00:28 | vagrantc has joined #ltsp | |
00:36 | johnny_ has quit IRC | |
01:40 | Shingoshi has joined #ltsp | |
02:11 | Iv^a_na has joined #ltsp | |
02:20 | Iv^a_na has left #ltsp | |
02:21 | laga_ is now known as laga | |
03:20 | bobby_C has joined #ltsp | |
03:28 | otavio_ has joined #ltsp | |
03:31 | K_O-Gnom has joined #ltsp | |
03:45 | ogra_cmpc has quit IRC | |
04:11 | Q-FUNK has joined #ltsp | |
04:18 | <vagrantc> hrm. autologin and guest login are borked on etch :(
| |
04:18 | ogra_cmpc has joined #ltsp | |
04:26 | bobby_C has quit IRC | |
04:43 | K_O-Gnom is now known as Admiral | |
04:43 | Admiral is now known as Admiral_Gnom | |
04:48 | vagrantc has quit IRC | |
04:50 | Admiral_Gnom has quit IRC | |
04:53 | DonSilver has joined #ltsp | |
05:04 | ogra_cmpc has quit IRC | |
05:08 | ogra_cmpc has joined #ltsp | |
05:20 | alekibango has quit IRC | |
05:31 | ogra_cmpc has quit IRC | |
05:33 | ogra_cmpc has joined #ltsp | |
05:55 | <rjune> dpkg runs a script to configure ldap when slapd gets installed, where can I find that script?
| |
05:56 | DonSilver has quit IRC | |
06:17 | <ogra_cmpc> rjune, thats likely a postinst script
| |
06:17 | locate slapd.postinst :)
| |
06:21 | rjune, but thats nothing you should touch directly, use dbconf preseeding if you want to change values
| |
06:21 | err
| |
06:21 | debconf
| |
06:21 | <rjune> I want to rerun it
| |
06:21 | <ogra_cmpc> sudo dpkg-reconfigure <packagename>
| |
06:22 | that re-runs all the debconf questions
| |
06:22 | <laga> ogra_cmpc: will there be another ltsp upload?
| |
06:22 | <ogra_cmpc> sure
| |
06:22 | several i assume
| |
06:22 | <laga> ogra_cmpc: before hardy, that is. :)
| |
06:22 | <ogra_cmpc> indeed
| |
06:22 | its feb ....
| |
06:22 | we release in april
| |
06:22 | <laga> great. got a very minor fix for my plugin.
| |
06:23 | ogra_cmpc: oh. true. :)
| |
06:23 | <ogra_cmpc> oki
| |
06:24 | * ogra_cmpc curses grub ... why doesnt that darn thing allow me to properly install it into a loop mounted filesystem image without wiping the partition table ... | |
06:28 | * laga curses various java implementations, their respective shortfalls and the confusion resulting from the whole mess | |
06:29 | K_O-Gnom has joined #ltsp | |
06:32 | * laga blinks | |
06:32 | <laga> just thought i'd removed all java VMs and suddenly my java app starts working
| |
06:44 | <rjune> ogra_cmpc: thanks, for some bizzare reason,it seems to have not created the passwd properly
| |
06:45 | <ogra_cmpc> ah
| |
06:45 | * ogra_cmpc cries .... | |
06:45 | ogra has quit IRC | |
06:45 | * ogra_cmpc just issued reboot -fp in the worng terminal .... | |
06:45 | <rjune> doh!
| |
06:45 | <ogra_cmpc> there goes my server
| |
06:45 | K_O-Gnom has quit IRC | |
06:48 | sepski has joined #ltsp | |
06:48 | DonSilver has joined #ltsp | |
06:49 | <rjune> at least it was reboot and not halt
| |
06:50 | ogra_cmpc_ has joined #ltsp | |
06:51 | ogra_cmpc has quit IRC | |
06:56 | K_O-Gnom has joined #ltsp | |
07:05 | Patina has quit IRC | |
07:06 | DonSilver__ has joined #ltsp | |
07:06 | DonSilver has quit IRC | |
07:06 | DonSilver__ is now known as DonSilver | |
07:07 | K_O-Gnom has quit IRC | |
07:07 | Patina has joined #ltsp | |
07:07 | ogra has joined #ltsp | |
07:07 | <rjune> ogra, how's ebox coming?
| |
07:07 | I thought it was gonna be user admin for 7.10
| |
07:19 | ogra_cmpc_ is now known as ogra_cmpc | |
07:19 | sepski has quit IRC | |
07:27 | DonSilver has quit IRC | |
07:29 | sepski has joined #ltsp | |
07:31 | K_O-Gnom has joined #ltsp | |
07:44 | cliebow__ has joined #ltsp | |
08:01 | cliebow_ has quit IRC | |
08:01 | cliebow has joined #ltsp | |
08:03 | cliebow_ has joined #ltsp | |
08:04 | Nuba1 has joined #ltsp | |
08:08 | <Nuba1> cool, alpha5 has my bluethooth recognised automatically... very nice
| |
08:18 | alekibango has joined #ltsp | |
08:18 | cliebow__ has quit IRC | |
08:20 | cliebow has quit IRC | |
08:26 | cliebow__ has joined #ltsp | |
08:34 | spectra has joined #ltsp | |
08:43 | cliebow_ has quit IRC | |
09:00 | <Nuba1> how can I increase sound levels in my thin clients? its very very faint right now
| |
09:12 | steph_ has joined #ltsp | |
09:18 | cliebow_ has joined #ltsp | |
09:34 | cliebow__ has quit IRC | |
09:45 | elisboa has joined #ltsp | |
10:34 | Q-FUNK has quit IRC | |
10:36 | <steph_> Hi everybody. Can someone tell me how the sound on the client side works? How the sound on the client can acces the local sound card? Is it ESOUND or not?
| |
10:37 | <Nuba1> steph pulseaudio I believe
| |
10:37 | <ogra_cmpc> steph_, depends on the version of ltsp you use
| |
10:38 | very old variants used nas ... later version 4.x used esound, early 5.x used esound as well ... the latest releases use alsa emulation
| |
10:39 | Nuba1, your gnome volume control doesnt suffice to adjust the volume ?
| |
10:40 | <Nuba1> ogra, no its almost inaudible
| |
10:41 | I set this in lts.conf but it doesn't seem to do anything:
| |
10:41 | VOLUME = 100
| |
10:41 | FRONT_VOLUME = 100
| |
10:41 | <ogra_cmpc> thats never been used by ltsp5
| |
10:42 | <Nuba1> ok, found it on the web somewhere, thought Id try it :-)
| |
10:42 | is there any way to set the volume to max, or is it already maxed out?
| |
10:43 | <ogra_cmpc> have a look at /etc/init.d/alsa-utils in the chroot
| |
10:43 | adjust master and pcm values and run sudo ltsp-update-image
| |
10:43 | <Nuba1> k... btw, I'm now running alpha5 which is really very stable and a vast improvement on gutsy
| |
10:43 | <ogra_cmpc> heh
| |
10:43 | <Nuba1> Everything ran out of the box, from local cds, to my ati drivers
| |
10:43 | usb sticks
| |
10:43 | <ogra_cmpc> there were not many code changes between gutsy and alpha5
| |
10:44 | <Nuba1> must have been a lot of fixes
| |
10:44 | <ogra_cmpc> you actually get gutsys ltsp with about ten bugfixes
| |
10:44 | <Nuba1> well, then they are 10 very noticable fixes
| |
10:44 | <ogra_cmpc> the packaging changes were massive though, but that didnt chqange any of the ltsp code
| |
10:44 | <Nuba1> thin client manager works too
| |
10:45 | well, maybe this is outside ltsp...
