IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 29 December 2007 (all times are UTC)
| 01:06 | cyberorg has quit IRC | |
| 01:09 | zodman has joined #ltsp | |
| 01:14 | cyberorg has joined #ltsp | |
| 01:46 | Q-FUNK has joined #ltsp | |
| 02:04 | chupa has joined #ltsp | |
| 02:06 | chup has quit IRC | |
| 03:07 | zodman has quit IRC | |
| 03:08 | Q-FUNK has quit IRC | |
| 03:21 | Q-FUNK has joined #ltsp | |
| 03:35 | Q-FUNK has quit IRC | |
| 03:47 | <DawnLight> what happens to clients upon a server's reboot?
| |
| 03:50 | <johnny> the clients hang
| |
| 03:50 | but they might be recoverable
| |
| 03:50 | i haven't been at my install when that's happened tho
| |
| 04:03 | chup has joined #ltsp | |
| 04:03 | chupa has quit IRC | |
| 04:42 | K_O-Gnom has joined #ltsp | |
| 05:20 | mcfloppy_ has quit IRC | |
| 05:20 | mcfloppy_ has joined #ltsp | |
| 05:38 | mcfloppy_ has quit IRC | |
| 05:48 | mikkel has joined #ltsp | |
| 05:48 | <ogra_cmpc> stgraber: you need the -nolzma option in ltsp-update-image (before -e)
| |
| 05:48 | stgraber: mksquashfs already defaults to lzma but the kernel cant decompress that yet
| |
| 05:50 | F-GT has quit IRC | |
| 05:56 | Q-FUNK has joined #ltsp | |
| 06:07 | cyberorg_ has joined #ltsp | |
| 06:07 | cyberorg has quit IRC | |
| 06:08 | cyberorg_ is now known as cyberorg | |
| 06:13 | <stgraber> ogra_cmpc: hmm, ok
| |
| 06:14 | F-GT has joined #ltsp | |
| 06:43 | Q-FUNK has quit IRC | |
| 07:31 | mcfloppy_ has joined #ltsp | |
| 07:31 | K_O-Gnom has quit IRC | |
| 07:32 | markvandenborre has joined #ltsp | |
| 07:32 | <markvandenborre> how feasible is the following scenario:
| |
| 07:33 | an ltsp server without a sound card
| |
| 07:33 | and clients with one
| |
| 07:33 | and still getting sound playback on the clients?
| |
| 07:33 | (no bidirectional sound needed)
| |
| 07:34 | <mcfloppy_> hmmm
| |
| 07:34 | my thin clients dont get the right xorg conf
| |
| 07:35 | XF86CONFIG_FILE = /etc/X11/xorg.conf.dual
| |
| 07:35 | X_CONF = /etc/X11/xorg.conf.dual
| |
| 07:35 | <ogra_cmpc> markvandenborre: works fine
| |
| 07:35 | <mcfloppy_> both versions dont work...... i've LTSP5
| |
| 07:35 | <markvandenborre> ogra_cmpc, great
| |
| 07:35 | <ogra_cmpc> mcfloppy_: and /opt/ltsp/i386/etc/X11/xorg.conf.dual exists ?
| |
| 07:36 | (and is in the image if you use gutsy or sid)
| |
| 07:36 | <mcfloppy_> buerots:~# ls /opt/ltsp/i386/etc/X11/xorg.conf.dual
| |
| 07:36 | . /opt/ltsp/i386/etc/X11/xorg.conf.dual
| |
| 07:37 | The client's image is a 4.0 Debian
| |
| 07:37 | <ogra_cmpc> using nbd or nfs ?
| |
| 07:37 | <mcfloppy_> nfs
| |
| 07:37 | <ogra_cmpc> then it should be used ,,,
| |
| 07:38 | <mcfloppy_> no, the client prints "using xorg.conf"
| |
| 07:38 | <ogra_cmpc> where?
