cheese indeed :)
| |
| 10:50 | <jammcq> and beer
| |
| 10:50 | mmmm yummy
| |
| 10:50 | <ogra> cheese makes everything better :)
| |
| 10:50 | <jammcq> indeed
| |
| 10:58 | Skarmeth has quit IRC | |
| 10:58 | Megant_ has joined #ltsp | |
| 11:00 | <vagrantc> ogra: we should probably implement a get_bool funcion in shell for the init scripts.
| |
| 11:00 | J45p3r_ has quit IRC | |
| 11:01 | <ogra> yeah
| |
| 11:01 | didnt we plan that in detroit ?
| |
| 11:01 | egypcio has quit IRC | |
| 11:01 | <vagrantc> looking at your lts-parameters.txt, i think your valid values for booleans are currently mosly lies :P
| |
| 11:01 | <ogra> heh, right, but they should give users a hand
| |
| 11:02 | * vagrantc thinks incorrect documentation can almost be worse that no documentation. | |
| 11:02 | <ogra> anyway, thin-client-configurator should make that all obsolete
| |
| 11:02 | <vagrantc> bah.
| |
| 11:03 | if i can't edit files with a text editor, it's useless.
| |
| 11:03 | <ogra> hey, i just redesigned the code, so you can have a curses interface
| |
| 11:03 | <vagrantc> that's nice, but i still prefer a text editor :P
| |
| 11:03 | <ogra> oki, i'll add the --hardcore switch that fires up $EDITOR
| |
| 11:03 | <cliebow_> exit
| |
| 11:04 | reenter
| |
| 11:04 | <vagrantc> ogra: probably not necessary, as long as the configuration file is sane. :P
| |
| 11:04 | <ogra> it will
| |
| 11:04 | Megant__ has quit IRC | |
| 11:04 | <ogra> boolean testing is on my list since detroit ...
| |
| 11:05 | i just didnt have the time .... edgy was way to short for all that extra stuff
| |
| 11:05 | <vagrantc> it shouldn't be all that hard to implement.
| |
| 11:05 | a simple case statement...
| |
| 11:05 | <ogra> right
| |
| 11:05 | <vagrantc> and i'll fix up the python variant, too.
| |
| 11:06 | <ogra> to something that also works in ubuntu ?
| |
| 11:06 | :)
| |
| 11:06 | <vagrantc> yes.
| |
| 11:06 | <ogra> cool
| |
| 11:06 | Megant_ has quit IRC | |
| 11:06 | Megant_ has joined #ltsp | |
| 11:06 | <vagrantc> i just rewrote some other program in python, and it gave me a more elegant idea for how to fix it that should be backwards compatible to python 1.5
| |
| 11:06 | <ogra> apart from fat-clients and the pulseaudio switch i dont really have ltsp specs this time
| |
| 11:07 | <vagrantc> nbd stuff?
| |
| 11:07 | <ogra> thats all there already ...
| |
| 11:07 | i merged all your code and threw mine away
| |
| 11:07 | <vagrantc> it still would be better to implement in nbd itself
| |
| 11:07 | <ogra> just need to get cryptsetup to main
| |
| 11:07 | right
| |
| 11:08 | <vagrantc> and get the timeout stuff working properly... though i thought of a simple nbd-ping thing we could run.
| |
| 11:08 | <ogra> right
| |
| 11:08 | thats what i thought as well
| |
| 11:08 | integrated into nbdswapserver
| |
| 11:08 | <vagrantc> basically just reading a random 1k from the raw device every now and then...
| |
| 11:09 | gah... y'all changed lots of stuff in ltspfs. looks like a backport will require a fair amount of work to be compatible with ltsp in debian :(
| |
| 11:10 | <ogra> it shouldnt
| |
| 11:10 | <vagrantc> binaries moved from /bin to /sbin ... additional code blocks in ldm ...
| |
| 11:10 | <ogra> ltsp-client-setup should have the LOCALDEV condition
| |
| 11:10 | oh207, hmm
| |
| 11:11 | <vagrantc> i really wish there had been more people actively working on it in debian.
| |
| 11:11 | <ogra> we could change that for a backport ...
| |
| 11:11 | <vagrantc> yeah
| |
| 11:12 | and i feel like a total IDIOT for not implementing a brain-dead simple ltsp-chroot. could of at least had one that sets the environment variables so that daemons don't get started.
| |
| 11:12 | <ogra> well, next time
| |
| 11:12 | <vagrantc> i had my eyes set on something fancier, and i missed the obvious simpler, more achievable goals.
| |
| 11:12 | <ogra> i wanted to look at ltsp-chroot for upgrades etc anyway
| |
| 11:13 | <vagrantc> it should set an environment variable, optionally respect $ROOT (which we should probably change to $LTSP_ROOT), and the fancy stuff i spent too much time on was mounting all the stuff in the chroot.
| |
| 11:14 | <ogra> $LTSP_ROOT sounds good
| |
| 11:14 | i'll try to use that in ltsp-manager as well
| |
| 11:14 | <vagrantc> i think the gentoo folks were worried that it would conflict with a variable they used. and it is a little too generic.
