IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 31 March 2007 (all times are UTC)
| 00:11 | freemind has joined #ltsp | |
| 00:13 | Fosforo has quit IRC | |
| 00:29 | Fosforo has joined #ltsp | |
| 00:35 | mistik1_ has joined #ltsp | |
| 00:36 | mistik1 has quit IRC | |
| 00:36 | mistik1_ is now known as mistik1 | |
| 00:45 | Fosfor1 has quit IRC | |
| 01:22 | freemind has quit IRC | |
| 01:29 | mistik1 has quit IRC | |
| 02:47 | Shingoshi has joined #ltsp | |
| 04:10 | vlt|l has quit IRC | |
| 04:41 | vlt|l has joined #ltsp | |
| 04:42 | <vlt|l> (sorry, if multiple posts, but seemed to be offline): Hello. I compiled hplip 1.7.3 in /opt/ltsp/i386 to support my HPLaserJet printer/scanner *on the client*. When I run `/etc/init.d/hplip start` there I get "[OK]" but syslog says "hpiod: unable to bind socket 2208: Cannot assign requested address". I tried to use this port 2208 manually with `netcat` and suceeded. What could be hplip's problem here?
| |
| 04:48 | vlt|l has left #ltsp | |
| 04:57 | Shingoshi has quit IRC | |
| 05:01 | Shingoshi has joined #ltsp | |
| 05:04 | vlt|l has joined #ltsp | |
| 05:11 | <vlt|l> Hello. I want to run hplip 1.7.3 I compiled and installed successfully the way described in the wiki on the hplip site to /opt/ltsp/i386 but get "hpiod: unable to bind socket 2208: Cannot assign requested address" when trying to run it *on the client*. I did an an strace for starting hpssd.py: http://rafb.net/p/QrG2Rd29.html -- Can anynone see what's hplip's problem here?
| |
| 05:18 | monteslu has quit IRC | |
| 05:21 | Shingoshi has quit IRC | |
| 05:25 | Shingoshi has joined #ltsp | |
| 05:36 | <vlt|l> Hello. (How) can I disable the "poweroff switch" in the login screen?
| |
| 05:52 | <cliebow> there is something called pessulus you can apt-get to control a lot of functions
| |
| 05:52 | in edubuntu
| |
| 06:01 | <vlt|l> cliebow: Thank you.
| |
| 06:03 | <cliebow> check it out anyway..not 100% it will alter gdm..
| |
| 06:10 | might look at gdmsetup as well
| |
| 06:10 | <vlt|l> Hello. I finally managed to run hplip sand saned on the cleint. Does anyone of you know how to tell `scanimage` (on the server) to access the remote saned? Can't find anything about it in scanimage's manpage.
| |
| 06:11 | cliebow: pessulus doesn't seem to offer control about that. I'll look at gdmsetup ...
| |
| 06:20 | robbie has joined #ltsp | |
| 06:28 | <vlt|l> It's `scanimage --device-name net:10.0.0.1:hpaio:/usb/...` in my case. Solved.
| |
| 06:30 | iMav has quit IRC | |
| 06:30 | <cliebow> http://ubuntu.wordpress.com/2006/03/20/disable-shutdown-for-normal-users/
| |
| 06:42 | gone to spread the gospel at melug
| |
| 06:42 | robbie has quit IRC | |
| 06:44 | plamengr has joined #ltsp | |
| 07:11 | bobby_C has joined #ltsp | |
| 07:25 | rpetre has joined #ltsp | |
| 07:37 | vlt|l has quit IRC | |
| 08:27 | monteslu has joined #ltsp | |
| 08:32 | bobby_C has quit IRC | |
| 08:33 | Fosfor1 has joined #ltsp | |
| 08:46 | irule has quit IRC | |
| 08:50 | Fosforo has quit IRC | |
| 08:54 | plamengr has quit IRC | |
| 08:56 | cliebow_ has joined #ltsp | |
| 08:56 | <cliebow_> Halloo
| |
| 08:57 | gepatino has joined #ltsp | |
| 08:57 | plamengr has joined #ltsp | |
| 09:25 | edgarin has quit IRC | |
| 09:31 | plamengr has left #ltsp | |
| 09:32 | cliebow_ has quit IRC | |
| 10:09 | muh2000_ has joined #ltsp | |
| 10:11 | <gepatino> hi everybody
| |
| 10:11 | does anyone tried to run a thin client using vnc, instead of full x version?
| |
| 10:12 | monteslu has quit IRC | |
| 10:12 | <gepatino> i mean, instead of running just a X server (with or without ssh) at the client, wouldn't it be faster to run a vncclient in the client side?
| |
| 10:13 | for low bandwith networks, or relly slow clients?
| |
| 10:14 | monteslu has joined #ltsp | |
| 10:16 | cliebow_ has joined #ltsp | |
| 10:25 | muh2000 has quit IRC | |
| 10:36 | monteslu has quit IRC | |
| 10:37 | monteslu has joined #ltsp | |
| 10:44 | edgarin has joined #ltsp | |
| 10:46 | ccjoe has quit IRC | |
| 10:46 | cliebow__ has joined #ltsp | |
| 10:48 | cliebow__ has quit IRC | |
| 10:55 | vagrantc has joined #ltsp | |
| 11:01 | gepatino has quit IRC | |
| 11:05 | iMav has joined #ltsp | |
| 11:08 | gentgeen__ has joined #ltsp | |
| 11:09 | <stillflame> vagrantc: okay, so looking at the workstation instructions, i don't want to install the 'desktop' package again. does just gdm and icewm sound good? the dependencies don't pull in any nasties that i can see.