| |
| 07:39 | * ogra_cmpc isnt aware of such a string in teh bootscripts | |
| 07:39 | <mcfloppy_> moment.. i reboot
| |
| 07:40 | hmmm..... the text comes when x start
| |
| 07:40 | <ogra_cmpc> weird
| |
| 07:40 | <mcfloppy_> hmm
| |
| 07:40 | <ogra_cmpc> must be a change vagrant made in teh debian port
| |
| 07:40 | <mcfloppy_> x dont start...
| |
| 07:41 | <ogra_cmpc> originally nothing prints such a sentence
| |
| 07:41 | <mcfloppy_> cannot create regular file '/etc/X11/xorg.conf': Read-only file system
| |
| 07:41 | <ogra_cmpc> ah, thats something different :)
| |
| 07:41 | <mcfloppy_> hehe yes.....
| |
| 07:41 | but Xorg prints also the config location
| |
| 07:41 | <ogra_cmpc> seems like the nfs biond mounts dont work
| |
| 07:42 | *bind
| |
| 07:42 | <mcfloppy_> hmm
| |
| 07:42 | <ogra_cmpc> files that need to be rw are bind mounted into a tmpfs in teh old implementation
| |
| 07:42 | (since nfs doesnt like unionfs that was necessary)
| |
| 07:43 | <mcfloppy_> okay.....
| |
| 07:43 | ./var/opt/ltsp 192.168.0.0/255.255.255.0(rw,no_root_squash,async)
| |
| 07:43 | ./var/opt/ltsp/swapfiles 192.168.0.0/255.255.255.0(rw,no_root_squash,async)
| |
| 07:43 | <ogra_cmpc> i wonder if vagrant did something wrong porting that over
| |
| 07:43 | <mcfloppy_> ./var/opt/ltsp/swapfiles 192.168.0.0/255.255.255.0(rw,no_root_squash,async)
| |
| 07:43 | is in my export
| |
| 07:43 | s
| |
| 07:43 | ./opt/ltsp 192.168.0.0/255.255.255.0(rw,no_root_squash,async)
| |
| 07:43 | so...
| |
| 07:43 | ;)
| |
| 07:43 | 3 rows
| |
| 07:44 | <ogra_cmpc> th ebind mounting is done after nfs is mo8unted
| |
| 07:44 | the exports etc are not related here
| |
| 07:44 | it looks to me like its expecting nbd but uses nfs
| |
| 07:45 | is that backported from sid or plain etch packages ?
| |
| 07:45 | <mcfloppy_> hmmm good question
| |
| 07:45 | i dont know
| |
| 07:45 | <ogra_cmpc> well, you should know which packages you installed :)
| |
| 07:45 | <mcfloppy_> i think i ve taken the package from ltsp.org ;)
| |
| 07:45 | <ogra_cmpc> dpkg -l ltsp-server
| |
| 07:46 | whats the version
| |
| 07:46 | <mcfloppy_> 5.0
| |
| 07:46 | <ogra_cmpc> 5.0 ??
| |
| 07:46 | there is no such package
| |
| 07:46 | must be 5.0.something
| |
| 07:46 | <mcfloppy_> hmmm
| |
| 07:47 | http://paste.debian.net/45764
| |
| 07:49 | <ogra_cmpc> so its 5.0.31... :)
| |
| 07:49 | ~0.etch.1 means its a backport
| |
| 07:50 | <mcfloppy_> should i switch to NBD?