| |
10:45 | <ogra_cmpc> it did in gutsy as well (at least for me)
| |
10:45 | (apart from vnc)
| |
10:45 | <Nuba1> for me, gutsy had broken usb, local devices, thin client manager, ati drivers
| |
10:46 | <ogra_cmpc> the usb fix is well documented ... its a one line change in /usr/sbin/ltspfsmounter
| |
10:46 | <Nuba1> I know, I'm just mentioning that it all seems to work out of the box now
| |
10:47 | <ogra_cmpc> ati drivers are neither tested nor does anything nonfree get installed by defaul
| |
10:47 | <Nuba1> I know I'm not the only one that has noticed.... the mailing list was up in arms about gutsy
| |
10:48 | I'm just saying, at least from what I've seen, hardy should calm a lot of voices
| |
10:49 | might also be because I'm running x64...
| |
10:51 | also, bluetooth and wireless suddenly worked on my laptops in hardy
| |
10:52 | alsa-utils is set to 70% and 80%, I've changed it to 100%
| |
10:52 | maybe that will help
| |
10:59 | <ogra_cmpc> whats your reason to run the 64bit OS ?
| |
11:00 | do you use a lot multithreaded software that could make any use of it (apache or lots of gimp image processing)
| |
11:00 | (beyond that using 64bit is really not worth the hassle)
| |
11:09 | Q-FUNK has joined #ltsp | |
11:09 | <Nuba1> yes and yes
| |
11:10 | I've run both 64 bit and not 64 bit and noticed serious differences in cpu time
| |
11:10 | <ogra_cmpc> i still doubt its worth the effort
| |
11:10 | really ?
| |
11:10 | <Nuba1> well, its the difference between me doing pkill -u 'user' and not
| |
11:10 | yes really
| |
11:10 | <ogra_cmpc> (me knows that the ubuntu kernel team regulary checks teh difference )
| |
11:11 | pkill -u usre ???
| |
11:11 | <Nuba1> I'd have to do pkill -u user after every class if it wasnt for 64 bit
| |
11:11 | yeah make sure all processes are not running at alarmingly high rates
| |
11:11 | or a script to kill all user processes
| |
11:11 | <ogra_cmpc> thats pretty weird, i cant really imagine that 32 vs 64bit can have anything to do with that
| |
11:12 | <Nuba1> then there's the fact I run 8 gigs of ram
| |
11:12 | <ogra_cmpc> in any case that advantage you will get through using 64bit is really minor
| |
11:12 | <Nuba1> I couldnt get 32 bit to recognise the extra ram
| |
11:12 | <ogra_cmpc> i wouldnt suggest 64bit OS to anyone using a normal ltsp server
| |
11:12 | <Nuba1> no matter what howtos people told me to look at on the net
| |
11:12 | <ogra_cmpc> which kernel package did you use _
| |
11:12 | ?
| |
11:13 | <Nuba1> I tried several
| |
11:13 | <ogra_cmpc> -generic supports up to 4gig afaik .... -server up to 64gig
| |
11:13 | <Nuba1> couldn't get generic to work
| |
11:13 | <ogra_cmpc> huh ?
| |
11:13 | thats the default kernel
| |
11:13 | <Nuba1> I mean, couldn't get it to recognise the extra ram
| |
11:13 | <ogra_cmpc> (its ifdentical on 32 and 64 bit)
| |
11:14 | <Nuba1> yes, and 64 bit it recognises the extra ram
| |
11:14 | <ogra_cmpc> right, its limited ot 4gig
| |
11:14 | so you should have used the -server kernel
| |
11:14 | <Nuba1> I tried
| |
11:14 | <ogra_cmpc> well, that runs up to 16 CPUs and up to 64GIG ....
| |
11:14 | <Nuba1> is there a problem with running 64 bit?
| |
11:15 | <ogra_cmpc> its a waste :)
| |
11:15 | there is not much software using it at all
| |
11:15 | <Nuba1> well, I run moodle extensively, which goes through apache
| |
11:15 | <ogra_cmpc> and the hassle you get into making the desktop work is really not worth the time you have to put into it
| |
11:15 | <Nuba1> so it might make a difference :-)
| |
11:16 | well, it was pretty much automatic in heron
| |
11:16 | no hassle at all
| |
11:16 | <ogra_cmpc> yeah, apache is one of the two apps i know making use of the 64bit registers
| |
11:16 | <Nuba1> i had to to change nothing on the desktop to make it run
| |
11:16 | <ogra_cmpc> so you have flash sound and java applets ?
| |
11:16 | <Nuba1> the only hassle was adding the -arch i386
| |
11:16 | <ogra_cmpc> indeed
| |
11:17 | <Nuba1> flash sound, sure
| |
11:17 | java, not tested
| |
11:17 | <ogra_cmpc> sorry for that i'll find a solution for that in hard+1
| |
11:17 | <Nuba1> I dont use it
| |
11:17 | <ogra_cmpc> wow, theer are some 100s of people that surely would loce to know how you made flash sound work
| |
11:17 | (its not even expected to work on ltsp atm)
| |
11:17 | <Nuba1> doing -arch i386 is minor hassle :-) making a fat environment with a switch would be nice for hardy+1 though ;-)
| |
11:18 | I'm not the only one getting flash+sound to work
| |
11:18 | check the mailing list
| |
11:18 | its a simple apt-get install gnash
| |
11:18 | <ogra_cmpc> i didnt read mails since friday
| |
11:18 | and i wont today :P
| |
11:18 | ah
| |
11:18 | gnash
| |
11:18 | ok
| |
11:18 | <Nuba1> is that cheating ;-) ?
| |
11:18 | <ogra_cmpc> you said flash :P
| |
11:19 | <Nuba1> to me its the same
| |
11:19 | <ogra_cmpc> one starts with f the other one with g :)
| |
11:19 | <Nuba1> and, from what I remember libflash is now fixed in hardy isn't it?
| |
11:19 | <ogra_cmpc> the one with f wont work
| |
11:19 | <Nuba1> for 64 bti I mean
| |
11:19 | <ogra_cmpc> nope
| |
11:19 | there is no lib32 binary
| |
11:20 | (simple packaging issue, but nobody put time into it yt)
| |
11:20 | <Nuba1> yeah, I remember you trying to help me to compile it for 64 bit
| |
11:20 | <ogra_cmpc> no
| |
11:20 | <Nuba1> wouldnt work
| |
11:20 | <ogra_cmpc> thats not what i mean
| |
11:20 | its for these peiople running teh 32bit flash via nspluginwrapper onj 64bit
| |
11:21 | you need the 32bit libflashsupport in the lib32 dir ...
| |
11:21 | but by default we only have the plain 64bit version
| |
11:21 | which doesnt workj with the 32bit plugin
| |
11:22 | <Nuba1> there is no 64 bit flash?
| |
11:22 | btw, this now works (with terrible hacks): https://help.ubuntu.com/community/UbuntuLTSP/LTSPFatClients
| |
11:22 | <ogra_cmpc> never was
| |
11:23 | <Nuba1> yeah another company that ignores 64 bit computing
| |
11:23 | then again, vista is no better off
| |
11:23 | <ogra_cmpc> well, MS has the same prob
| |
11:24 | there is not much SW that actually uses 64bit
| |
11:24 | just compiling the packages on a 64bit arch doesnt gain you anything
| |
11:24 | <Nuba1> lazy programmers :-p
| |
11:24 | <ogra_cmpc> the code must make use of the extra functions/registers
| |
11:25 | else its totally pointless to use the 64bit variant ... as long as you have any desktop users
| |
11:25 | <Nuba1> apart from apache what else uses 64bit functions?
| |
11:25 | <ogra_cmpc> gimp
| |
11:25 | gcc
| |
11:26 | surely some of the grid computing things like boinc
| |
11:26 | but you can count it and it wont extend the amount of fingers you own i think
| |
11:26 | <Nuba1> well, I teach gimp and inkscape extensively
| |
11:26 | still, thats why I created the fat clients
| |
11:26 | bandwidth kills the usabillity
| |
11:28 | what about gcompris, childsplay, tuxpaint, tuxmath
| |
11:28 | ?