| |
| 11:14 | <ogra> thin-client-configurator (sorry)
| |
| 11:14 | redguy_work has quit IRC | |
| 11:15 | <vagrantc> well, now that etch is frozen, i can look at working on a backport :)
| |
| 11:15 | redguy_work has joined #ltsp | |
| 11:15 | <ogra> for ubuntu i'm planning full update-manager integration anyway
| |
| 11:15 | <vagrantc> all the things i've been putting off :)
| |
| 11:15 | <ogra> right, same here ... just that i'm at the beginning of a schedule
| |
| 11:15 | * ogra wonders if debian has update-manager at all | |
| 11:16 | <vagrantc> ogra: you know, i always know there's some phase of a cycle you are in that prevents you from doing anything, so you don't need to explain exactly which phase :P
| |
| 11:16 | * vagrantc has no idea what update-manager is | |
| 11:16 | <ogra> the dist-upgrade tool we use in ubuntu
| |
| 11:18 | * vagrantc applies for a tty operator job | |
| 11:19 | <str4nd> Megant_: ask about ctrl+alt+f7 :)
| |
| 11:19 | i just can't handle it :(
| |
| 11:20 | <vagrantc> ogra: i'm wondering if we shouldn't do some sort of checking for SCREEN_NN that conflict with stuff that init and/or upstart are running ..
| |
| 11:20 | <ogra> not really
| |
| 11:20 | Avatara has joined #ltsp | |
| 11:20 | <ogra> upstart should care if you want to access a tty
| |
| 11:20 | <vagrantc> anyone coming from a ltsp 4.x background will get burned by it.
| |
| 11:20 | <ogra> it will get the event ...
| |
| 11:21 | <vagrantc> ogra: so ltsp-client init scripts should do some upstart event?
| |
| 11:21 | <ogra> hmm ?
| |
| 11:21 | why should they get burned by it ?
| |
| 11:21 | it works here
| |
| 11:22 | <vagrantc> if upstart is configured to run something on tty1, and they set SCREEN_01=shell ... what happens?
| |
| 11:22 | <ogra> the only tty init/upstart use is tty1 as it was with sysvinit
| |
| 11:22 | <str4nd> whai our ltsp 4 not vork fine
| |
| 11:22 | Megant_: ...
| |
| 11:22 | <ogra> and it should free that tty if its done
| |
| 11:22 | the only thing i could imagine here would be a race condition
| |
| 11:23 | but thats a bug that would need fixage
| |
| 11:23 | <vagrantc> the ltsp-client scripts blindly start something on the tty...
| |
| 11:23 | <ogra> which is the correct way
| |
| 11:23 | its just important to make sure that tty is available
| |
| 11:23 | which is upstarts job
| |
| 11:24 | <vagrantc> so, upstart starts running something on tty1, and ltsp-client tries to run something on tty1... what happens?
| |
| 11:24 | in that order.
| |
| 11:24 | <ogra> it breaks
| |
| 11:24 | <vagrantc> RIGHT. :P
| |
| 11:25 | <ogra> our job is to make sure to start ltsp-client late enough that the tty is available
| |
| 11:25 | <Megant_> So we have a problem. Our thin clients' X.orgs restart sometimes, but all together.
| |
| 11:25 | <vagrantc> ogra: i would think our job is to make sure that the tty is available before spawning something on it ... ?
| |
| 11:25 | <ogra> as it was our job to make sure sysvinit is done with everything before using something
| |
| 11:25 | right
| |
| 11:25 | meduxa has joined #ltsp | |
| 11:26 | <ogra> you said the same thing in other words
| |
| 11:27 | <vagrantc> ogra: no, you said it should run late enough, i said it should be available.
| |
| 11:27 | there's a difference.
| |
| 11:27 | <Megant_> Could someone help?
| |
| 11:27 | <str4nd> ogra: help?
| |
| 11:27 | <ogra> if its late enough to be available there isnt anymore :)
| |
| 11:27 | <vagrantc> ogra: there's no way to configure upstart to run a process on a tty such that it stays running until you shut down the machine?
| |
| 11:27 | <ogra> str4nd, i'm working on ltsp5, there is not much i know about 4.x
| |
| 11:27 | vagrantc, no idea ... upstart is still nonexistent
| |
| 11:28 | there is a binary that replaces init
| |
| 11:28 | <vagrantc> ogra: huh?
| |
| 11:28 | <ogra> but the whole concept isnt implemented yet
| |
| 11:28 | <vagrantc> ok.
| |
| 11:28 | <ogra> ubuntu edgy still uses sysvinit scripts but the upstart binary
| |
| 11:28 | <vagrantc> well, at any rate, i don't care what the underlying init or init-like thing is, we should have some mechanism to check if a tty is available before using it.
| |
| 11:28 | <ogra> for the tty's the event code is there
| |
| 11:29 | no
| |
| 11:29 | upstart will care
| |
| 11:29 | we'll need to change the initscript to be event driven
| |
| 11:29 | bjohnson has quit IRC | |
| 11:29 | <vagrantc> yeah, that won't be any sort of fork :P
| |
| 11:29 | <ogra> upstart has commands like: on tty
| |
| 11:29 | Avatara has quit IRC | |
| 11:29 | <ogra> or "on defaultroute"
| |
| 11:30 | i.e.: on tty startx
| |
| 11:30 | <vagrantc> sounds cool.
| |
| 11:30 | <ogra> would be a valid upstart initscript
| |
| 11:30 | <str4nd> ogra: (k)ubuntu edgy + ltsp5 works?
| |
| 11:30 | * vagrantc suspects it will be difficult to maintain upstart/sysv compatibility | |
| 11:30 | <ogra> and i think it would be worth it to make ltsp depend on upstart in debian ;)
| |
| 11:31 | str4nd, well, given that i type from a thin client thats an upgraded edgy ... i'd assume so :)
| |
| 11:31 | <vagrantc> ogra: yes, as debian is just an ubuntu sub-project, and totally malleable to ubuntu's will. :P
| |
| 11:31 | <ogra> :PPP
| |
| 11:31 | i suspect debian will switch very soon to upstart
| |
| 11:32 | <vagrantc> meanwhile, i want good backports to etch :P
| |
| 11:32 | Gadi has joined #ltsp | |
| 11:32 | Skarmeth has joined #ltsp | |
| 11:33 | <ogra> pfft backports ...