| |
| 07:50 | <ogra_cmpc> no
| |
| 07:50 | .31 didnt have nbd support iirc and etch is missing a lot kernel stuff to make it even work
| |
| 07:51 | .31 is the version ubuntu shipped in feisty (7.04)
| |
| 07:51 | nbd cam with gutsy
| |
| 07:51 | *came
| |
| 07:52 | check the values in /etc/default/ltsp-client-core or so in the chroot
| |
| 07:52 | it should set rw_dirs and files
| |
| 07:53 | actually /opt/ltsp/i386/etc/default/ltsp-client-setup
| |
| 07:53 | <mcfloppy_> yes
| |
| 07:53 | ive found
| |
| 07:53 | rw_dirs="/var/cache/man /var/lib/xkb /var/lock /var/run /var/log /var/spool /var/tmp /tmp /var/lib/discover /etc/hotplug/.run /var/lib/nfs /et$
| |
| 07:53 | <ogra_cmpc> right
| |
| 07:53 | check that bindfiles gets xorg.conf assigned
| |
| 07:53 | <mcfloppy_> bindfiles="/etc/network/interfaces /etc/hostname /etc/hosts /etc/nbd-client /etc/syslog.conf /etc/fstab /etc/resolv.conf"
| |
| 07:53 | :(
| |
| 07:54 | <ogra_cmpc> look at the bottom :)
| |
| 07:54 | the last if condition
| |
| 07:54 | it either sets copy_dirs="$copy_dirs /etc/X11 /var/lib/x11" or bindfiles="$bindfiles /etc/X11/xorg.conf"
| |
| 07:55 | unless vagrant dropped that (which i dont see a reason for)
| |
| 07:55 | <mcfloppy_> | |
| 07:55 | it was comented?!?
| |
| 07:55 | hmm i try without comments
| |
| 07:55 | <ogra_cmpc> what is root_write_method set to ?
| |
| 07:55 | <mcfloppy_> root_write_method="bind_mounts"
| |
| 07:55 | root_write_method="bind_mounts"
| |
| 07:56 | hmm now it works
| |
| 07:56 | thank you
| |
| 07:56 | <ogra_cmpc> heh
| |
| 08:56 | markvandenborre has quit IRC | |
| 09:04 | K_O-Gnom has joined #ltsp | |
| 09:20 | K_O-Gnom has quit IRC | |
| 09:31 | DawnLight has quit IRC | |
| 09:31 | DawnLight has joined #ltsp | |
| 09:35 | ogra has joined #ltsp | |
| 09:48 | zodman has joined #ltsp | |
| 09:49 | Egyptian[Home] has quit IRC | |
| 09:49 | Egyptian[Home1 has quit IRC | |
| 09:50 | Egyptian[Home] has joined #ltsp | |
| 09:51 | Egyptian[Home1 has joined #ltsp | |
| 10:00 | zodman has quit IRC | |
| 10:48 | outofrange has joined #ltsp | |
| 11:09 | Top|OLPC has joined #ltsp | |
| 11:15 | ltspbot has joined #ltsp | |
| 11:24 | zodman has joined #ltsp | |
| 11:40 | markvandenborre has joined #ltsp | |
| 11:40 | <markvandenborre> hi, I'm trying to optimise performance of flash video on my edubuntu 7.10 setup
| |
| 11:40 | video is working fine in general
| |
| 11:41 | only flash video using the flash plugin is really choppy
| |
| 11:41 | <jammcq> there's a DIRECT_X setting you can add to lts.conf
| |
| 11:41 | <markvandenborre> if I download the video and play it using totem, everything works very smoothly
| |
| 11:41 | <jammcq> unfortunately, I don't know the details about exactly how to specify it
| |
| 11:41 | <markvandenborre> jammcq: me googling about this
| |
| 11:41 | <jammcq> it'll make an enormous difference in X performance
| |
| 11:42 | <markvandenborre> google says nothing about it
| |
| 11:42 | what's the principle behind it?
| |
| 11:42 | (may help me)
| |
| 11:42 | may help me find out more
| |
| 11:42 | <jammcq> normally, LTSP-5 tunnels all X traffic through SSH
| |
| 11:43 | with the Directx option enabled, it'll do a secure login through ssh, but then switch back to just X over the wire, not tunneled through ssh
| |
| 11:43 | <markvandenborre> ah, right
| |
| 11:43 | that one, ok, should be enough for me to find my way around
| |
| 11:44 | jammcq: still a bit afraid though that this will prove not to be the bottleneck
| |
| 11:44 | seeing that the same video in full screen mode
| |
| 11:44 | plays back very smoothly
| |
| 11:44 | <jammcq> I almost guarantee you'll be happy
| |
| 11:44 | <markvandenborre> if I download it...