| |
11:28 | <ogra_cmpc> plain32bit
| |
11:29 | <Nuba1> ah well, then I noticed the difference because I'm using gimp and apache a lot
| |
11:29 | <ogra_cmpc> rpobably
| |
11:30 | <Nuba1> what I do notice, relatively often, is that nautilus goes to 100%
| |
11:30 | for user sessions...
| |
11:30 | burnerx has quit IRC | |
11:38 | <Nuba1> mysql also takes advantage of 64 bit
| |
11:41 | <steph_> ogra_cmpc: I asked you a question about the sound on the client side. I'm using LTSP 5. So I assume that it use pulse, right?
| |
11:46 | indradg has joined #ltsp | |
12:04 | <steph_> ogra_cmpc: You gave me an answer (nas, esound, alsa emulation).Thanks. But how the sound "travels"? From the client to the server then back to the client? Or the client connects directly to the local soundcard? Or something else?
| |
12:06 | <ogra_cmpc> steph_, in the session there is an automatic setup of the virtual alsa device which connects in the backend to a pulse stream ... on the client that stream connects to the alsa layer again
| |
12:21 | <steph_> ogra_cmpc: So everything is on the client side, right?
| |
12:21 | <ogra_cmpc> no
| |
12:21 | its on both sides
| |
12:22 | one doesnt work without the other
| |
12:23 | <steph_> I think I can figure how it works. That's why the package ltsp-server is installed, right?
| |
12:24 | <ogra_cmpc> ltsp-server only brings the one script that sets up the alsa device during login
| |
12:24 | ltsp-client inside the client environment sets up the HW side of things where the session script attaches to
| |
12:25 | ldm has the scripts that actually trigger the setup
| |
12:25 | <steph_> HW ???
| |
12:25 | <ogra_cmpc> hardware
| |
12:25 | <steph_> sorry, that was a stu*** question :|
| |
12:25 | <ogra_cmpc> so that the session has access to the client soundcard over the net
| |
12:27 | <steph_> We are getting there. That's what I want to understand. The session gets access the client soundcard over the network. right?
| |
12:27 | <ogra_cmpc> right
| |
12:27 | ltsp-client sets up everzthingt to make that possible during boot
| |
12:28 | ldm than uses that setup to tell the session where to attach to
| |
12:28 | during login
| |
12:29 | on ubuntu the autodetector of gstreamer will then automatically attach its sound output anhd volume control to that
| |
12:29 | <steph_> So I just need, on the session, to use ALSA (who's a kind of emulation)
| |
12:29 | <ogra_cmpc> what exactly are you trying to do
| |
12:29 | (and what distro/ltsp release are you using there)
| |
12:30 | ltsp in ubuntu since feisty (7.04) sets up everyting out of the box
| |
12:31 | debian should have that default since lenny and with lots of manual fiddling with the etch backports packages
| |
12:31 | <steph_> First, I would like to understand (thanks, you help me a lot). Second, I would like to use "light" audio appz in a classroom. I tried with ubuntu studio, but sound on audio appz came out on the server's speaker.
| |
12:32 | <ogra_cmpc> well, not sure what ubuntu studio doe to the ubuntu defaults if it comes to sound setup
| |
12:32 | i can imagine that tthey change a lot since they aim for music and video editing
| |
12:32 | <laga> JACK maybe
| |
12:33 | <ogra_cmpc> i.e. install jackd which might let you sit there being screwed
| |
12:33 | laga, heh, snap
| |
12:34 | <steph_> From what I read, i made a fresh installation of edubuntu and installed studio packages.
| |
12:34 | <ogra_cmpc> steph_, did you add an lts.conf or did anything else to override the defaults ?
| |
12:34 | <steph_> I tried with jack with the same result.
| |
12:34 | <ogra_cmpc> javck wont work
| |
12:35 | <steph_> I figured there was a module to load or a file to configure... but wich one.
| |
12:36 | <ogra_cmpc> you can surely change things to your needs, but did you test the setup *before* you added any non standard stuff (like teh studio metapackages, or any lts.conf entries)
| |
12:37 | i have only seen it not working for people who broke their setup through lts.conf yet ...
| |
12:37 | <steph_> Well, the sound came out from the client's speaker (like beep, bop, boum :)
| |
12:37 | <laga> ogra_cmpc: BTW, what does LTSP do if it finds an existing /etc/X11/xorg.conf file? ignore it and create its own?
| |
12:37 | <ogra_cmpc> laga, it copies its own over that unless you set CONFIGURE_X=false
| |
12:38 | <steph_> Only after I've install audio meta packages
| |
12:38 | <ogra_cmpc> (see the ltsp-client-core initscript)
| |
12:39 | steph_, which ones ?
| |
12:39 | <steph_> -confirm- this package is not install on the server. How can I verify on the client?
| |
12:39 | <laga> ogra_cmpc: thanks. is there any reason you don't just trust bulletproof X?
| |
12:39 | <steph_> ubuntu-studio metapackes
| |
12:42 | <ogra_cmpc> laga, i'd even prefer to run without xorg.conf completely, but that doesnt work with certain mouse and keyboard setups where you need manual configuration
| |
12:42 | <steph_> ogra_cmpc. It is said to not install ltsp-client-core on a regular machine. It is intended for installation inside a LTSP client. Do I need to install this package with chroot?
| |
12:42 | <ogra_cmpc> laga, we used to use the way the livecd uses (simply dpkg-reconfigure -phigh xserver-xorg durin boot)
| |
12:43 | but thats not bearable speed wise
| |
12:43 | it takes nearly a minute to shovel the complete debconf db over the net ... its way to slow
| |
12:44 | <laga> ogra_cmpc: ouch.
| |
12:44 | <ogra_cmpc> the new way is to let Xorg --configure do the grunt work and sed through the resulting file now
| |
12:44 | that takes less than 10secs during boot
| |
12:44 | <laga> ogra_cmpc: ok, for mythbuntu-diskless, i'll just set CONFIGURE_X=false and hope that bulletproof X works on first boot.
| |
12:44 | <ogra_cmpc> you will be stuck with us keyboard
| |
12:44 | (in case you care)
| |
12:45 | steph_, it will alreqady be there, no need to install it
| |
12:46 | <laga> ogra_cmpc: how is that different from the live disks? :) remember that i've got a persistent copy on write branch where the xorg.conf can be saved
| |
12:46 | <steph_> ogra_cmpc: ok.
| |
12:46 | <ogra_cmpc> its installed in the client environment under /opt/ltsp/i386
| |
12:46 | <steph_> ogra_cmpc: there is a lts.conf in /opt/ltsp/i386/etc. Is there something special I need to look for?
| |
12:46 | vagrantc has joined #ltsp | |
12:46 | <ogra_cmpc> laga, the livecd sets keyboard layout and variant according to the selected locale you picked at the gfxboot screen
| |
12:47 | steph_, read the warning in the lts.conf
| |
12:47 | steph_, and note that you usually womnt need the lts.conf at all
| |
12:47 | <laga> ogra_cmpc: really? i guess i should try the live disks more often.
| |
12:47 | <ogra_cmpc> if it worked before you installed the studio packages its not an ltsp thing you look for
| |
12:48 | <laga> ogra_cmpc: how does ltsp currently figure out the correct keyboard layout?
| |
12:48 | <ogra_cmpc> it gets pulled from /etc/default/console-setup
| |
12:48 | which we copy from the server during ltsp-build-client
| |
12:49 | <vagrantc> RyanRyan52: so, some good news, and bad news ...
| |
12:49 | RyanRyan52: good news is ltsp and ldm build fine on etch ...
| |
12:49 | RyanRyan52: bad news is that the guest login button and autologin don't work at all on etch.