| |
| 11:34 | ... which reminds me ... i have to do two as well
| |
| 11:34 | <vagrantc> heh.
| |
| 11:34 | <ogra> well, its rather bugfixes ...
| |
| 11:35 | but i have to backport them ...
| |
| 11:38 | egypcio has joined #ltsp | |
| 11:39 | Avatara has joined #ltsp | |
| 11:55 | vykarian has joined #ltsp | |
| 11:55 | <vykarian> hy all
| |
| 11:56 | does someone has a 2.4 kernel series .config file? the kernel packet only has 2.6 confs
| |
| 11:56 | jammcq u have?
| |
| 12:03 | <jammcq> LTSP no longer suports the 2.4 kernels
| |
| 12:04 | 2.4 was missing too many features, like initramfs and udev
| |
| 12:07 | <vykarian> right, ty
| |
| 12:07 | wuuf has joined #ltsp | |
| 12:10 | <wuuf> hi all - anyone alive?
| |
| 12:11 | * vagrantc is a zombie | |
| 12:11 | <wuuf> :-)
| |
| 12:12 | <cliebow_> no 8~)
| |
| 12:12 | <wuuf> I figured I check here before posting to mailing list - As I've seen a few people with the same problem, but they don't really get answered
| |
| 12:13 | * wuuf is in the middle of an epic battle with an on-board ATI card | |
| 12:13 | <cliebow_> wuuf: give it a shot...
| |
| 12:13 | <wuuf> well, for some reason, if I tell LTSP to use the radeon or ATI drivers, I get a black screen
| |
| 12:13 | but I can interact with it
| |
| 12:14 | <cliebow_> can you snag a pci id so we know what you're up against?
| |
| 12:14 | <wuuf> Xorg.0.0.log basically pukes about refresh rates being out of range
| |
| 12:14 | however, I've never had that spill over into a console
| |
| 12:15 | http://68.147.244.84/~woof/LTSP/LTSP.lspci
| |
| 12:15 | board is an Intel D101GGC
| |
| 12:16 | it's an ATI X300, basically
| |
| 12:17 | however, even knoppix doesn't see it quite right
| |
| 12:17 | http://68.147.244.84/~woof/LTSP/knoppix.lspci
| |
| 12:17 | <cliebow_> what are you using for a driver/
| |
| 12:18 | <wuuf> right now, it works with vesa
| |
| 12:18 | at first I tried radeon
| |
| 12:18 | and then ati
| |
| 12:18 | I figured the radeon should work - though I've seen tonnes of problems posted about it
| |
| 12:19 | even tried weird things like, using ATI but with no DRI
| |
| 12:19 | (which, apparently is supposed to fix one thing or the other)
| |
| 12:20 | thing is, I'm not even running a full X session on these guys - I'm just running rdesktop
| |
| 12:20 | <cliebow_> will it work from alive cd..where you could grab the xorg.conf
| |
| 12:20 | <wuuf> yeah
| |
| 12:20 | http://68.147.244.84/~woof/LTSP/knoppix.xorg.conf
| |
| 12:21 | <cliebow_> i have to switch away from this pos..brb
| |
| 12:21 | <wuuf> I've pasted all that stuff into a custom X for LTSP, still no luck
| |
| 12:21 | sep has quit IRC | |
| 12:21 | <wuuf> ok
| |
| 12:21 | :-)
| |
| 12:22 | thankfully, yesterday I found a esd->alsa patch which fixed up the sound
| |
| 12:22 | sep has joined #ltsp | |
| 12:22 | night|away has quit IRC | |
| 12:24 | <Envite> Out for a coffee
| |
| 12:24 | <wuuf> also, I've noted that there are plenty of posts on getting the nvidia binaries to work with LTSP, but no fglrx
| |
| 12:24 | cliebow__ has joined #ltsp | |
| 12:24 | <wuuf> ATI cards seem to be a bit of a bitch to work with, overall :-\
| |
| 12:24 | <cliebow__> post that one more time?
| |
| 12:24 | <wuuf> http://68.147.244.84/~woof/LTSP/knoppix.xorg.conf
| |
| 12:25 | > I've pasted all that stuff into a custom X for LTSP, still no luck
| |
| 12:26 | <cliebow__> wuuf: did xvidtune tell you anything about a proper config?
| |
| 12:26 | <wuuf> well, I don't have xvitune log, but I do have these :
| |
| 12:27 | http://68.147.244.84/~woof/LTSP/knoppix.xrandr
| |
| 12:27 | http://68.147.244.84/~woof/LTSP/knoppix.xdpyinfo
| |
| 12:27 | has anyone configured a card like this?
| |
| 12:28 | I've seen no real answer regarding fglrx configuration on LTSP
| |
| 12:28 | <cliebow__> nothing jumps right out..perhaps the big guns will lok at it
| |
| 12:29 | <wuuf> though there was some mention of r200 drivers permitting that card to use radeon drivers without complaining
| |
| 12:29 | cliebow__ : you ever seen anything on configuring fglrx on LTSP?
| |
| 12:30 | <cliebow_> no never run across it
| |
| 12:38 | <wuuf> mm
| |
| 12:43 | <shogunx> hey sbalneav... how are you?