| |
| 11:44 | <jammcq> full-screen mode on a thin client?
| |
| 11:44 | <markvandenborre> yeah
| |
| 11:45 | youtube video, so not too high res
| |
| 11:45 | but it's a really old pII 350
| |
| 11:45 | with a fairly well supported old ati video card
| |
| 11:46 | now since I have like 100 of these machines laying around, it's worth the effort trying to optimise
| |
| 11:46 | thx a lot for the hint!
| |
| 11:47 | <jammcq> how do you download youtube videos?
| |
| 11:47 | i've only ever watched them as a stream
| |
| 11:48 | <chup> there is a firefox plugin ive been told. all I know though.
| |
| 11:49 | <stgraber> "Fast Video Download" it's called
| |
| 11:50 | <markvandenborre> jammcq: or there is keepvid.com
| |
| 11:50 | stuff like that
| |
| 11:51 | but it's fairly trivial to create a script to do that
| |
| 11:51 | <outofrange> On ltsp5 XDM isn't running on the server, right? That is, I can't take a normal workstation and get a remote login on the ltsp server?
| |
| 11:51 | <markvandenborre> I have a hardcore freebsd friend who uses lynx as his day-to-day browser
| |
| 11:51 | he has made some stuff to handle this in lynx
| |
| 11:51 | <jammcq> outofrange: correct
| |
| 11:51 | you'd have to configure GDM to allow remote connections
| |
| 11:51 | <markvandenborre> jammcq: have been looking at /opt/ltsp/i386/usr/share/doc/ltsp-client-core/examples
| |
| 11:52 | I don't directly see any mention of this DIRECT_X stuff there
| |
| 11:52 | <jammcq> yeah, I'm searching myself
| |
| 11:52 | <stgraber> markvandenborre: let me boot my hardy thin client I know I have that somewhere on it :)
| |
| 11:53 | <outofrange> You mean LDM_DIRECTX?
| |
| 11:53 | <markvandenborre> outofrange: just found http://www.mail-archive.com/ltsp-discuss@lists.sourceforge.net/msg31592.html
| |
| 11:53 | it mentions that too
| |
| 11:53 | stgraber: would really appreciate
| |
| 11:54 | <jammcq> http://www.mail-archive.com/edubuntu-users@lists.ubuntu.com/msg02663.html
| |
| 11:54 | that has it
| |
| 11:54 | LDM_DIRECTX=True
| |
| 11:55 | in your lts.conf file
| |
| 11:55 | <stgraber> LDM_DIRECTX=True in [Default] should do it
| |
| 11:55 | that's what I have here at least
| |
| 11:55 | <jammcq> and you don't need the patch that is mentioned in the email, cuz it's all been rolled into the upstream ldm
| |
| 11:55 | <outofrange> Hum. I just noticed I have LDM_DIRECTX = True only on some clients.
| |
| 11:56 | <markvandenborre> great, thx jammcq, stgraber
| |
| 11:56 | and outofrange of course
| |
| 11:56 | <jammcq> ok, gotta run. see ya'll later
| |
| 11:56 | * stgraber is checking on youtube (it was laggy without) | |
| 11:57 | <outofrange> Sure wish I could get volume control working.
| |
| 11:58 | <markvandenborre> will be back in a minute
| |
| 11:58 | (working from a thin client right now)
| |
| 11:58 | markvandenborre has left #ltsp | |
| 12:05 | bobby_C has joined #ltsp | |
| 12:07 | joebaker has quit IRC | |
| 12:08 | joebaker has joined #ltsp | |
| 12:09 | markvandenborre has joined #ltsp | |
| 12:10 | <markvandenborre> hm, flash video is still quite choppy
| |
| 12:10 | it's marginally better
| |
| 12:11 | I must confess that I am running this libflashsupport package, to get sound on my local machine
| |
| 12:11 | err, on my client...