| |
| 11:10 | <vagrantc> stillflame: sure.
| |
| 11:14 | <rpetre> any idea why the X server run by "startx" keeps messing my keyboard, although if i run it from shell (with tha same parameters) works fine?
| |
| 11:14 | <vagrantc> rpetre: what linux distro?
| |
| 11:14 | <rpetre> debian
| |
| 11:15 | both the server and the slient
| |
| 11:15 | <vagrantc> what SCREEN_NN did you set startx to?
| |
| 11:15 | <rpetre> 01
| |
| 11:15 | <vagrantc> a login is already running there.
| |
| 11:15 | you need to put it on 07+
| |
| 11:15 | <rpetre> oh
| |
| 11:15 | thank you
| |
| 11:15 | <vagrantc> presuming this is ltsp5 built with ltsp-server
| |
| 11:15 | <rpetre> yes, nice job, btw ;)
| |
| 11:16 | <vagrantc> we try :)
| |
| 11:16 | <rpetre> you saved my butt, although a bit more documentation will be helpful
| |
| 11:16 | <vagrantc> we would gladly accept any documentation :)
| |
| 11:17 | rpetre: please contribute to http://wiki.debian.org/LTSP
| |
| 11:17 | <rpetre> heh, never thought to look there
| |
| 11:17 | i will
| |
| 11:18 | <vagrantc> there's also a little bit in /usr/share/doc/ltsp-server ... though i needs more
| |
| 11:18 | it
| |
| 11:19 | J45p3r has joined #ltsp | |
| 11:19 | <rpetre> yes, that's what i was talking about
| |
| 11:19 | and thank you for the _07 tip, it works ok
| |
| 11:19 | <vagrantc> rpetre: the other option is to disable more things in /opt/ltsp/ARCH/etc/inittab
| |
| 11:19 | <rpetre> the ssh connections kill my server
| |
| 11:20 | true, but i like it more like that
| |
| 11:20 | it behaves like a real system
| |
| 11:20 | <vagrantc> yah, tunneling everything through ssh has it's downsides, for sure.
| |
| 11:21 | <rpetre> currently the only bug i see is the long X configuration time and the messages from debconf (i think)
| |
| 11:21 | <vagrantc> yeah.
| |
| 11:21 | <rpetre> but i have no better idea
| |
| 11:22 | <vagrantc> we try as much as possible to use standard tools to configure X... using the debconf infrastructure... and debconf is slow
| |
| 11:22 | next generation might be better if cdebconf becomes a viable alternative
| |
| 11:22 | <rpetre> yes, i was delighted to see such a debian-way approach :)
| |
| 11:23 | <vagrantc> that's the whole idea behind ltsp5 ... use the distro's approach for as much as possible.
| |
| 11:23 | <rpetre> also, i have the dhcpd and tftpd on a different server, this causes a bit of extra work when upgrading
| |
| 11:23 | <vagrantc> indeed.
| |
| 11:23 | <rpetre> but i have no better idea for making this easier, as well
| |
| 11:24 | <vagrantc> i hope to split the packages up a little more... we'll see.
| |
| 11:24 | J45p3r has quit IRC | |
| 11:28 | <rpetre> well, thank you again, and as soon as i put together some install/usage notes, i'll put them on the wiki
| |
| 11:28 | rpetre has left #ltsp | |
| 11:38 | cliebow_ has quit IRC | |
| 12:00 | iMav has quit IRC | |
| 12:06 | plamengr has joined #ltsp | |
| 12:08 | kaminski-ltsp-br has quit IRC | |
| 12:21 | vagrantc has quit IRC | |
| 12:25 | yhan has joined #ltsp | |
| 12:26 | <yhan> hi, I try to setup the sound on a terminal with a sound card cyrix 5510, how can I debug it ?
| |
| 12:52 | monteslu has quit IRC | |
| 13:10 | <jammcq> g'afternoon kids
| |
| 13:10 | <yhan> hi dad
| |
| 13:20 | <ogra> hey jammcq
| |
| 13:21 | <jammcq> ogra: hey
| |
| 13:22 | did you see my comments yesterday, about Debian ltsp-5 working so much faster than ubuntu ltsp-5 on that e2300 ?
| |
| 13:22 | <ogra> jammcq, i'm fighting with ldm :/ ... apparently its not the greeter thats slow, i just tested with a plain C app thats not taking any memory and is extremly speedy ... takes exactly 21secs to start as long as the pyton greeter
| |
| 13:22 | <jammcq> hmm
| |
| 13:22 | and yet, with Debian, the python ldm only takes a total of 23 secs
| |
| 13:22 | <ogra> its something else that slows us down
| |
| 13:22 | <jammcq> yes, I agree... Now, we just need to figure out what it is
| |
| 13:23 | <ogra> from starting X to get a login screen ?
| |
| 13:23 | how do you measure that ?
| |
| 13:23 | <jammcq> from seeing the 'X' in the middle of the screen, to getting the complete login screen is 23 secs
| |
| 13:23 | <ogra> right
| |
| 13:23 | <jammcq> it's about 105 seconds on Ubuntu
| |
| 13:23 | <ogra> same here ...
| |
| 13:23 | <jammcq> that's on a eBox e2300
| |
| 13:23 | <ogra> that must be a kernel issue ... ldm is exactly the same in both
| |
| 13:23 | yhan_ has joined #ltsp | |
| 13:24 | <jammcq> the t1220 is MUCH faster
| |
| 13:24 | http://wiki.ltsp.org/twiki/bin/view/Ltsp/BootSpeedeBox2300
| |
| 13:24 | check that page
| |
| 13:24 | <ogra> ubuntus ldm has one additional button, but i doubt that has any effect
| |
| 13:24 | <jammcq> so, maybe it's NFS slowness
| |
| 13:24 | <ogra> yeah, thats odd
| |
| 13:24 | <yhan_> how can I use a usb sound device on a terminal ?