| |
| 12:44 | <ogra> vagrantc, hmm, seems my answer to your mail doesnt get through ... i'm not subscribed to debian-edu (yet)
| |
| 12:44 | and pkg-ltsp-devel was only cc'ed
| |
| 12:45 | <shogunx> does edgy's ltsp implementation still use the lbus daemons? I noticed ltspmounter getting passed via a socket, but that does not seem to do when not using ldm, so i patched together the /lib/udev scripts to support lbus also, if you want it, as I could not get localdev working otherwise.
| |
| 12:45 | <ogra> but i didnt even get a reject message ...
| |
| 12:45 | shogunx, no it uses the ssh tunnel
| |
| 12:46 | non ldm setups are not supported yet
| |
| 12:46 | <shogunx> ogra... right, but no ssh tunnel with gdm, correct?
| |
| 12:46 | <ogra> right
| |
| 12:46 | no sound either
| |
| 12:46 | and many other no's
| |
| 12:47 | looks like we'll get an unencrypted ssh mode for ltsp in feisty ... so we probably will abandon xdmcp completely ... lets see
| |
| 12:47 | <shogunx> right, hence my kludgearound that would add support for other *dms. sound worked fine via esd for me, but for things unlike xmms... the apps with no esound support, esddsp looks like the only way.
| |
| 12:48 | <ogra> no, ldm sets some env variables ....
| |
| 12:48 | thats why it works out of the box with gnome for example ... since we4 force gstreamer to use the esd output sink
| |
| 12:48 | <shogunx> all i have left is to extract the ${SIZE} data accurately for the call to lbus.fifo
| |
| 12:49 | <ogra> but that will change a lot in feisty anyway
| |
| 12:49 | esd is history
| |
| 12:49 | for lbus jammcq would be your bettr target btw
| |
| 12:49 | <shogunx> eh? new method to handle sound?
| |
| 12:50 | <wuuf> well, there's an alsa->esd package
| |
| 12:50 | <shogunx> ogra... i got it working for my purposes... usbsticks. just wondering if you guys wanted the patched files... it mostly trivial, but does the job.
| |
| 12:50 | <wuuf> you use ESD as the sound Daemon but it allows you to use alsa
| |
| 12:54 | <shogunx> i think the only issue I have had with esd was trying to support a SIP softphone, but we hashed that out here before... looks running as a local app is the only reasonable way to do that.
| |
| 12:56 | ogra... so lbus is essentially deprecated then?
| |
| 12:56 | brich has joined #ltsp | |
| 12:56 | <ogra> yep
| |
| 12:57 | <shogunx> ack. and i can implement ldm via lts.conf to go that route?
| |
| 12:57 | <ogra> shogunx, http://developer.novell.com/wiki/index.php/Edgy/HOWTO:_PulseAudio
| |
| 12:57 | thats for sip phones and skype
| |
| 12:58 | our implementation will be a lot more trivial, but along these lines .... and give you full control over the volume and MIC input
| |
| 12:59 | <shogunx> ogra... i'll test that out. nice.
| |
| 12:59 | <ogra> shogunx, afaik sbalneav wanted to use xatoms instead of lbus, but thats something he can better talk about
| |
| 13:00 | <shogunx> yeah, you mentioned that. interesting, integrating that deep into X11
| |
| 13:00 | <ogra> anyway, gtg ...
| |
| 13:00 | <shogunx> thanx.
| |
| 13:00 | <ogra> ciao ...
| |
| 13:01 | <shogunx> nice talking again.
| |
| 13:03 | <Envite> Here again
| |
| 13:04 | * jammcq thinks dbus is the future, instead of lbus or xatoms | |
| 13:06 | <shogunx> not sure i understand deprecating lbus. it worked pretty well.
| |
| 13:07 | <jammcq> well, I like lbus, afterall, I wrote it :) but dbus is being used now by Gnome and KDE and it's really the standard for message passing
| |
| 13:07 | <Gadi> unencrypted ssh mode?
| |
| 13:07 | it must have been a while since I picked my head up
| |
| 13:08 | <vagrantc> Gadi: yes, essentially using ssh, but not doing any encryption.
| |
| 13:08 | <jammcq> no, I think a vpn is something that we'll eventually need to move towards
| |
| 13:08 | <shogunx> jammcq like ipsec style?
| |
| 13:08 | <vagrantc> Gadi: so all the X forwarding and such works, but the overhead is trivial
| |
| 13:08 | <Gadi> then, we just have to see if the poor video is due to the encryption or the proxy ;)
| |
| 13:08 | <jammcq> yeah, or openvpn
| |
| 13:09 | <vagrantc> setting up VPNs for local device access?
| |
| 13:09 | <Gadi> ldm has nothing over xdmcp when it comes to video performance
| |
| 13:09 | <jammcq> vpns for ALL traffic betwen thin client and server
| |
| 13:09 | <vagrantc> Gadi: have you tried it without encryption?
| |
| 13:09 | * Gadi just heard of ssh without encryption | |
| 13:09 | <Gadi> ;)
| |
| 13:10 | <jammcq> Gadi: it's not compiled in by default
| |
| 13:10 | <ogra> Gadi, i'm not quite gone yet, dont blame ldm for ssh's failures
| |
| 13:10 | :P
| |
| 13:10 | <Gadi> ogra: did not mean to imply that
| |
| 13:10 | <vagrantc> why not blame LDM for a design decision to use ssh?
| |
| 13:10 | <ogra> i know :)
| |
| 13:10 | <Gadi> just dont want xdmcp dumped until the video is comparable
| |
| 13:10 | <ogra> vagrantc, pfft ...
| |
| 13:11 | <vagrantc> ogra: it's fair.