| |
| 12:11 | <jammcq> good think I said "almost guarantee"
| |
| 12:11 | <markvandenborre> :)
| |
| 12:11 | that's about the only non-standard package I have installed
| |
| 12:12 | is there any other way I could try and troubleshoot this?
| |
| 12:12 | <outofrange> Is it just youtube or all flash?
| |
| 12:12 | <markvandenborre> seems to me that the problem is with this damn binary flash stuff itself
| |
| 12:13 | outofrange: I'll double check, but I'm afraid it's all of it
| |
| 12:13 | could it be because I have this latest version?
| |
| 12:13 | <stgraber> the problem with flash is that it uses X11 for the output so it'll always be laggy till we have local apps working
| |
| 12:13 | I'm playing a video using mplayer and the XV output here and it's just perfect
| |
| 12:13 | (on a 1440x900 monitor :))
| |
| 12:13 | <markvandenborre> ah, maybe it's because all the rest is accelerated
| |
| 12:14 | I see
| |
| 12:14 | how do I get shell access to the client?
| |
| 12:14 | maybe there's more things I can do around there...
| |
| 12:15 | <outofrange> Flash works great on my fast clients, but doesn't on the slower clients. But, on the faster clients (P4 3GHz!) I've even had youtube stall and freeze.
| |
| 12:15 | <markvandenborre> outofrange: this is a really slow type of client
| |
| 12:15 | like, pII 350
| |
| 12:15 | <stgraber> hmm, it's on a 2600+ amd CPU here, but I think the display size may cause problem with non-accelerated video stream ...
| |
| 12:16 | markvandenborre: you can add SCREEN_2=shell
| |
| 12:16 | <outofrange> To get shell access I copied in my password int the chroot shadow file. Then just ctrl-alt-f1 and log in.
| |
| 12:16 | <markvandenborre> but the video is relatively sane in comparison to the cpu
| |
| 12:16 | <stgraber> and you'll have a root shell on tty2
| |
| 12:16 | <markvandenborre> stgraber: and are there faster ways even to get to the x configuration on the client
| |
| 12:17 | I've already pushed colour depth to 16 from 32
| |
| 12:17 | <stgraber> no, this is the fastest way (a line to add to lts.conf)
| |
| 12:17 | <markvandenborre> stgraber: I mean, diagnosing what is currently autogenerated
| |
| 12:17 | <outofrange> I've been thinking of adding sshd to the chroot.
| |
| 12:17 | stgraber: are you using gnome desktop?
| |
| 12:18 | <stgraber> markvandenborre: except if you can get the settings you want using xrandr I doubt there is a faster way than getting root access
| |
| 12:18 | outofrange: yep
| |
| 12:18 | <outofrange> Does your gnome panel volume applet work?
| |
| 12:18 | <markvandenborre> is lowering refresh rate any use for X11/flash performance
| |
| 12:18 | ?
| |
| 12:19 | <stgraber> outofrange: yes it does, but I'm running Hardy maybe it was fixed after gutsy was released
| |
| 12:21 | <outofrange> Ah, I see. I just have not been able to figure out why mine doesn't work. If you right-click and select preferences does it show Pulse as a device? Mine only shows ALSA and OSS.
| |
| 12:22 | <stgraber> IIRC it's because you have access to the server soundcard (read/write access to /dev/dsp) (my server doesn't have a sound card) and because you are missing a pulse library
| |
| 12:23 | what's the output of : dpkg -l | grep pulse ?
| |
| 12:23 | <outofrange> On client or server?
| |
| 12:23 | libpulse-mainloop-glib0 libpulse0 on server
| |
| 12:24 | <stgraber> try installing : gstreamer0.10-pulseaudio
| |
| 12:24 | <outofrange> On the server?