| |
12:49 | <laga> ogra_cmpc: that's smart. hum. now i'm wondering how to implement a similar thing..
| |
12:50 | <vagrantc> RyanRyan52: and we need guest login to switch over at freegeek
| |
12:50 | <ogra_cmpc> laga, well, you have an ubuntu-standard system below your mythtv setup i assume
| |
12:50 | that should automatically configure console-setup
| |
12:50 | so you should be able to just do the same
| |
12:51 | err
| |
12:51 | your plugin sits in the middle of the setup, the locales plugin should already apply to you ...
| |
12:51 | so it should already get set
| |
12:52 | * ogra_cmpc would suspect lirc support being more intresting than keymapping for laga | |
12:53 | <laga> ogra_cmpc: yes. but i need to find a good way to get a nice X config on first boot and let the user create their own xorg.conf if they want/need.
| |
12:54 | ogra_cmpc: lirc support is quite good in gutsy and hardy, fortunately :)
| |
12:54 | <ogra_cmpc> ah :)
| |
12:55 | laga, btw, where in .de do you sit ?
| |
12:55 | <laga> ogra_cmpc: saarland
| |
12:55 | <ogra_cmpc> ah
| |
12:55 | * ogra_cmpc is in kassel | |
12:55 | <laga> where am i sitting? in front of my computer ;)
| |
12:55 | <ogra_cmpc> :p
| |
12:55 | <laga> ah. thought you were in london because you work for canonical
| |
12:55 | <vagrantc> /me blinks
| |
12:55 | <ogra_cmpc> neh
| |
12:56 | <vagrantc> the m68k buildd is building ltsp
| |
12:56 | <ogra_cmpc> i only have to in the office once in six months
| |
12:56 | the majority of time i work at home
| |
12:56 | <laga> ogra_cmpc: nice
| |
12:56 | <ogra_cmpc> vagrantc, ugh
| |
12:56 | vagrantc, how will you ever test that ?
| |
12:57 | <vagrantc> ogra_cmpc: i don't care
| |
12:57 | ogra_cmpc: what it means is i will be able to upload ldm without a sponsor
| |
12:57 | * ogra_cmpc imagines clients taking a week to boot :) | |
12:57 | <vagrantc> ogra_cmpc: m68k was the last architecture that hadn't built a package since the ldm split
| |
12:57 | <ogra_cmpc> yeqah
| |
12:58 | <steph_> ogra_cmpc: Can you give me an insight? In the session, I changed everything in the sound preferences to ALSA. But there is a drop down list box who shows me the soundcard....of the server??
| |
12:58 | <vagrantc> well, i assume i'll be able to upload ldm after m68k has ltsp packages built
| |
12:58 | <ogra_cmpc> steph_, indeed ... dont select it :)
| |
12:58 | Shingoshi has quit IRC | |
12:58 | <vagrantc> i don't know for certain that's why my upload was rejected, but i think it's pretty likely
| |
12:58 | Shingoshi has joined #ltsp | |
12:58 | <ogra_cmpc> steph_, also check the users ~/.asoundrc.asoundconf
| |
12:59 | <vagrantc> i should really just finish out my new maintainer application and become a debian developer
| |
12:59 | <ogra_cmpc> it should only have two active lines
| |
12:59 | ++
| |
13:01 | steph_, aplay /usr/share/sounds/login.wav .... does that give you proper output on the client ?
| |
13:01 | (run that inside a gnome-terminal in a running ltsp session)
| |
13:01 | <steph_> Well, you teached me something. That's why there is a hw:0 (the server soundcard) and a hw:1 (the local soundcard)
| |
13:03 | lns has joined #ltsp | |
13:04 | <steph_> Is there a way (a .conf or something else) to disabled hw:0 on the session: I don't want my students to mess my settings?
| |
13:05 | <ogra_cmpc> just make sure ~/.asoundrc.asoundconf is empty on login
| |
13:05 | ldm/ltsp will care for the rest
| |
13:05 | after login you should see a pcm and ctl device in there
| |
13:06 | <steph_> just to understand, what pcm and ctl mean?
| |
13:07 | <ogra_cmpc> they are the virtual alsa devices pcm is the playback device, ctl is the volume control device
| |
13:08 | <steph_> Now I have a audio appz running well. Thanks. In the preferences of another audio appz, there is something like /dev/dsp. Can I connect to another device?
| |
13:08 | <ogra_cmpc> it needs to point to alsa
| |
13:08 | <steph_> ??
| |
13:08 | <ogra_cmpc> well, if there is a selection option for the device, point it to alsa
| |
13:09 | <steph_> My question was: There is no drop down list box ! lol
| |
13:09 | <lns> ogra, was there ever a package compiled for libflashsupport on AMD64? Just curious as I'm updating my wiki.
| |
13:09 | <ogra_cmpc> lns, yes, but thats plain amd64 ... not usable with the nspluginwrapper 32bit setup
| |
13:10 | we will need a package that installs the 32bit version to lib32 for that to properly work
| |
13:11 | 32bit flash wont work with the 64bit lib
| |
13:11 | <lns> ogra_cmpc, gotcha. thanks =)
| |
13:11 | <ogra_cmpc> steph_, what app is that ?
| |
13:11 | <lns> ogra_cmpc, yeah I remember working with you on compiling the revolutionlinux source for libflashsupport and it seemed to work but still no sound
| |
13:12 | <ogra_cmpc> well, we now have a proper package for libflsahsupport in the archive for hardy ...
| |
13:12 | but that only helps with 32bit
| |
13:12 | <lns> oh207, nice
| |
13:12 | well at least that won't be another manual step
| |
13:13 | always a good thing(tm)
| |
13:13 | <ogra_cmpc> or if you convince adobe to provide a native 64bit build
| |
13:13 | <lns> LAWLZ
| |
13:13 | right
| |
13:13 | they'll do that after they port shockwave director to linux
| |
13:13 | <steph_> ogra_cmpc: You will laugh ! It's a metronom for my students. GTick !! Is there something I can write like /dev/alsa_emulation_or_the_soundcard_hw:1?
| |
13:14 | <ogra_cmpc> steph_, try calling it with padsp in front of the command
| |
13:14 | i.e.: padsp gtick
| |
13:14 | mcfloppy has joined #ltsp | |
13:14 | <ogra_cmpc> and file a whishlist bug at gtick's upstream bugtracker that they add proper alsa support ;)
| |
13:15 | accessing /dev/dsp directly is soooo 90s
| |
13:15 | <steph_> lol
| |
13:16 | You are patient and funny ;)
| |
13:16 | <ogra_cmpc> and lifeless :) (else i wouldnt hang around here on a sunday evening)
| |
13:17 | <lns> ogra_cmpc, does that mean that once you get a life you're going to stop hanging out with us? :( hehe
| |
13:17 | <steph_> hilarious !
| |
13:18 | misui has joined #ltsp | |
13:18 | <ogra_cmpc> vagrantc, btw, the ubuntu jeos guys have apparently solved the grub loopback issue
| |
13:18 | <steph_> maybe you can call 1-800-FREUD
| |
13:18 | <laga> oh, it's sunday?
| |
13:18 | <ogra_cmpc> vagrantc, in a really weird way, by cutting stage1 and 2 into pieces through dd
| |
13:18 | ogra@ceron:~/devel/packages/ubuntu-vm-builder-0.2$ grep dd ubuntu-vm-builder|wc -l
| |
13:18 | 45
| |
13:19 | each of these is only to modify the grub image to ignore the fct that there is no bios device for the loopback
| |
13:19 | <vagrantc> ogra_cmpc: wow. the ubuntu/debian diff on ltspfs is tiny! :)
| |
13:19 | <ogra_cmpc> its the most insane code i've see in a while
| |
13:19 | vagrantc, i synced yours
| |
13:20 | only chqanged the ldm dep since i had to
| |
13:20 | <vagrantc> ogra_cmpc: i think i noticed a while back, but it hadn't yet posted the diff to debian
| |
13:20 | <ogra_cmpc> vagrantc, if you could depend on >= 5.0.40 that would help i guess
| |
13:20 | just dropping the tail
| |
13:21 | <vagrantc> ogra_cmpc: too late to switch back ... once epoch, always epoch.