| |
| 13:11 | * ogra is *really* gone now ... feel free to blame ldm for everything ;) | |
| 13:11 | <dberkholz> have y'all tried changing ssh's encryption algorithm?
| |
| 13:11 | might be some that are less intensive
| |
| 13:11 | <Gadi> my only 2 beefs with EdgyLTSP so far are the long boot times (in my testing) and poor video performance (by comparison)
| |
| 13:11 | <jammcq> dberkholz: been there, done that
| |
| 13:11 | <ogra> dberkholz, we already use blowfish
| |
| 13:11 | <Gadi> everything else is magically delicious!
| |
| 13:12 | <ogra> there seems to be no fastrer one
| |
| 13:12 | <vagrantc> i know why ssh-encrypted X sessions are good- i've been doing ldm-like things for years. but it has drawbacks, and people should own up to the drawbacks of using it.
| |
| 13:12 | <dberkholz> don't some chips have hardware encryption modules?
| |
| 13:12 | <jammcq> yeah, drawbacks like supporting 30 clients instead of 120
| |
| 13:12 | <ogra> 120 clients ? on what hardware with what desktop ?
| |
| 13:13 | <Gadi> didn't he say he was gone?
| |
| 13:13 | ;)
| |
| 13:13 | <jammcq> i've got 140 clients on a single server running Icewm
| |
| 13:13 | using OOo, firefox and a bunch of other things
| |
| 13:13 | <ogra> yeah, but if you guys bring up such controversial stuff, how can i ?
| |
| 13:13 | <shogunx> and xdmcp is moving by the wayside for security issues?
| |
| 13:13 | <ogra> jammcq, and you dont achieve that with ssh ?
| |
| 13:14 | i cant imagine that wouldnt work as well with ssh
| |
| 13:14 | <Gadi> does anyone else actually have >1min boot times still?
| |
| 13:14 | <jammcq> I haven't tried, but people who have, claim that about 30 is the upper limit
| |
| 13:14 | <ogra> as long as you dont use a memory hogging desktop
| |
| 13:14 | <vagrantc> Gadi: anyone, meaning... ?
| |
| 13:14 | <Gadi> or am I just doing something ridiculously wrong?
| |
| 13:14 | I dunno - I feel like such a noob
| |
| 13:14 | <shogunx> Gadi... about 20 second on both servr and client here.
| |
| 13:14 | <ogra> jammcq, they should file bugs so i can investigate :)
| |
| 13:14 | <Gadi> I dont know what to expect from LTSP5
| |
| 13:14 | lol
| |
| 13:15 | <ogra> anyway, final call ... i'm gone now
| |
| 13:15 | <Gadi> shogunx: running ldm?
| |
| 13:15 | <jammcq> yeah, right :)
| |
| 13:15 | <vagrantc> i just don't understand why boot time is so important. i understand people complain about it.... but really?
| |
| 13:15 | <Envite> i hope ltsp5 allows us to specify icons for both gnome and kde when using local devices
| |
| 13:15 | <jammcq> ogra: is still lurking, I can just feel it
| |
| 13:15 | <shogunx> Gadi running gdm.
| |
| 13:15 | <ogra> jammcq, :P
| |
| 13:15 | <jammcq> see !!!!!
| |
| 13:15 | <Gadi> vagrantc: because when I log out of ldm, it takes close to 30 seconds for the login screen to come back!
| |
| 13:15 | <vagrantc> Gadi: ah, yeah. LDM is slow as molasses.
| |
| 13:16 | <Gadi> thats important
| |
| 13:16 | <vagrantc> that part is annoying.
| |
| 13:16 | <Gadi> thats a deal breaker
| |
| 13:16 | <vagrantc> Gadi: have you tried sdm? :)
| |
| 13:16 | <Gadi> :)
| |
| 13:16 | <Envite> I always use GDM
| |
| 13:16 | <Gadi> have you made it prettier?
| |
| 13:16 | <jammcq> the boot time sets the users expectations. a slow boot == slow running computer
| |
| 13:16 | <dberkholz> i'm a fan of entrance
| |
| 13:16 | <jammcq> a fast boot == fast computer
| |
| 13:16 | <Gadi> shogunx: local gdm or XDMCP?
| |
| 13:16 | <jammcq> and LTSP-4.2 boots lickity split fast
| |
| 13:16 | <vagrantc> Gadi: no, but people who want it to look pretty can do so :P
| |
| 13:17 | <dberkholz> jammcq: so what you should do is save a suspended image of each box on the server, and never boot from scratch
| |
| 13:17 | <vagrantc> Gadi: it's got hooks for that...
| |
| 13:17 | <shogunx> i like lbus too, jammcq, well written. looks like dbus would be capable of more tight integration into the desktop, but is it capable of the networking aspects? all the articles i see talk solely about IPC.
| |
| 13:17 | Gadi, xdmcp... server has 1Gb/s link and terminlas have 100Mb/s
| |
| 13:17 | <Gadi> shogunx: thats cheating ;)
| |
| 13:18 | <jammcq> shogunx: yeah, its' got a network transport that needs to be enabled at compile time, and from the people i've talked to, it hasn't gotten alot of shake out yet. But, I think it's something we should at least be considering
| |
| 13:18 | <vagrantc> jammcq: yeah, i guess users don't actually notice the speed of a computer while actually *using* it.
| |
| 13:18 | <jammcq> vagrantc: no joke, perception is 99% of the battle. if it boots fast, it IS fast
| |
| 13:18 | <Gadi> XDMCP lacks localdev support still, right?
| |
| 13:18 | <shogunx> Gadi why is that cheating?
| |
| 13:18 | <Envite> Gadi: ?