| |
| 12:24 | <stgraber> yes
| |
| 12:24 | <outofrange> Ok
| |
| 12:25 | That package doesn't exist, but this does: libgstreamer-plugins-pulse0.10-0
| |
| 12:25 | <stgraber> hmm, ok install this one then :)
| |
| 12:25 | looks like they renamed some packages with gutsy
| |
| 12:27 | <outofrange> I assume that library is used by the current user -- i.e. I don't need to reboot the server (just re-login with the client)?
| |
| 12:27 | <stgraber> indeed
| |
| 12:30 | <outofrange> Yes!
| |
| 12:30 | stgraber: thanks so much. I've been trying to get that working for a while.
| |
| 12:31 | <stgraber> np
| |
| 12:31 | <outofrange> But, I had to configure "Sound Preferences" to work. Is there any trick to make "Autodetect" pick Pulse?
| |
| 12:31 | Kind of a pain to configure all users
| |
| 12:32 | <stgraber> it should take pulse if it can't access the real sound device
| |
| 12:34 | <outofrange> X11 connection rejected because of wrong authentication. I get this when some users log in and then their volume applet doesn't show up.
| |
| 12:34 | Maybe they are not in the right group?
| |
| 12:35 | <stgraber> audio group maybe
| |
| 12:36 | or it's an ISA sound card undetected by the thin client
| |
| 12:37 | <outofrange> No, the user is in audio. The only difference is vidieo lpadmin admin netdev powerdev.
| |
| 12:38 | Messages like "wrong authentication" could be more explicit, ;)
| |
| 12:38 | <stgraber> indeed :)
| |
| 12:39 | <outofrange> Is this something I should ask about on the list? And if so, which list? ltsp-discuss or edubuntu?
| |
| 12:44 | markvandenborre has left #ltsp | |
| 12:46 | <outofrange> sure wish ldm provided both login and password on the same screen. I keep typing the wrong username and can't go back.
| |
| 12:47 | <johnny> yeah.. i agree
| |
| 13:00 | <outofrange> Hum, ldm.log also says "Warning: No xauth data; using fake authentication data for X11 forwarding.
| |
| 13:00 | It also says "LTSPROCKS" ;)
| |
| 13:09 | ogra has quit IRC | |
| 13:09 | ogra_cmpc has quit IRC | |
| 13:10 | ogra_cmpc has joined #ltsp | |
| 13:14 | slipttees has joined #ltsp | |
| 13:40 | J45p3r has joined #ltsp | |
| 13:50 | zodman has quit IRC | |
| 13:54 | markvandenborre has joined #ltsp | |
| 13:54 | <markvandenborre> I'm still experimenting with flash and modest thin clients
| |
| 13:55 | please tell me if this is not the appropriate channel to talk about that
| |
| 13:55 | <johnny> it is.. and isn't..
| |
| 13:56 | <jammcq> I'd say it's the perfect place to talk about it
| |
| 13:56 | <johnny> if it's on the topic of ltsp at all, then you're prolly ok
| |
| 13:56 | <markvandenborre> I can imagine that, especially with the fact that it's non-free
| |
| 13:56 | <johnny> thin clients != ltsp
| |
| 13:56 | in all cases
| |
| 13:56 | <markvandenborre> johnny: ltsp, obviously
| |
| 13:56 | <jammcq> if the videos work fine on a full computer, but don't work so well on LTSP terminals, then this is the place
| |
| 13:56 | <johnny> yes.. that too :)
| |
| 13:56 | i've ignored flash on my clients for now
| |
| 13:56 | <markvandenborre> jammcq: I have experienced that flash videos like http://www.schooltv.nl/beeldbank/clip/20030701_kerstbal01
| |
| 13:56 | <jammcq> certainly, johnny makes a good point. If you aren't using LTSP as your thin client OS, then this wouldn't be the irght place
| |
| 13:57 | <outofrange> markvandenborre: do you have modern hardware to test with as well?
| |
| 13:57 | As a client, that is.