| |
13:21 | <ogra_cmpc> oh, i forgot
| |
13:21 | right, you carry that ugly thing with you
| |
13:21 | <vagrantc> well, ldm surely isn't 5.x
| |
13:22 | <ogra_cmpc> well, it was before
| |
13:22 | <steph_> ogra_cmpc: so I need pulseaudio-utils for padsp, right?
| |
13:22 | <vagrantc> yes, but i consider that a bug
| |
13:22 | hence the epoch
| |
13:22 | <ogra_cmpc> steph_, might be, i cant remember ... just run it in a terminal, the command-no-found function of the shell should tell you the packagename you need
| |
13:23 | vagrantc, it sad that you cant really benefit from laga's work
| |
13:23 | fat mythtv clients are a cool enhancement
| |
13:23 | iMav has joined #ltsp | |
13:24 | <laga> porting mythbuntu ;) to debian shouldn't be that hard
| |
13:24 | <ogra_cmpc> laga, well, they dont have the seed infrastructure and metapackages ... that would be a ton of manual work
| |
13:24 | <laga> you'll just need aufs and some small packages.. could get away with using marillat's mythtv packages
| |
13:25 | <ogra_cmpc> doing what germinate does by hand is a hard thing i imagine
| |
13:25 | <laga> ogra_cmpc: if you replace mythbuntu-desktop with mythtv-frontend it should be a lot easier. of course, they'll be missing lots of the polish that mythtv was given in ubuntu
| |
13:25 | <ogra_cmpc> but indeed you can have plain mythtv stuff in there
| |
13:26 | right
| |
13:26 | <laga> mythtv is such a neat toy. sometimes you have to recompile your kernel, then you need to deal with lirc, X, mysql.. it's taught me a lot about linux. xyou don't have to jump trhough all these hoops anymore these days
| |
13:26 | <ogra_cmpc> anyway, there is war of the worlds on tv and i should actually spend some hours without lappie on my knees ...
| |
13:27 | * laga is recording it. ;) | |
13:27 | <ogra_cmpc> heh4e
| |
13:27 | i'm so upset they switched off their HD last week
| |
13:27 | <laga> yes, that's annoying.
| |
13:27 | can still go premiere HD ;)
| |
13:28 | <ogra_cmpc> i saw the movie already ... twice ... on 7" displays on planes
| |
13:28 | <lns> Woo, Openoffice 2.4 will have the pixmap memory leak fix =) http://www.openoffice.org/issues/show_bug.cgi?id=85321
| |
13:28 | <ogra_cmpc> and was really looking forward to see it in good quality on a big screen ... but well
| |
13:28 | <lns> I love being a part of fixing bugs, even if just reporting them =p
| |
13:28 | <ogra_cmpc> lns, the current ubuntu package should have it since three weeks
| |
13:29 | Egyptian[Home] has joined #ltsp | |
13:29 | <ogra_cmpc> laga, i dont pay for watching tv ... no premiere for me
| |
13:29 | <lns> ogra_cmpc, OOo 2.4 in gutsy?
| |
13:29 | or just the patch
| |
13:29 | <ogra_cmpc> lns, currently the patch ...
| |
13:29 | <lns> oh cool!
| |
13:29 | wasn't aware of that
| |
13:29 | <laga> ogra_cmpc: schon GEZahlt? :)
| |
13:29 | <ogra_cmpc> 2.4 is due next week or so
| |
13:29 | laga, pfft
| |
13:29 | * lns starts the drumroll | |
13:29 | <lns> nice..love it
| |
13:30 | now we just wait for FF3
| |
13:30 | <ogra_cmpc> thats witing for the deps to be rebuilt for main
| |
13:30 | should be in next week or during the next two
| |
13:30 | <lns> wow
| |
13:31 | <ogra_cmpc> hardy will in anyy case release with ff3 and ooo 2.4
| |
13:31 | <lns> of course =)
| |
13:31 | that will be nice...and i'll have all my labs on a stable supported version again =p
| |
13:31 | <ogra_cmpc> (we need to support that crap for three years )
| |
13:31 | <lns> lol
| |
13:32 | <ogra_cmpc> anyway, off for tv ...
| |
13:32 | <lns> later ogra, off for breakfast
| |
13:32 | <steph_> ogra_cmpc: can I use JACK with LTSP ?
| |
13:32 | oh sorry
| |
13:33 | ciao! and thanks
| |
13:35 | Q-FUNK has quit IRC | |
13:35 | <RyanRyan52> vagrantc: I will mess with it later today...whats happening?
| |
13:36 | <steph_> Is it possible to use JACk and LTSP?
| |
13:36 | <vagrantc> RyanRyan52: when attempting autologin or clicking on the guest login button, X restarts
| |
13:38 | bobby_C has joined #ltsp | |
13:47 | ccherret1 has joined #ltsp | |
13:48 | ccherrett has quit IRC | |
13:54 | elisboa has quit IRC | |
13:59 | <vagrantc> ogra_cmpc: have you got a ltsp-ubuntu-packaging branch you can push somewhere? i'd like to try and merge as much as i can
| |
13:59 | ogra_cmpc: ditto for ldm
| |
14:04 | indradg is now known as idg|Zzzz | |
14:05 | ccherret1 is now known as ccherrett | |
14:09 | cliebow__ has joined #ltsp | |
14:11 | cliebow_ has quit IRC | |
14:49 | K_O-Gnom has quit IRC | |
14:51 | misui has left #ltsp | |
15:06 | bobby_C has quit IRC | |
15:07 | indradg_ has joined #ltsp | |
15:22 | Q-FUNK has joined #ltsp | |
15:24 | idg|Zzzz has quit IRC | |
16:16 | sepski has quit IRC | |
16:23 | <RyanRyan52> vagrantc: should I use the etch ltsp-server or should I use a backport?
| |
16:29 | steph_ has quit IRC | |
16:31 | <vagrantc> RyanRyan52: backport
| |
16:31 | RyanRyan52: i've got the latest ltsp backport ...
| |
16:31 | elisboa has joined #ltsp | |
16:31 | <RyanRyan52> ok
| |
16:32 | <vagrantc> RyanRyan52: deb http://llama.freegeek.org/~vagrant/debian etch-backports/
| |
16:32 | RyanRyan52: it's also got my latest ldm build, which seems to work fine with everything except autologin and guestlogin
| |
16:33 | RyanRyan52: i was thinking of renaming LDM_ALLOW_GUEST to LDM_GUESTLOGIN to be more consistant with LDM_AUTOLOGIN
| |
16:33 | <RyanRyan52> okay
| |
16:33 | <vagrantc> oh, i haven't tested LDM_DIRECTX yet ...
| |
16:38 | Q-FUNK has quit IRC | |
16:38 | Q-FUN1 has joined #ltsp | |
16:38 | Q-FUN1 is now known as Q-FUNK | |
16:41 | achandrashekar has joined #ltsp | |
16:41 | <achandrashekar> Johnny: thanks for the help..I got it all working with edubuntu 7.10 :)
| |
16:41 | <johnny> rox0r..
| |
16:41 | <achandrashekar> johnny: the push was all i needed to get it done
| |
16:41 | <johnny> now gimme some clients :)
| |
16:41 | lol
| |
16:42 | jk
| |
16:42 | <achandrashekar> :)
| |
16:44 | Now for the configurations and lock down of the ltsp clients
| |
16:44 | and we'll see if the ldap integration gets in the way.