| |
| 13:25 | for instance with a usb headset
| |
| 13:25 | <jammcq> yhan_: what version of LTSP ?
| |
| 13:25 | hmm, microphone too ?
| |
| 13:25 | <yhan_> the one who came with ubuntu 6.10
| |
| 13:25 | <ogra> if you boot with it and are lucky :)
| |
| 13:25 | <yhan_> mike would be the cherry ontop of the sundae
| |
| 13:25 | <ogra> there is no code for detection yet
| |
| 13:26 | no way to get a mic to work in a default 6.10
| |
| 13:26 | 7.04 builds the base for that feature, but also isnt doing anything automatically for mics yet
| |
| 13:26 | <yhan_> actually on my term, which has a cyrix 5530 chipset, I can't have the sound working
| |
| 13:26 | <ogra> (7.04 switched to pulseaudio)
| |
| 13:27 | <yhan_> and I was wondering if a usb headset would be the solution
| |
| 13:27 | I have a driver for oss for the 5530 chipset, but not for alsa
| |
| 13:28 | <jammcq> yeah, the 5530 is kind of a bastard chipset
| |
| 13:28 | the geodes put some of the logic in the bios, so supporting the chipset is difficult in the kernel. you have to support not only the chipset, but also the bios
| |
| 13:28 | <yhan_> the video card works with vesa but not with the "native" driver
| |
| 13:28 | <jammcq> right
| |
| 13:28 | same problem
| |
| 13:28 | <yhan_> the term is quite slow :(
| |
| 13:28 | <edgarin> jammcq, hi
| |
| 13:28 | <jammcq> well, that too
| |
| 13:28 | edgarin: HOWDIE !!!
| |
| 13:29 | <yhan_> do you guys use the compression with ssh while tunneling to the application server ?
| |
| 13:30 | <ogra> i wouldnt with that term ...
| |
| 13:30 | compression needs cpu power ...
| |
| 13:31 | <yhan_> if it was enabled by default, I would disable it
| |
| 13:32 | to get a bit more cpu for the video rendering
| |
| 13:32 | the term has only a cyrix 233 mhz cpu
| |
| 13:32 | not that fast
| |
| 13:32 | <edgarin> jammcq, all fine and you?
| |
| 13:32 | <jammcq> doing great
| |
| 13:32 | yhan has quit IRC | |
| 13:33 | <yhan_> an another question, is that a way to speed up the boot time ?
| |
| 13:33 | <jammcq> heh
| |
| 13:33 | that's exactly what we are working on right now
| |
| 13:33 | <yhan_> my term takes more or less 6-10minutes to boot
| |
| 13:34 | <jammcq> yep
| |
| 13:34 | <ogra> 6-10min ! ?
| |
| 13:34 | phew
| |
| 13:34 | <jammcq> have you actually timed it? or is that just an estimate ?
| |
| 13:34 | <yhan_> it's an estimate
| |
| 13:34 | look, I'm gonna to time it
| |
| 13:34 | <jammcq> if you get a chance, please give us a more accurate timing
| |
| 13:34 | * ogra has nothing that boots longer than 90sec here ... | |
| 13:34 | <yhan_> ogra, with a 233 mhz machine ...
| |
| 13:35 | <jammcq> ogra: the T-1220 takes 93 seconds
| |
| 13:35 | <ogra> my smallest is 300
| |
| 13:35 | <yhan_> now
| |
| 13:35 | term is booting
| |
| 13:35 | the term is a visara 1730
| |
| 13:36 | I have a PII-400 (regular pc ) which boots faster
| |
| 13:36 | <ogra> whee
| |
| 13:36 | <yhan_> this same server, same network switch
| |
| 13:36 | <ogra> our feisty kernel doesnt usderstand "proto"
| |
| 13:37 | as NFSOPT
| |
| 13:37 | * ogra tries nolock only | |
| 13:38 | <ogra> but i think thats set already by default
| |
| 13:38 | <jammcq> ogra: can you try: actimeo=600,nocto
| |
| 13:38 | eTyack has told me that helps on some clients on ltsp-4.2
| |
| 13:39 | <ogra> takes still ages for the greeter to come up
| |
| 13:39 | i dont realy belive its nfs
| |
| 13:39 | *really
| |
| 13:39 | <jammcq> what's your hunch ?
| |
| 13:39 | <ogra> the other services are fast
| |
| 13:39 | no idea
| |
| 13:40 | <jammcq> well, we gotta rule things out
| |
| 13:40 | <ogra> but nfs doesnt seem to be it
| |
| 13:40 | <jammcq> is it possible that it's actually compiling the LDM stuff?
| |
| 13:40 | <ogra> the thing is that ldm starts the X server ... and five lines below it starts the greeter
| |
| 13:40 | <jammcq> ok, so it's not really "LDM" that is slow, it's the greeter
| |
| 13:40 | <ogra> there is absolutely *nothing* inbetween that could slow anything down
| |
| 13:41 | but that exactly the palce where its slow
| |
| 13:41 | no
| |
| 13:41 | i'm using a greeter thats 19k C binary stuff
| |
| 13:41 | <jammcq> ok, so that's not it
| |
| 13:41 | <ogra> eating less than a megabyte
| |
| 13:41 | <jammcq> which client are you using?
| |
| 13:41 | <ogra> the red/black one
| |
| 13:41 | <jammcq> k, that's the t-1220
| |
| 13:41 | <ogra> th eopenchrome one ...