| |
| 13:18 | <vagrantc> jammcq: yeah, i realize this.
| |
| 13:18 | <dberkholz> vagrantc: haven't you heard how os x initially shipped with no hardware-accelerated video, but because of lack of artifacts people thought it was fast?
| |
| 13:19 | <Envite> I use gdm with xdmcp and i have lda
| |
| 13:19 | <jammcq> heh
| |
| 13:19 | <vagrantc> dberkholz: no...
| |
| 13:19 | <Gadi> shogunx: because ldm is half the slowness ;)
| |
| 13:19 | <jammcq> Envite: that's on ltsp-4.2 prolly, right?
| |
| 13:19 | <Envite> sure
| |
| 13:19 | but it worked on 4.1 too
| |
| 13:19 | <dberkholz> vagrantc: now you have =)
| |
| 13:19 | <jammcq> I think that localdevs still isn't working on ltsp-5 with xdmcp
| |
| 13:19 | <vagrantc> Gadi: now, are you saying LDM is slow because of the pretty GUI login, or because of the encrypted ssh, or both?
| |
| 13:19 | <shogunx> Gadi... i kludged it in via ltsp4.2.2's lbus daemons and additions to /lib/udev scripts that /etc/udev/rules.d/88-ltsp.rules calls
| |
| 13:20 | <jammcq> although, sbalneav has deployed ltsp-5 with xdmcp, so maybe he's got it working, or at last "in the lab"
| |
| 13:20 | <vagrantc> Gadi: cause, the pretty GUI login is totally separate from the ssh connection- it all happens locally.
| |
| 13:20 | <Gadi> vagrantc: its slow on all counts
| |
| 13:20 | slow to load
| |
| 13:20 | and video is slow through ssh+proxy
| |
| 13:20 | noticeably slower than XDMCP
| |
| 13:20 | <Envite> is ltsp moving towards NX?
| |
| 13:21 | <jammcq> no, NX is not free
| |
| 13:21 | <shogunx> jammcq... i essentially forward ported lbus to handle localdev with ltsp5/edgy. it works.
| |
| 13:21 | <Envite> nx IS free
| |
| 13:21 | <Gadi> jammcq: I think he means philosohpically
| |
| 13:21 | or not
| |
| 13:21 | <jammcq> hmm, it's NOT free
| |
| 13:21 | they have a 2-user license that you can get for free
| |
| 13:21 | <Envite> nonononono
| |
| 13:21 | <jammcq> and *some* of their technology is free
| |
| 13:21 | <shogunx> hmm... is that ssh connection getting -C passed to it? that might speed things up.
| |
| 13:21 | <Envite> nx is a protocol
| |
| 13:22 | <vagrantc> shogunx: just slows it down moe.
| |
| 13:22 | more.
| |
| 13:22 | <Gadi> ldm is like NX without the happy packet dance
| |
| 13:22 | <jammcq> ssh is the protocol
| |
| 13:22 | <Envite> and the protocol in free
| |
| 13:22 | <jammcq> ns goes over NX
| |
| 13:22 | err
| |
| 13:22 | NX goes over ssh
| |
| 13:22 | <Envite> no
| |
| 13:22 | nx is a protocol that happens over ssh
| |
| 13:22 | <shogunx> vagrantc, yeah, i guess if bandwidth is not the rate-limiter...
| |
| 13:22 | <jammcq> and yes, the NX specification is free and open, but who's got the time to implement a completely free version of it?
| |
| 13:22 | <Envite> taht is another protocol that happens over tcp
| |
| 13:22 | nx libraries are free too
| |
| 13:22 | freenx is a free nx server
| |
| 13:23 | <jammcq> but there still is NO free complete solution
| |
| 13:23 | <vagrantc> shogunx: CPU is the limitation with the ssh connections
| |
| 13:23 | <Envite> actually, only the client is propietary
| |
| 13:23 | <jammcq> what good is NX without a client?
| |
| 13:23 | <Envite> but knx will soon be a free client
| |
| 13:23 | <jammcq> well, when that happens, we can talk
| |
| 13:23 | <Envite> (it IS, but is not production stable)
| |
| 13:23 | <jammcq> and *mabye* we are interested
| |
| 13:24 | <Envite> BUT we do not need a complete server nor a complete client for ltsp
| |
| 13:24 | <vagrantc> Envite: some people use NX with ltsp. but until a completely free (as in freedom) NX implementation is available, i doubt if it will become the default.
| |
| 13:24 | <jammcq> NX places a HUGE burdon on the server
| |
| 13:24 | it runs an Xserver on the LTSP server for EACH user
| |
| 13:24 | <shogunx> jammcq dbus method looks pretty smooth.
| |
| 13:25 | <jammcq> and it does nothing to help us encrypt the NFS traffic
| |
| 13:25 | <Envite> jammcq: oh, I see
| |
| 13:25 | <jammcq> whereas, a vpn type solution would cover all traffic
| |
| 13:25 | <Envite> i talked thinking only in ssh encryption
| |
| 13:25 | <jammcq> i'm not saying we have to go to a vpn, I just think it's something interesting, that we should investigate
| |
| 13:25 | nx has some good points
| |
| 13:25 | <shogunx> yep... ipsec interfaces comes up right after the regular interface, and does not waste too many cycles.