| |
16:47 | <cliebow__> Good job getting it going!
| |
16:47 | Rock on!
| |
16:51 | <vagrantc> hrm. seem to be having some troubles with LDM_DIRECTX ... maybe gadi was right
| |
16:53 | <RyanRyan52> vagrantc: about what?
| |
16:54 | <vagrantc> RyanRyan52: about not passing "ssh -X" when using LDM_DIRECTX ...
| |
16:55 | RyanRyan52: might be having xauth related issues somehow
| |
16:55 | <RyanRyan52> oh...what is LDM_DIRECTX?
| |
16:55 | <vagrantc> RyanRyan52: it uses ssh to log in, but doesn't use ssh X11 forwarding for the session
| |
16:55 | <RyanRyan52> oh
| |
16:56 | <vagrantc> so it gives a pretty significant performance boost at some cost in security
| |
16:58 | <RyanRyan52> Is there a backport of ltsp-client somewhere?
| |
16:59 | <vagrantc> RyanRyan52: at the sources.list entry on llama i mentioned earlier
| |
16:59 | RyanRyan52: so use ltsp-build-client --extra-mirror "deb http://llama.freegeek.org/~vagrant/debian etch-backports/
| |
16:59 | "
| |
17:00 | RyanRyan52: although you'll probably need to use the pkg-ltsp.alioth.debian.org backport as well
| |
17:00 | RyanRyan52: oh, and --accept-unsigned-packages
| |
17:00 | <RyanRyan52> I got it working now, thanks.
| |
17:00 | <vagrantc> RyanRyan52: http://wiki.debian.org/LTSP/Howto ... at the bottom it mentions how to install backports
| |
17:01 | <achandrashekar> cliebow__: now to deal with such things as setting up appropriate groups and such. I did notice that when I created a user for me as teacher there is a menu item for lockdown. Hope that menu item does NOT show for students.
| |
17:01 | <johnny> it won't work for them
| |
17:01 | even if it does show
| |
17:01 | afaik
| |
17:02 | i'll talk to you more about the situation a bit later
| |
17:03 | <achandrashekar> johnny: good enough...you'll have to explain how to create the "admin" user, and it appears if i create a local user, not ldap then i can browse their home...and i dont want that..so we can talk a bit later. Really appreciate it. :)
| |
17:04 | <johnny> the admin user ? you're it
| |
17:04 | whoever can sudo can be admin
| |
17:05 | why are you creating local users?
| |
17:05 | <vagrantc> there's an "admin" group
| |
17:05 | <johnny> anybody who you want to be able to sudo would go in that group
| |
17:06 | vagrantc, howdy!
| |
17:06 | * vagrantc waves | |
17:06 | <johnny> vagrantc and ogra are my heroes of the month
| |
17:06 | <vagrantc> heh
| |
17:06 | <johnny> thanks guys..
| |
17:06 | i wish i could help you a bit more
| |
17:07 | still screwing up with sabayon in 2.22 :(
| |
17:08 | silly gvfs ..
| |
17:10 | Q-FUNK has quit IRC | |
17:29 | elisboa has quit IRC | |
17:30 | achandrashekar has quit IRC | |
17:52 | moritz has joined #ltsp | |
17:52 | <moritz> Good evening
| |
17:54 | I am new to LTSP and there are a couple of aspects I don't understand yet - maybe someone can give me a hint into the right direction. My first question is about the /opt/ltsp/ARCH chroot tree. What is it about? Am I right in assuming, that it has nothing to do with the system the user sees after logging in at his terminal client?
| |
17:55 | <vagrantc> moritz: /opt/ltsp/ARCH is the thin-client's root filesystem, but yes, the system the user sees basically has nothing to do with that
| |
17:56 | moritz: once you log in, you're logged into the server itself
| |
17:56 | <moritz> Alright
| |
17:57 | But then: why does the text on http://wiki.ltsp.org/twiki/bin/view/Ltsp/MueKow mentions "installing firefox in the chroot environment"?
| |
17:57 | to quote:
| |
17:57 | "For example, If you want to install firefox into an LTSP tree on a Debian system, you'd run the following commands: chroot /opt/ltsp/i386 apt-get install mozilla-firefox "
| |
17:57 | why would one want to install firefox in the chroot tree?
| |
17:58 | since it is only used to provide the remote-login to the server?
| |
18:01 | <vagrantc> moritz: well, you *could* install and run applications locally, but it is not the typical environment ...
| |
18:02 | moritz: if you had a thin-client with decent processor and memory, you could run all applications locally ...
| |
18:02 | moritz: some folks are doing that ... some call it diskless workstations
| |
18:02 | <moritz> ah. that's about "LocalApps"
| |
18:02 | <vagrantc> and localapps is something different ... that's running some applications locally, and some on the server
| |
18:02 | which is a little trickier
| |
18:03 | <moritz> ok
| |
18:05 | besides trying to understand the whole concept, I wonder why I fail to boot ltsp with qemu
| |
18:05 | :-)
| |
18:05 | <vagrantc> moritz: how are you trying to boot ltsp with qemu?
| |
18:05 | <moritz> as i understand it, there are several ways to do so.
| |
18:06 | 1) with the qemu-ltsp script
| |
18:06 | 2) by setting up network bridges and using qemu with the tap device
| |
18:06 | when I use the qemu-ltsp script, i see plenty of startup messages
| |
18:06 | including a login prompt
| |
18:07 | and then - i guess it's trying to start the x server - i only see a black screen
| |
18:07 | nothing else
| |
18:07 | and when trying 2), it doesn't even boot.
| |
18:08 | my system log (on the terminal server) tells me that no single dhcp request is processed
| |
18:08 | which probably means that the network bridging/tap device stuff doesn't really work
| |
18:08 | but i'm not really familiar with that
| |
18:09 | http://www.ltsp.org/twiki/bin/view/Ltsp/Clients#Qemu_as_a_client <- that's the howto i was following for 2)
| |
18:16 | chup is now known as chupacabra | |
18:17 | <monteslu> jammcq, your pistons are back. Absolutely stomped my suns this afternoon. Fugly game.
| |
18:18 | <vagrantc> moritz: i used to use qemu a lot ... in fact i wrote that qemu-ltsp script ...
| |
18:18 | moritz: but i've mostly switched to virtualbox
| |
18:18 | <moritz> hah, interesting
| |
18:19 | virtualbox.... looking up.... another virtualization software!
| |
18:20 | <vagrantc> moritz: yeah, it was recently open-sourced
| |
18:20 | <moritz> never heard of it
| |
18:21 | how does it compare to qemu and the proprietary solutions?
| |
18:23 | <vagrantc> it's a little more like vmware ... it's GUI-only, pretty much. it requires host-side kernel modules (but doesn't require any in the guest OS)
| |
18:23 | it's way faster than qemu, in my experience. and i'm a long-time qemu fan :)
| |
18:24 | <moritz> not following news in the free software world for a couple of months and such things happen. =)
| |
18:26 | be right back, need to re-login in order to be member of the vboxusers group =)
| |
18:26 | moritz has quit IRC | |
18:27 | moritz has joined #ltsp | |
18:28 | <RyanRyan52> vagrantc: I added some debugging code and it confused me more...it responds "Ya0,hrm0,(hrm1-7),hrm8" then dies without a Ya1. Is there something I'm missing? http://llama.freegeek.org/~ryan52/bzr/ltsp/ldm-debug/
| |
18:29 | <vagrantc> RyanRyan52: that looks like an empty directory ... ?
| |
18:29 | oh, bzr branch
| |
18:30 | <moritz> vagrantc: how do i go about booting ltsp with virtualbox?