| |
| 13:42 | <jammcq> yep
| |
| 13:42 | <ogra> right
| |
| 13:42 | <jammcq> so... why does Debian run it so fast ?
| |
| 13:42 | <ogra> to sad openchrome is so unstable
| |
| 13:42 | bad option: actimeo
| |
| 13:42 | btw^
| |
| 13:43 | i'd guess for the kernel still...
| |
| 13:43 | an surely debian doesnt use usplash
| |
| 13:43 | but thats a tradeoff i'm willing to pay
| |
| 13:43 | <jammcq> i've got usplash turned off
| |
| 13:43 | is NFSOPTS being passed to mount? or to loading the nfs kernel module?
| |
| 13:44 | <ogra> should be to mount
| |
| 13:44 | <yhan_> ok it's booted
| |
| 13:44 | <ogra> i'd love to be able to set proto
| |
| 13:44 | <yhan_> so 9mins to boot
| |
| 13:44 | <ogra> :(
| |
| 13:44 | why doesnt that work
| |
| 13:45 | grmpf and why dont i have the kernel docs here
| |
| 13:45 | <jammcq> who's 'mount' are you using?
| |
| 13:45 | <ogra> klibc i think
| |
| 13:46 | which is silly
| |
| 13:46 | <jammcq> i'm using busybox I think
| |
| 13:46 | <ogra> we could just use mount :)
| |
| 13:46 | <jammcq> the "real" one?
| |
| 13:46 | <ogra> sure
| |
| 13:46 | its initramfs
| |
| 13:46 | <jammcq> you just need to include it
| |
| 13:46 | is it already there?
| |
| 13:46 | <ogra> copying that is trivial
| |
| 13:46 | hmm, might even be there, not sure
| |
| 13:46 | <jammcq> hmm, you might need some libs too
| |
| 13:47 | <ogra> libc ... what else ?
| |
| 13:48 | <jammcq> hmm, maybe that's all you need
| |
| 13:48 | <ogra> and libc is included anyway i think
| |
| 13:48 | (thats why its so silly to use klibc :) )
| |
| 13:50 | <jammcq> ogra: are you using "-o proto..."
| |
| 13:50 | <ogra> yes
| |
| 13:50 | nolock works in the same line
| |
| 13:51 | nfsmount -o nolock ${roflag} ${NFSOPTS} ${NFSROOT} ${rootmnt}
| |
| 13:51 | thats what initramfs calls
| |
| 13:52 | <jammcq> so you end up with: nfsmount -o nolock -o ro -o proto=tcp ${NFSROOT} ${rootmnt}
| |
| 13:52 | <ogra> ah
| |
| 13:52 | right
| |
| 13:52 | <jammcq> can it handle multiple '-o' ?
| |
| 13:52 | <ogra> but it should be able to
| |
| 13:52 | lets try
| |
| 13:52 | <jammcq> I always do it with one
| |
| 13:52 | -o nolock,ro,proto=tcp
| |
| 13:52 | <ogra> you can look at it yourself btw: /usr/share/initramfs-tools/scripts/nfs
| |
| 13:52 | <jammcq> yep
| |
| 13:53 | i'm there already
| |
| 13:53 | looks like it's already gonna do -o nolock -o ro
| |
| 13:53 | <ogra> right
| |
| 13:54 | wah
| |
| 13:54 | <jammcq> so, which kernel config file should I have used?
| |
| 13:54 | <ogra> nfsmount: need a server
| |
| 13:54 | unbelivable
| |
| 13:54 | smells slightly buggy
| |
| 13:55 | so you need -o in there
| |
| 13:56 | bad option "proto" ....
| |
| 13:56 | tsk
| |
| 13:57 | * ogra tries something | |
| 13:58 | <ogra> seems it survives changing nfsmount to mount :)
| |
| 14:07 | yay, i just shoved off 5 seconds :)
| |
| 14:08 | there is a sleep(5) in /usr/sbin/ldm :)
| |
| 14:08 | no idea why ...
| |
| 14:08 | surely not needed there
| |
| 14:10 | Fosfor1 has quit IRC | |
| 14:13 | Fosforo has joined #ltsp | |
| 14:14 | <jammcq> hmm, that could mean that Debian ldm could get down to 18 secs :)
| |
| 14:15 | <ogra> mine is down to 15
| |
| 14:15 | on the 1220
| |
| 14:15 | still way to much
| |
| 14:15 | <jammcq> yep
| |
| 14:15 | <ogra> i'm fine with 10 and i'd *like* 5
| |
| 14:16 | <jammcq> 5-10 would be wonderful
| |
| 14:16 | <ogra> but i still dont understad *why* its so slow
| |
| 14:16 | there is nothing that could slow it down
| |
| 14:17 | i mean, look at /usr/sbin/ldm ... if ou comment out the sleep and all xauth stuff, there is nothing between starting X and starting the greeter ...
| |
| 14:17 | (thats what i did here )
| |
| 14:18 | apart from speeding up 5 secs because i dropped the sleep there is no noticeable change in dropping xauth ...
| |
| 14:18 | ... neither a slowdown
| |
| 14:18 | i wonder if X is our prob
| |
| 14:20 | <jammcq> deb is using 7.1, ubuntu using 7.2 ?
| |
| 14:20 | can't see how that would matter
| |
| 14:20 | <ogra> 7.0 i think
| |
| 14:21 | <jammcq> deb is using 7.1, I just checked
| |
| 14:21 | plamengr has quit IRC | |
| 14:21 | <ogra> deb froze some time ago
| |
| 14:21 | ok
| |
| 14:21 | <jammcq> biglap:/# dpkg -l | grep xorg
| |
| 14:21 | ii xorg 7.1.0-13 X.Org X Window System
| |
| 14:21 | <ogra> wel, its my only explanation
| |
| 14:21 | <jammcq> that's chroot'd to the deb chroot
| |
| 14:21 | ok, what does it take to build a new kernel?