| |
| 13:25 | <jammcq> and when there's a totally free, well written solution, then i'm interested
| |
| 13:26 | i'm actually a very happy NX user
| |
| 13:27 | <Gadi> jammcq: I thought you just NX'd socially
| |
| 13:27 | <shogunx> hehe
| |
| 13:27 | <jammcq> I usually NX by myself, late at night
| |
| 13:28 | <Gadi> I NXd back in college - but only because I needed the money
| |
| 13:28 | Imnot proud of it
| |
| 13:29 | <shogunx> thats what you get for hanging out near the bathrooms doing remote X sessions from the payphones... these things are bound to happen:)
| |
| 13:30 | freemindX has joined #ltsp | |
| 13:33 | <shogunx> is the current ssh based implementation something along the lines of ssh -X server then X -query localhost?
| |
| 13:34 | <brich> jammcq: richard here. May I ask you about ltsp5 for suse?
| |
| 13:35 | <jammcq> sure, ask away
| |
| 13:36 | <brich> My question is about some additional documentation (besides the wiki).
| |
| 13:36 | <jammcq> yeah, would be nice to have some, wouldn't it :)
| |
| 13:36 | <brich> Yes :)
| |
| 13:36 | Some time ago you wrote something, that there might be more information
| |
| 13:37 | after the november hackfest
| |
| 13:37 | <jammcq> yeah, that was when I thought I'd have to time to write it up. hasn't happened yet
| |
| 13:37 | <brich> So you have some more, or is the information in the wiki alll there is?
| |
| 13:37 | Okay, that's clear than.
| |
| 13:37 | <jammcq> most of the information is collected in the minds of some of the guys who hang out here in the channel
| |
| 13:38 | <brich> Alright, than the wiki is our starting point. If more info is needed I'll
| |
| 13:38 | ask this channel for advice :)
| |
| 13:38 | <jammcq> guys like Vagrantc, ogra, sbalneav they are working daily on this stuff. Others, such as Gadi, dberkholz and a few more also work with this stuff
| |
| 13:39 | to be successful at integrating LTSP-5 into a distro, I think it's a necessity to integrate the people into our dev team
| |
| 13:39 | get them in here, see who the players are, and start sponging off of them
| |
| 13:39 | <dberkholz> i know more theory than this specific implementation, which i'm still learning
| |
| 13:40 | andresmujica has left #ltsp | |
| 13:41 | muh2000 has joined #ltsp | |
| 13:42 | <brich> jammcq: what do you mean with here? Is that the irc channel or an office near to you?
| |
| 13:43 | <jammcq> here == #ltsp on irc.freenode.net
| |
| 13:43 | <brich> jammcq: I go the msg ;)
| |
| 13:54 | <shogunx> jammcq what happens if you do not pass ${SIZE} from lbuscd to lbussd? will the mount fail?
| |
| 13:57 | vykarian has quit IRC | |
| 13:58 | wuuf has quit IRC | |
| 13:59 | zOOd has joined #ltsp | |
| 13:59 | <zOOd> hello
| |
| 14:00 | I get a mounting /proc failed, device or ressource busy on the client.
| |
| 14:00 | where could it come from
| |
| 14:01 | <bricode> Wow, a lot happens when you leave for a couple of hours.
| |
| 14:01 | I didn't get to put in my vote for video and multimedia running client side.
| |
| 14:01 | I got the fat client stuff working last night.
| |
| 14:02 | Pretty sweet video and multimedia.
| |
| 14:02 | Was able to do VoIP calls and play reasonable resolution video.
| |
| 14:03 | Only thing broken is Flash...mostly because of no video acceleration.
| |
| 14:09 | <zOOd> btw, is there a way to log kern messages of the client ?
| |
| 14:12 | <vagrantc> shogunx: ldm essentially does: ssh -X x-window-manager (slightly more complicated, but that's basically it).
| |
| 14:12 | shogunx: it's essentially a gui frontend for "ssh -X"
| |
| 14:13 | i really ought to do some re-writing of sdm to make it easier to make pretty. :)
| |
| 14:13 | Envite has quit IRC | |
| 14:13 | Envit1 has joined #ltsp | |
| 14:23 | Sp0oKeR has joined #ltsp | |
| 14:32 | <vagrantc> zOOd: if you enable remote syslogging, usually syslog will feed kernel messages back to your server.
| |
| 14:33 | http://wiki.ltsp.org/twiki/bin/view/Ltsp/Syslog
| |
| 14:41 | <shogunx> dberkholz entrance like from E17? that has xdmcp support?