| |
18:30 | <RyanRyan52> moritz: you download a debian/ubuntu install cd and create a new virtual machine...it will hel you
| |
18:31 | moritz: Then you follow that distros guide for installing ltsp
| |
18:32 | moritz: then you add another interface to the server in settings thats connected to the internal netword and create a thinclient virtual machine with no hard drive that only has one network card thats connected to the internal network
| |
18:32 | <moritz> that's the interesting point: the network settings. i will give it a try, thanks
| |
18:47 | moritz has quit IRC | |
18:57 | cooteian has joined #LTSP | |
18:58 | <cooteian> I got a thin client booted over the network, but I can't seem to login.
| |
18:58 | <vagrantc> cooteian: linux distro and release?
| |
18:58 | <cooteian> the server is a mythbuntu 7.04 i think
| |
18:59 | <RyanRyan52> vagrantc: So do you have any idea what's going on with ldm?
| |
18:59 | <vagrantc> cooteian: laga's your expert with mythbuntu
| |
18:59 | i think
| |
18:59 | <cooteian> not really. It pops up to the ubuntu login screen, but I don't know if it's gdm or ldm
| |
18:59 | <vagrantc> RyanRyan52: i haven't looked into it much
| |
18:59 | <RyanRyan52> oh
| |
19:00 | cliebow__ has quit IRC | |
19:00 | <cooteian> basically I'm trying to set up a silent client in the tv room and let the server work away in the basement and make as much noise as it wants
| |
19:01 | <vagrantc> right
| |
19:01 | <cooteian> whenever I try to login, it tries and comes back wiht a segmentation fault
| |
19:01 | <vagrantc> RyanRyan52: i'll check out the ldm-debug branch
| |
19:01 | oh, 7.04 is the old python ldm ...
| |
19:02 | cooteian: try ltsp-update-sshkeys
| |
19:02 | <cooteian> I've tried that, don't you have to edit the file and add the ip address to the beginning?
| |
19:02 | <vagrantc> it should handle that for you
| |
19:03 | <cooteian> ok, ill run that. What username do I use at the login screen?
| |
19:03 | <vagrantc> cooteian: whatever your user is on the server
| |
19:03 | <cooteian> ok, i'll try that, thanks a mint
| |
19:06 | <vagrantc> RyanRyan52: maybe putting some of the debugging into more of the conditionals ... ?
| |
19:07 | RyanRyan52: i.e. in the if statements
| |
19:08 | <RyanRyan52> vagrantc: but I'm confused because its dying right after the load_guestinfo function returns. It doesn't even make it to the thing after load_guestinfo
| |
19:08 | I don't know what to check for...its dying in such a wierd place...
| |
19:09 | <vagrantc> RyanRyan52: it's dieing between when load_guestinfo returns and the very next thing ... for sure ?
| |
19:09 | <RyanRyan52> thats what the debug statements tell me...and I would assume they don't lie :P
| |
19:09 | <vagrantc> RyanRyan52: the last debug log you've got in there is right before the return ...
| |
19:09 | <RyanRyan52> yes, but in the main funtion there are other ones (ya0 before load_guestinfo and ya1 after)
| |
19:10 | <vagrantc> right
| |
19:10 | it is certainly perplexing where it's failing ...
| |
19:10 | works fine on sid at least :)
| |
19:10 | <RyanRyan52> ya
| |
19:11 | <vagrantc> we could run ldm from strace or something ...
| |
19:12 | <RyanRyan52> How do I do that...I know how to use strace...but how do I do it from a thinclient?
| |
19:12 | and with a window manager...
| |
19:12 | err, login manager
| |
19:13 | <vagrantc> RyanRyan52: edit the chroot's /usr/lib/ltsp/screen.d/ldm to call ldm with strace
| |
19:13 | RyanRyan52: and then append the strace output to a log on the client or something
| |
19:13 | <RyanRyan52> okay
| |
19:17 | <cooteian> vagrant, I tried running ltsp-update-sshkeys, ran /etc/init.d/ssh restart and rebooted the client, to no avail
| |
19:17 | <vagrantc> cooteian: you can log in using ssh as the user and password ?
| |
19:18 | <cooteian> how do you mean?
| |
19:18 | type "ssh" into the login screen?
| |
19:18 | <vagrantc> cooteian: log in from another machine with that username and password
| |
19:18 | using ssh
| |
19:19 | i.e. ssh USER@SERVER
| |
19:19 | just to make sure ssh works with that user and host
| |
19:19 | <cooteian> I am sshing into the server as we speak, thats' how i ran update-sshkeys
| |
19:20 | vagrant, sorry I just got a call from my old man and I have to go pick up a pizza, thanks for the help
| |
19:21 | <vagrantc> good luck
| |
19:47 | <Ryan52> vagrantc: strace broke it...it made it keep connecting to ldminfod according to syslog, but it was dead (it would not respond) on the thinclient...any other ideas?
| |
19:47 | <vagrantc> ugh
| |
19:47 | hope this version of ldm hits lenny and lenny gets released soon? :)
| |
19:48 | <Ryan52> heh
| |
19:48 | works for me :)
| |
19:48 | <vagrantc> *sigh*
| |
19:49 | i was really hopeing to be able to switch freegeek over sooner than later ... ldm in sid is feature-complete for freegeek's needs, i think.
| |
19:49 | * Ryan52 will make it print all of its variables and make sure they are all how they should be... | |
19:49 | * Ryan52 wants freegeek to use ldm too... | |
19:50 | <vagrantc> i think we could run a lenny ltsp environment with etch servers without much trouble.
| |
19:50 | other than lenny would be a constant upgrade cycle.
| |
19:50 | * RyanRyan52 still wants to figure out why it doesn't work in etch...its bugging him now | |
20:22 | spectra has quit IRC | |
20:24 | topslakr_ has joined #ltsp | |
20:28 | <RyanRyan52> vagrantc: I fixed it
| |
20:29 | vagrantc: Since load_guestinfo is only being called from one place I just moved all of its code into main where it was called and it works now...creepy
| |
20:30 | cooteian has quit IRC | |
20:31 | * RyanRyan52 wonders why it did that...its just soooooo wierd... | |
20:41 | <RyanRyan52> vagrantc: http://llama.freegeek.org/~ryan52/bzr/ltsp/ldm-etch/
| |
20:42 | vagrantc: I tested it with sid and etch, using autologin, guest and normal login and everything works
| |
20:42 | vagrantc: but you should still double check because I think I'm going insane...
| |
20:46 | <vagrantc> RyanRyan52: heh
| |
20:46 | crazy.
| |
20:46 | that's ... obnoxious
| |
20:47 | RyanRyan52: i think i'll just have to maintain a patch for etch ... that's too ugly to push upstream
| |
20:55 | <RyanRyan52> okay...
| |
21:04 | ian_ has joined #ltsp | |
21:06 | <ian_> hello all, I was in here earlier with questions about logging into a client...I re-ran ltsp-update-sshkeys and rebooted everything, but I still can't log in
| |
21:06 | Whenever I try, it says it's validating the password, and after about 5 minutes it blanks the screen and says segmentation fault
| |
21:35 | anybody there?
| |
21:36 | <vagrantc> ian_: can't think of anything more to try ... segfaulting is really unusual
| |
21:36 | <ian_> what does it mean exactly?
| |
21:37 | <vagrantc> i don't even know what it means *exactly*, but it's a pretty major failure.
| |
21:37 | you might want to run memtest on your thin-client
| |
21:37 | sometimes bad memory will cause weird segfaults
| |
21:38 | ogra has quit IRC | |
21:38 | <ian_> ok, it's a pretty old machine, that might be connected to it
| |
21:38 | <warren> you can easily enough add memtest to your pxe menu
| |
21:38 | ogra has joined #ltsp | |
21:38 | <warren> (if your pxe is anything like the one in fedora)
| |
21:38 | <ian_> pxe menu?