| |
| 14:23 | <ogra> https://wiki.ubuntu.com/KernelCustomBuild
| |
| 14:23 | and take the -386 config as base
| |
| 14:23 | thats what we use in ltsp
| |
| 14:24 | <jammcq> I don't have any thin clients here at home, they're all at the office
| |
| 14:24 | <ogra> -generic is compiled for 586 and higher
| |
| 14:24 | there are many things different
| |
| 14:26 | <jammcq> i'm comparing debians config against ubuntus config
| |
| 14:26 | to see if anything jumps out at me
| |
| 14:26 | CONFIG_PARAVIRT=y is one that I see
| |
| 14:28 | hmm, CONFIG_BLK_DEV_RAM_SIZE
| |
| 14:28 | deb uses 8192, ubuntu uses 65536
| |
| 14:28 | it's nice having several different chroot's on the same machine
| |
| 14:28 | easy to compare things
| |
| 14:28 | * ogra tries to install linux-image-lowlatency in the chroot for fun | |
| 14:28 | <jammcq> is that a module?
| |
| 14:29 | or a whole kernel?
| |
| 14:29 | <ogra> no, a kernel
| |
| 14:29 | for the audiofreaks
| |
| 14:29 | <jammcq> might be interesting
| |
| 14:29 | <ogra> i dont expect it to be astonishing fast or so, but surely worth a try
| |
| 14:30 | <jammcq> hmm, both clients I test with have 8139 chipsets.
| |
| 14:30 | CONFIG_8139TOO_PIO
| |
| 14:30 | 'y' on deb, 'not set' on ubuntu
| |
| 14:30 | <ogra> whats that for ?
| |
| 14:30 | <jammcq> no clue
| |
| 14:30 | <ogra> heh
| |
| 14:30 | <jammcq> also, CONFIG_8139TOO_TUNE_TWISTER
| |
| 14:30 | 'y' on deb
| |
| 14:32 | <ogra> hmm ... this network module stuff reminds me ... i wanted to check it runs at fullduplex .... now i'm away from my client .. grmpf
| |
| 14:34 | <jammcq> CONFIG_DEBUG_KERNEL=y
| |
| 14:34 | on ubuntu
| |
| 14:34 | not set on debian
| |
| 14:34 | <ogra> only until release
| |
| 14:34 | <jammcq> ok, any chance our problems will go away on release?
| |
| 14:34 | <ogra> will be unset in the last upload
| |
| 14:36 | i'm pretty sure you wont notice debug on/off on a usual system ...
| |
| 14:36 | but it could be one thing causing our slowdown ...
| |
| 14:38 | <jammcq> I don't expect a single magic bullet will fix our problem. I'm thinking it's many things
| |
| 14:43 | mistik1 has joined #ltsp | |
| 14:48 | <yhan_> am I wrong or esd uses alsa as backend ?
| |
| 14:50 | noticed that on the esound website, A working sound device is needed
| |
| 14:51 | so if I have an OSS driver, which works on locally, what do I need to modify in lts.cfg or in an other file to make it work ?
| |
| 15:10 | tried this on the terminal /usr/bin/esd -nobeeps -public -tcp -d /dev/dsp
| |
| 15:10 | - using device /dev/dsp
| |
| 15:10 | ALSA lib pcm.c:2146:(snd_pcm_open_noupdate) Unknown PCM /dev/dsp
| |
| 15:11 | lfesdail has joined #ltsp | |
| 15:13 | lfesdail has quit IRC | |
| 15:29 | yhan_ has quit IRC | |
| 15:42 | <ogra> jammcq, hmm, the kernel docs state pretty clearly that its a bad ide to have nfs as a module if you run a diskless client
| |
| 15:43 | * ogra builds a kernel with just nfs builtin as only change ... | |
| 15:50 | * jammcq thinks that document is pre-initrd days, when you couldn't load the nfs module if you didn't have a rootfs to load it from | |
| 15:50 | <jammcq> but, sure would be cool if that fixes it
| |
| 15:53 | <ogra> well, it states 2006 as last copyright date
| |
| 15:55 | <jammcq> I still think that particular point is from long ago
| |
| 15:55 | but, maybe i'll help
| |
| 15:55 | <ogra> lets see
| |
| 15:55 | you still have the nfs error in your syslog
| |
| 15:55 | so that might fix it
| |
| 15:55 | apart from that i still suspect soemthing like halfduplex
| |
| 15:56 | even though that would also affect the session
| |
| 15:56 | but the session is usually snappy
| |
| 15:56 | <jammcq> yeah, it even runs ok on the ebox
| |
| 15:57 | <ogra> usually the server is the limiting bit on that ... so much i learned :)
| |
| 15:58 | hey, canonical moved offices today :)
| |
| 15:58 | http://en.wikipedia.org/wiki/Millbank_Tower ;)
| |
| 15:58 | some more space than the 40sqm we had before ;)
| |
| 15:59 | <jammcq> somebody should update the wikipedia article to point out that it's the headquarters for Canonical/ubuntu
| |
| 15:59 | <ogra> heh
| |
| 16:00 | not sure they want to make it to public :)
| |
| 16:00 | but i think we have a full floor there now ... lots of conference room space
| |
| 16:00 | <jammcq> umm, UDS 2008
| |
| 16:00 | <ogra> heh
| |
| 16:01 | unikely ... but in the future all sprints will take place there
| |
| 16:01 | <jammcq> looks like I'll be back at Google in June for a Desktop architecture mtg
| |
| 16:01 | <ogra> gah, all these modules take ages to build, the kernel was done 20 min ago ...