| |
| 14:44 | FernandoMM has quit IRC | |
| 14:46 | Envit1 has left #ltsp | |
| 14:46 | andresmujica has joined #ltsp | |
| 14:47 | bobby_C has joined #ltsp | |
| 14:52 | dm_ has joined #ltsp | |
| 14:53 | dm_ has left #ltsp | |
| 14:53 | Gadi has left #ltsp | |
| 14:54 | jammcq has quit IRC | |
| 14:58 | <cliebow_> Hasts
| |
| 14:58 | Hasta
| |
| 14:58 | cliebow_ has quit IRC | |
| 15:00 | cliebow__ has quit IRC | |
| 15:02 | brich has quit IRC | |
| 15:24 | meduxa has quit IRC | |
| 15:33 | arc_ has quit IRC | |
| 15:45 | bricode has quit IRC | |
| 15:50 | freemindX has quit IRC | |
| 16:11 | sbalneav has quit IRC | |
| 16:13 | J45p3r has joined #ltsp | |
| 16:23 | Sp0oKeR has quit IRC | |
| 16:27 | pipedream has quit IRC | |
| 16:29 | wizzy has quit IRC | |
| 16:30 | pipedream has joined #ltsp | |
| 16:34 | wizzy has joined #ltsp | |
| 16:36 | ccjoe has quit IRC | |
| 16:36 | oh207 has quit IRC | |
| 16:36 | gentgeen__ has quit IRC | |
| 16:36 | BadMagic has quit IRC | |
| 16:36 | Arsenick-TC2L has quit IRC | |
| 16:36 | tazle has quit IRC | |
| 16:36 | mdz has quit IRC | |
| 16:37 | BadMagic has joined #ltsp | |
| 16:37 | tazle has joined #ltsp | |
| 16:44 | Egyptian[Home] has quit IRC | |
| 16:45 | Egyptian[Home] has joined #ltsp | |
| 16:48 | ccjoe has joined #ltsp | |
| 16:48 | oh207 has joined #ltsp | |
| 16:48 | gentgeen__ has joined #ltsp | |
| 16:48 | Arsenick-TC2L has joined #ltsp | |
| 16:48 | mdz has joined #ltsp | |
| 16:49 | ccjoe has quit IRC | |
| 16:49 | oh207 has quit IRC | |
| 16:49 | ccjoe has joined #ltsp | |
| 16:49 | oh207 has joined #LTSP | |
| 16:50 | Arsenick-TC2L has quit IRC | |
| 16:50 | gentgeen__ has quit IRC | |
| 16:50 | Arsenick-TC2L has joined #ltsp | |
| 16:50 | mdz has quit IRC | |
| 16:53 | gentgeen__ has joined #ltsp | |
| 16:55 | vykarian has joined #ltsp | |
| 16:58 | * msinhore is away: até amanhã | |
| 17:00 | <vykarian> nite
| |
| 17:05 | mdz has joined #ltsp | |
| 17:07 | Arsenick-TC2L has quit IRC | |
| 17:07 | Arsenick-TC2L has joined #ltsp | |
| 17:11 | bobby_C has quit IRC | |
| 17:13 | andresmujica has left #ltsp | |
| 17:29 | cliebow_ has joined #ltsp | |
| 17:29 | <cliebow_> irssi
| |
| 17:30 | !sound
| |
| 17:30 | <ltspbot> cliebow_: "sound" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/Sound
| |
| 17:30 | <cliebow_> !seen cliebow
| |
| 17:30 | <ltspbot> cliebow_: cliebow was last seen in #ltsp 11 hours, 6 minutes, and 48 seconds ago: <cliebow> yeah..bbiaab
| |
| 17:33 | <cliebow> !seen cliebow_
| |
| 17:33 | <ltspbot> cliebow: cliebow_ was last seen in #ltsp 3 minutes and 0 seconds ago: <cliebow_> !seen cliebow
| |
| 17:41 | <rjune> hey
| |
| 17:41 | <cliebow_> hey
| |
| 17:41 | <cliebow> heh trying irssi
| |
| 17:41 | <cliebow_> me too
| |
| 17:42 | and xchat
| |
| 17:42 | <rjune> LOL
| |
| 18:00 | slipttees has joined #ltsp | |
| 18:00 | <cliebow> rjune: what is my password? must have typed it iin here 50 times
| |
| 18:01 | <rjune> CHOKE
| |
| 18:01 | Like I remember
| |
| 18:01 | brb, I have to clean the pop off my keyboard
| |
| 18:01 | <cliebow> heheh
| |
| 18:19 | slipttees has quit IRC | |
| 18:26 | vaineh has quit IRC | |
| 18:31 | jammcq has joined #ltsp | |
| 18:39 | efra has quit IRC | |
| 18:56 | <cliebow_> like over here
| |
| 19:08 | arc_ has joined #ltsp | |
| 19:16 | JMBarbosa has joined #ltsp | |
| 19:26 | JMBarbosa has quit IRC | |
| 19:28 | FernandoMM has joined #ltsp | |
| 19:32 | vykarian has quit IRC | |
| 19:34 | JMBarbosa has joined #ltsp | |
| 19:46 | jammcq has quit IRC | |
| 19:49 | cliebow_ has quit IRC | |
| 20:01 | patrickb has joined #ltsp | |
| 20:09 | patrickb_ has quit IRC | |
| 20:11 | Stevec has joined #ltsp | |
| 20:34 | Skarmeth has quit IRC | |
| 20:37 | JMBarbosa has quit IRC | |
| 20:38 | Cipher_ has joined #ltsp | |
| 20:38 | Cipher_ has quit IRC | |
| 20:39 | Stevec has left #ltsp | |
| 20:40 | kitchen has joined #ltsp | |
| 20:40 | <kitchen> Hi everyone
| |
| 20:41 | J45p3r has quit IRC | |
| 20:41 | <kitchen> Is there anyone here who would know why my scanner attached to a terminal will only scan in binary, when attached to a standalone linux box works fine in color??
| |
| 20:42 | I've read the wiki and it is recognized in a terminal shell as the proper epson...
| |
| 20:45 | When in xsane I try to change from binary to gray or color I get an error message: "Failed to set value of option mode: Invalid argument"
| |
| 21:07 | Also with XINETD_SERVICES = "saned" added to default, the scanner will scan in binary but XMMS audio fails!
| |
| 21:10 | patrickb_ has joined #ltsp | |
| 21:17 | FernandoMM has quit IRC | |
| 21:26 | patrickb has quit IRC | |
| 22:19 | RiXtEr has joined #ltsp | |
| 22:20 | drma has joined #ltsp | |
| 22:24 | drma has quit IRC | |
| 22:40 | MrMoo has quit IRC | |
| 22:42 | jammcq has joined #ltsp | |
| 22:42 | <jammcq> howdie
| |
| 22:54 | RiXtEr has quit IRC | |
| 23:01 | kitchen has quit IRC | |
| 23:44 | vagrantc has quit IRC | |
| 23:46 | robbie has joined #ltsp | |