| |
21:38 | I don't know if i have a pxe menu
| |
21:38 | <warren> you probably don't see it by default
| |
21:39 | <ian_> my pxelinux.cfg file needs to be edited, i suppose
| |
21:39 | there's just one line
| |
21:39 | <vagrantc> ian_: apt-get install memtest86+
| |
21:39 | <warren> ooh, I'm the owner of memtest86+ in fedora. I can easily make it auto-add to the pxe menu. =)
| |
21:40 | <ian_> when I install it, i do it on the server right?
| |
21:40 | <vagrantc> ian_: and then: cp /boot/memtest86+ /var/lib/tftpboot/ltsp/i386/memtest
| |
21:40 | <ian_> I get a bit confused about what needs to be where, as there is the filesystem on the server and on the client machines in /opt/ltsp
| |
21:40 | <vagrantc> (pxelinux sometimes has issues with +)
| |
21:41 | ian_: i will always be explicit.
| |
21:41 | <warren> ian_, memtest86+ needs to boot instead of the thin client's kernel
| |
21:41 | ian_, so yes, install it on the server
| |
21:41 | <vagrantc> ian_: and then edit /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
| |
21:41 | well, you can install it in the chroot if you want, and then run ltsp-update-kernels ... but no need to confuse ian_ any extra :)
| |
21:41 | <ian_> ok, that makes sense now
| |
21:42 | that's something I need to research a bit more, I've heard of chroot, and I think it's probably important. I'll do some googling
| |
21:42 | Thanks again for all your help
| |
21:42 | <RyanRyan52> ian_: a seg fault is when a program tries to access a memory area that it is not allowed to access
| |
21:43 | <ian_> could it be a perimssions issue?
| |
21:43 | <vagrantc> ian_: and then: echo DEFAULT memtest > /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
| |
21:43 | <RyanRyan52> no...not that kind of allowing
| |
21:44 | the linux kernel gives each program section of ram and when a program tries to access another programs area it gets in trouble by the kernel
| |
21:44 | <ian_> that sounds like something has gone pretty haywire then.
| |
21:44 | so when it tries to login, it's trying to access something that it shouldn't?
| |
21:44 | <RyanRyan52> its usually caused by the programmer screwing up, but I've seen it on bad hardware too.
| |
21:45 | <ian_> I might want to start over with a clean install
| |
21:45 | <RyanRyan52> ian_: do memtest
| |
21:45 | <ian_> it's gotten a bit confused in this machine anyway
| |
21:45 | I'll try the memtest first
| |
21:45 | quick last question, then I'll stop bothering you
| |
21:46 | <RyanRyan52> ian_: bother away :D
| |
21:46 | <vagrantc> memtest will take a while ... depending on how fast your processor and how much memory you have
| |
21:46 | <RyanRyan52> ian_: or it will go on forever if you let it :P
| |
21:46 | <ian_> If I want to run programs on the client, they need to be installed in /opt/ltsp/i386/usr/bin not on /usr/bin on the server?
| |
21:47 | <RyanRyan52> ian_: if you are logging in with ldm, then everything you do after that is on the server
| |
21:47 | <ian_> I have very little memory, somehting like 128 meg on the client, and 512 on the server
| |
21:47 | <RyanRyan52> ian_: so install them on the server
| |
21:47 | <ian_> I don't know if I'm loggin in with ldm
| |
21:47 | <RyanRyan52> ian_: you probably are then
| |
21:48 | <ian_> ok, so If i have mythtv installed right now, then I don't need to reinstall it on the /opt directory to get it running on the client?
| |
21:49 | <vagrantc> ian_: everything runs on the server, unless you've done a LOT of work.
| |
21:49 | ian_: just treat the thin clients like extra monitors, keyboards, mice, etc. for your server.
| |
21:50 | <ian_> I haven't done a lot of work
| |
21:50 | That's what I thought, but it's my first foray into the thin client world
| |
21:51 | Thanks again fellas
| |
21:59 | <warren> Hmm, in dhcpd.conf...
| |
21:59 | can the option option have any name?
| |
21:59 | or there is a set of possible options?
| |
22:08 | <RyanRyan52> warren: I think you have to define names to have corresponding code that the client can understand...but I'm not completely sure
| |
22:13 | * RyanRyan52 can't find any examples for it for dhcp3-server...only version 2 | |
22:15 | cyberorg has quit IRC | |
22:25 | <warren> vagrantc, what is the Debian/Ubuntu dhcpd.conf syntax to tell the client "Use NBD at this IP:port (not NFS)"?
| |
22:25 | vagrantc, or you need to rebuild your initramfs to switch between the two?
| |
22:26 | vagrantc, the way I see it, a single initrd should be able to support whatever the dhcp server tells it to do.
| |
22:29 | <vagrantc> RyanRyan52: the patch for etch seems to be working
| |
22:29 | <warren> ah..
| |
22:29 | man dhcp-options
| |
22:30 | it appears that I can define any option I want
| |
22:30 | I just need the client to interpret my custom option
| |
22:30 | however I would really want to make it uniform with ubuntu if possible
| |
22:30 | <vagrantc> warren: currently, the way to determine is actually with boot prompt options ... shouldn't be too hard to add dhcp-based stuff
| |
22:31 | <warren> vagrantc, our initrd already contains code to boot a NFS root without any IP or paths hard coded and without the help of boot prompt defined things either.
| |
22:31 | <vagrantc> warren: well, there's a basically hard-coded default, which is overridden with boot prompt options ... can probably throw dhcp into the middle.
| |
22:31 | <warren> I rather have a single initrd that doesn't have any hard-coded anything
| |
22:31 | and the dhcp server tells it what to do
| |
22:32 | <vagrantc> warren: the ip addresses and such come from dhcp
| |
22:32 | warren: the only thing hard-coded right now is weather to NFS or NBD
| |
22:32 | <warren> vagrantc, but nbd or nfs comes from boot prompt?
| |
22:32 | ah
| |
22:32 | yeah
| |
22:32 | I want that to come from dhcp as well
| |
22:33 | vagrantc, if your boot prompt tells it NBD, it still gets the IP address and port from root-path?
| |
22:33 | <vagrantc> i would like that as well, although i think there should be a "reasonable default" hard-coded if dhcp doesn't supply which
| |
22:33 | <warren> instead of IPADDR:/opt/ltsp/i386 it gets IPADDR:PORT?
| |
22:33 | vagrantc, ok, reasonable default isn't bad.
| |
22:33 | <vagrantc> warren: i don't know the NBD stuff very well. i've only used it a small handful of times.
| |
22:34 | warren: could probably do some trickery with root-path dhcp option ...
| |
22:34 | <warren> well, I want to define something that we can both use
| |
22:34 | <vagrantc> root-path NBD:IPADDR:PORT
| |
22:34 | <warren> it seems that we can define any custom dhcp option we want
| |
22:34 | I'm not sure it is a good idea to overload the existing root-path
| |
22:34 | <vagrantc> i'd rather not throw new options into the mix
| |
22:35 | but whatever
| |
22:35 | <warren> I'm going to ask our technical lead on this part what he thinks
| |
22:42 | vagrantc, if you look at "man dhcp-options" you see all kinds of crazy options that are defined as "official"
| |
22:42 | vagrantc, including a bunch for Novell
| |
22:42 | vagrantc, I really don't see a problem with us just agreeing upon a new option.
| |
22:42 | vagrantc, root-path is pretty well defined for NFS. We might want to define something like nbd-root-path
| |
22:43 | vagrantc, otherwise older software will blow up in unpredictable ways with our overloading
| |
22:52 | <vagrantc> ok
| |
22:53 | * vagrantc wonders how hard it will be to get klibc-utils to support new dhcp options | |
22:53 | <vagrantc> warren: would be good to contact nbd upstream ... which *might* be the debian maintainer.
| |
22:54 | warren: i know the debian maintainer has been working on nbdroot support
| |
22:54 | <warren> oh
| |
23:11 | ltspbot has joined #ltsp | |