| |
| 16:01 | nice !
| |
| 16:01 | google was great
| |
| 16:01 | good place to be
| |
| 16:01 | <jammcq> it was a pretty decent location
| |
| 16:03 | <ogra> i think we should spec an -ltsp kernel ... for UDS ... even though that gets me in the dangerous position that i have to maintain a kernel :(
| |
| 16:03 | i suspect setting it to M386 would giveus a lot
| |
| 16:03 | <jammcq> why would M386 give you more than M486 ?
| |
| 16:04 | <ogra> the smalles setting we have is M486 (in the i386 flavor)
| |
| 16:05 | because many clone CPUs will be rather 386 than 486 ?
| |
| 16:05 | <jammcq> really ?
| |
| 16:05 | we've been using M486 for a long time
| |
| 16:05 | <ogra> hmm, k
| |
| 16:05 | <jammcq> but there's memory settings you can tweak
| |
| 16:05 | <ogra> well, we wont need SMP
| |
| 16:05 | <cliebow> jaammcq:i wowwed quite a few with ltsp today
| |
| 16:05 | <jammcq> and you can turn off virtualization and other things
| |
| 16:05 | cliebow: oh?
| |
| 16:05 | <cliebow> just my lappie and an ibok
| |
| 16:05 | <jammcq> out spreading the good word ?
| |
| 16:05 | <cliebow> at melug..
| |
| 16:06 | maine lug in fairfield..spreading the gospel..
| |
| 16:06 | <jammcq> so, yer like a big celebrity there, eh?
| |
| 16:06 | <cliebow> small pond..
| |
| 16:07 | <jammcq> heh
| |
| 16:07 | <cliebow> only 20 or so peoples..b ut several directors of vocational schools..
| |
| 16:08 | <jammcq> dtrask had an excellent email to the k12 list last week, talking about teaching concepts, instead of applications
| |
| 16:08 | some people just really get it
| |
| 16:10 | <cliebow> what dirves me nuts is the educ apps arent there in linux...
| |
| 16:10 | and with sound problematical ..til n ow..
| |
| 16:11 | and my clients many wont even run a 2.6 kernel
| |
| 16:12 | J45p3r has joined #ltsp | |
| 16:28 | markit has joined #ltsp | |
| 16:36 | markit has left #ltsp | |
| 17:03 | Regnad has joined #ltsp | |
| 17:04 | <cliebow> regdnad kcin?
| |
| 17:09 | <Regnad> Yeah, that's the one
| |
| 17:10 | No anchovies? You've got the wrong man!
| |
| 17:11 | Regnad has left #ltsp | |
| 17:15 | <jammcq> hmm, dunno what that was
| |
| 17:21 | cliebow_ has joined #ltsp | |
| 17:28 | <edgarin> cliebow_, saludos!
| |
| 17:31 | <cliebow_> Hola!
| |
| 19:07 | ccjoe has joined #ltsp | |
| 19:10 | <jammcq> ogra: ping
| |
| 19:17 | dererk has joined #ltsp | |
| 19:17 | <dererk> hey :)
| |
| 19:19 | sbalneav has joined #ltsp | |
| 19:19 | <jammcq> !s
| |
| 19:19 | <ltspbot> jammcq: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 19:19 | <sbalneav> Evening jammcq
| |
| 19:20 | <jammcq> sbalneav: what's shakin
| |
| 19:21 | <sbalneav> Not much, just doing some work on the edubuntu manual at the moment
| |
| 19:21 | <jammcq> neat-o
| |
| 19:21 | i'm working on building a new kernel for ltsp-5
| |
| 19:21 | muh2000_ is now known as muh2000 | |
| 19:21 | <jammcq> did you see our converstation yesterday about the Debian boot being so much faster than the Ubuntu boot ?
| |
| 19:21 | <cliebow_> jammcq;i been reading that
| |
| 19:22 | you looking at nfs weirdness?
| |
| 19:22 | <jammcq> yeah, but I just dunno if that's causing it
| |
| 19:22 | ogra is trying to build the kernel with nfs statically linked, but looking at the Debian kernel, the configs match Ubuntu
| |
| 19:23 | so i'm trying, by turning of things like virtualization
| |
| 19:23 | and some high memory settings
| |
| 19:23 | <cliebow_> ogra sais he found a 5 sec sleep there somewhere?
| |
| 19:23 | <jammcq> yeah, in LDM
| |
| 19:24 | 5 secs is nice, but the full boot is 4mins and 45 secs
| |
| 19:24 | so we need a little more than 5 secs
| |
| 19:24 | <cliebow_> Holy
| |
| 19:24 | my ibook is around 60 seconds..with i thought latest
| |
| 19:25 | to boot from the Sony
| |
| 19:25 | Mar 21 build
| |
| 19:32 | vagrantc has joined #ltsp | |
| 19:37 | <sbalneav> jammcq: oh, crumb. http://news.bbc.co.uk/2/hi/americas/6514999.stm
| |
| 19:37 | <jammcq> hmm, I think Marlon was affected by this today
| |
| 19:38 | I got a message from him that he was waiting 9 hours for a flight
| |
| 19:38 | lets just hope they get it all figured out in the next 10 days
| |
| 19:38 | Long queues snaked outside airport halls, particularly in Rio de Janeiro and Sao Paulo
| |
| 19:40 | <cliebow_> yeccch
| |
| 20:02 | Shingoshi has quit IRC | |
| 20:03 | Shingoshi has joined #ltsp | |
| 20:05 | Shingoshi has quit IRC | |
| 20:11 | cliebow_ has quit IRC | |
| 20:11 | Shingoshi has joined #ltsp | |
| 20:44 | ogra has quit IRC | |
| 20:45 | ogra has joined #ltsp | |
| 20:49 | dm_ has joined #ltsp | |
| 21:02 | edgarin_ has joined #ltsp | |
| 21:03 | dm has quit IRC | |
| 21:18 | edgarin has quit IRC | |
| 21:58 | robbie has joined #ltsp | |
| 22:24 | J45p3r has quit IRC | |
| 22:32 | <dererk> Hey :)
| |
| 22:32 | I've a little doubt
| |
| 22:33 | which is the latest version available?
| |
| 22:33 | <vagrantc> of ltsp ?
| |
| 22:34 | <dererk> Yes :)
| |
| 22:34 | <robbie> hey all
| |
| 22:34 | jammcq: sbalneav Faithful had another boy
| |
| 22:34 | <dererk> robbie, hi :)
| |
| 22:35 | <robbie> hi
| |
| 22:35 | <jammcq> robbie: Woooooo Hoooooooo!
| |
| 22:35 | <vagrantc> it's a complicated answer... essentially, some distros (ubuntu, debian) have an ltsp5 based variant, you can grab tarballs of those and install them on any distro, or install ltsp 4.2 on most other distros instead
| |
| 22:35 | <jammcq> and you named him "Jim", right?
| |
| 22:35 | <robbie> james
| |
| 22:35 | \he, i just realised
| |
| 22:35 | <jammcq> no shit ?
| |
| 22:35 | <robbie> yeah
| |
| 22:35 | <jammcq> ok then
| |
| 22:35 | <robbie> are you a james or a jim ?
| |
| 22:36 | <vagrantc> dererk: so... hows that for an answer?
| |
| 22:36 | <robbie> by birth
| |
| 22:36 | <dererk> vagrantc, I understand that 4.2 is out
| |
| 22:36 | <jammcq> glad to see i'm starting to get some respect around here :)
| |
| 22:36 | 'James A. McQuillan'
| |
| 22:36 | officially
| |
| 22:36 | <dererk> But, what's 5 then?
| |
| 22:36 | <robbie> cool
| |
| 22:36 | <vagrantc> !ltsp5
| |
| 22:36 | <ltspbot> vagrantc: "ltsp5" is ltsp5 is The next version of LTSP, and the future of Thin Client computing on GNU/Linux. http://wiki.ltsp.org/twiki/bin/view/Ltsp/Ltsp5
| |
| 22:36 | <robbie> you know Faithful's name is james ?
| |
| 22:36 | <jammcq> yeah, but you named him after me, right?
| |
| 22:36 | <robbie> of course
| |
| 22:36 | <dererk> vagrantc, thanks :D
| |
| 22:36 | <jammcq> ok
| |
| 22:36 | <vagrantc> dererk: it's a shift in thinking ...
| |
| 22:37 | <jammcq> robbie: you gonna call him 'Jimmie' ?
| |
| 22:37 | <robbie> james
| |
| 22:37 | i like jim though
| |
| 22:37 | <jammcq> me too
| |
| 22:37 | <robbie> but dont tell his mum :)
| |
| 22:37 | <dererk> vagrantc, so, a drogs and alcohol night eh? :P
| |
| 22:37 | joking :)
| |
| 22:37 | <robbie> jimmy is good name to be called by mates when at pub
| |
| 22:37 | <dererk> I'm gonna read this wiki doc
| |
| 22:37 | <vagrantc> dererk: that's a good start.
| |
| 22:37 | <dererk> :)
| |
| 22:55 | twinprism has joined #ltsp | |
| 23:11 | Shingoshi has quit IRC | |
| 23:12 | Shingoshi has joined #ltsp | |
| 23:16 | toddobryan_ has joined #ltsp | |
| 23:16 | <toddobryan_> Hi. I'm using LTSP 5 on Ubuntu 6.10.
| |
| 23:17 | <sbalneav> yep
| |
| 23:17 | <toddobryan_> I have a problem with transparency. A program that works on the server has white pixels where they should be transparent on the clients. I really have no idea where to start troubleshooting.
| |
| 23:18 | <sbalneav> What's the program? Can you just turn off the transparency?
| |
| 23:18 | <toddobryan_> It's a programming environment--DrScheme.
| |
| 23:19 | <jammcq> sounds like a video driver issue
| |
| 23:19 | <toddobryan_> These are animations my students are writing and the images don't look right.
| |
| 23:19 | <sbalneav> That's what I'd say. What kind of video cards are in the clients?
| |
| 23:19 | <toddobryan_> They're NTAVO 6020P's.
| |
| 23:19 | <sbalneav> That's the video card?
| |
| 23:19 | <toddobryan_> Via chipset.
| |
| 23:20 | <jammcq> what version of ltsp ?
| |
| 23:20 | ah, 6.10
| |
| 23:20 | <toddobryan_> Yeah. With MueKow.
| |
| 23:21 | I'll brb.
| |
| 23:24 | vagrantc has quit IRC | |
| 23:25 | <toddobryan_> Sorry about that.
| |
| 23:27 | I can't seem to find the video card on their website.
| |
| 23:28 | VIA VT8235
| |
| 23:28 | Does that sound like a video chip?
| |
| 23:29 | I'm almost sure that we don't have the right driver. When the clients boot, we get to "Ready." on the terminal.
| |
| 23:29 | <jammcq> for a few seconds?
| |
| 23:30 | <toddobryan_> If we let the screen go to black, like X is starting, eventually we get an all black screen and a freeze. Ctrl+Alt+F1 doesn't bring the terminal back.
| |
| 23:30 | <jammcq> so how do you bring it up?
| |
| 23:30 | <toddobryan_> If, however, we hit Ctrl+Alt+F1 on the black screen while the curse is still showing, it goes back to the terminal and X starts.
| |
| 23:30 | curse->cursor
| |
| 23:32 | jammcq: Have you considered selling parts/kits on disklessworkstations?
| |
| 23:32 | <jammcq> parts like what?
| |
| 23:32 | <toddobryan_> jammcq: Would have loved to buy from you, but couldn't pass up the cost savings.
| |
| 23:33 | Like a case, mobo, cpu to make a client.
| |
| 23:33 | <jammcq> well, how much support is devonIT giving you on this issue?
| |
| 23:34 | <toddobryan_> Not sure if that would be cost effective, but I thought I'd mention it.
| |
| 23:34 | Well, they told us to use the vesa driver...
| |
| 23:34 | I actually haven't contacted them about the most recent problem because I just noticed it Thursday when we started these new programs.
| |
| 23:34 | <jammcq> and they'll give you your money back, if it doesn't work out?
| |
| 23:35 | <toddobryan_> Probably not.
| |
| 23:35 | <sbalneav> You're using the vesa driver?
| |
| 23:35 | <toddobryan_> But they were only $129 each. :-)
| |
| 23:35 | Yeah.
| |
| 23:35 | <sbalneav> That's your problem.
| |
| 23:36 | <toddobryan_> No transparency?
| |
| 23:36 | <jammcq> I'd be happy to sell you somethng that doesn't work for $129
| |
| 23:36 | <sbalneav> Not with vesa, no.
| |
| 23:36 | <toddobryan_> That was kinda what I figured.
| |
| 23:36 | Guess I need to figure out which driver to use.
| |
| 23:37 | <jammcq> see, at DisklessWorkstations.com, we put an awful lot of effort and testing into our products to make damn sure they work. sometimes we run into problems, and we take back the product and refund the money.
| |
| 23:37 | but, enough said. this isn't a commercial channel
| |
| 23:38 | <toddobryan_> I'm actually more impressed with the awful lot of effort you put into LTSP.
| |
| 23:38 | <jammcq> yeah, I wonder how much effort Devonit puts into ltsp
| |
| 23:39 | <toddobryan_> They're apparently getting more involved. When Atlanta bought a bunch of stuff from them they set up a lab to troubleshoot some problems.
| |
| 23:39 | <jammcq> hmm, never seen any of them show up here in #ltsp or on the ltsp-discuss mailing list, or even contact any of us LTSP developers
| |
| 23:39 | <toddobryan_> jammcq: you do paid consulting, right?
| |
| 23:39 | <jammcq> yep
| |
| 23:40 | well, my company does
| |
| 23:40 | <toddobryan_> A student's family donated some cash to my lab as a thank-you for the college recs I wrote.
| |
| 23:40 | Maybe I could hire you to figure out the video problem.
| |
| 23:41 | Oh, and we haven't got sound working, either.
| |
| 23:41 | <jammcq> figure out the video problem on another vendors hardware?
| |
| 23:41 | send an email to support@DisklessWorkstations.com
| |
| 23:41 | mistik1 has quit IRC | |
| 23:41 | <toddobryan_> OK. Will do.
| |
| 23:42 | Does anyone know how good Student Control Panel (or whatever it's now called) is going to be in Feisty?
| |
| 23:43 | <jammcq> I've heard that TCM (Thin Client Manager) is looking really good now
| |
| 23:43 | <sbalneav> What is it you'd like to know?
| |
| 23:43 | <toddobryan_> Well, the Edgy version is not super feature rich.
| |
| 23:44 | I'm looking forward to being able to monitor screens and such.
| |
| 23:45 | <sbalneav> There'sIt has a screen viewer in it
| |
| 23:46 | <toddobryan_> Do you know if you can monitor multiple servers?
| |
| 23:47 | We got two so that if one went out I wouldn't have to quit teaching while we got it fixed.
| |
| 23:48 | In a way, it's lucky we did. Turns out that teaching programming uses way more memory and CPU than I realized.
| |
| 23:49 | <sbalneav> Each TCM only monitors the stations attached to that server, I don't think it has support for monitoring a different server.
| |
| 23:49 | joebob777as7 has joined #ltsp | |
| 23:50 | <toddobryan_> I guess I'd have to open a VNC session to look at the other one.
| |
| 23:50 | <joebob777as7> hey I've got a quick question. I'm running a 64bit server and trying to connect a 32 bit client and I'm getting "your cpu does not support long mode"
| |
| 23:50 | can someone help real quick?
| |
| 23:53 | anyone?
| |
| 23:53 | <jammcq> what distro?
| |
| 23:53 | and what version of ltsp ?
| |
| 23:53 | <joebob777as7> newest ltsp and ubuntu
| |
| 23:53 | <jammcq> sounds like you've also got a 64-bit ltsp chroot
| |
| 23:54 | you'll need to build the 32-bit version. Ogra's the guy who can help, but he doesn't seem to be around right now
| |
| 23:54 | <joebob777as7> yes because it is a 64bit processor... do you know how to get a 32-bit ltsp chroot? and how my clients decipher which to choose?
| |
| 23:58 | <jammcq> it's based on the 'filename' and 'root-path' entries in dhcpd.conf, and sorry, I don't know how to build a 32-bit environment on a 64-bit system
| |
| 23:59 | <joebob777as7> ok thanks i'll wait for ogra
| |