| |
| 13:57 | <markvandenborre> outofrange: I could use my laptop as a client I suppose
| |
| 13:57 | but I discovered that movies using http://www.jeroenwijering.com/?item=JW_FLV_Player
| |
| 13:57 | <outofrange> I'm amazed how well everything works on my new hardware and how it doesn't on my really old Imacs.
| |
| 13:57 | <markvandenborre> perform much better
| |
| 13:58 | <johnny> i retired the imac public terms
| |
| 13:58 | cuz i didn't want to maintain two chroots
| |
| 13:58 | for just a few of them
| |
| 13:58 | <outofrange> Yes, that's the plan. We have 70 of them.
| |
| 13:59 | <markvandenborre> back in a sec
| |
| 13:59 | markvandenborre has left #ltsp | |
| 13:59 | <jammcq> wow, that's a cool video
| |
| 13:59 | <outofrange> How do I change the language? ;)
| |
| 14:06 | Heidi has joined #ltsp | |
| 14:06 | Heidi has left #ltsp | |
| 14:12 | chup is now known as chupacabra | |
| 14:22 | <DawnLight> is there support for running processes on the clients?
| |
| 14:27 | oh, i see http://wiki.ltsp.org/twiki/bin/view/Ltsp/LocalApps
| |
| 14:45 | Egyptian[Home] has quit IRC | |
| 14:50 | DawnLight has quit IRC | |
| 15:03 | DawnLight has joined #ltsp | |
| 15:08 | <outofrange> Hum, so it's my .Xauthority that is causing problems. Anyone know how it's used in the LTSP environment?
| |
| 15:14 | DawnLight has quit IRC | |
| 15:40 | vagrantc has joined #ltsp | |
| 15:46 | bobby_C has quit IRC | |
| 15:47 | outofrange has left #ltsp | |
| 15:49 | mikkel has quit IRC | |
| 15:53 | <otavio> vagrantc: hi
| |
| 15:54 | forgot them
| |
| 15:54 | vagrantc: resend them to me please
| |
| 15:54 | <vagrantc> otavio: oh.
| |
| 15:54 | otavio: ok. i've got a small update for ltsp anyways ...
| |
| 15:55 | otavio: how have you been?
| |
| 15:59 | otavio: ldm sent. will work on ltsp...
| |
| 15:59 | <johnny> hmm.. haven't seen sbalnaev and ogra in awhile
| |
| 15:59 | they gonna come back soon?
| |
| 15:59 | <vagrantc> i saw ogra_cmpc a day or two ago
| |
| 16:01 | <jammcq> Scott bought a cabin in canada, and I think he's taken his family there for the holidays
| |
| 16:01 | no internet access there
| |
| 16:01 | <johnny> fun
| |
| 16:02 | Q-FUNK has joined #ltsp | |
| 16:02 | markvandenborre has joined #ltsp | |
| 16:03 | <vagrantc> otavio: should XS-X-Vcs-Bzr point to upstream, or the debian packaging branch?
| |
| 16:04 | <markvandenborre> what is the best way to test xorg performance on a client?
| |
| 16:04 | I've found a few nice options that might just give me the 10-20% extra I need to get *$!# flash working more fluently
| |
| 16:05 | but how do I test if plain x11 performance has actually improved?
| |
| 16:06 | <jammcq> x11perf
| |
| 16:06 | Q-FUNK has quit IRC | |
| 16:06 | <markvandenborre> jammcq, how should I have found that?
| |
| 16:07 | chupacabra has quit IRC | |
| 16:07 | <markvandenborre> hm, never mind that question...
| |
| 16:07 | <jammcq> heh
| |
| 16:07 | all you gotta do is ask
| |
| 16:07 | chupacabra has joined #ltsp | |
| 16:07 | <markvandenborre> it might have sounded a bit negative, but I was just being curious
| |
| 16:08 | but if I'm in a session, that'll give me remote performance, right?
| |
| 16:08 | I can run glxinfo and glxgears and stuff too, with like 3 frames per minute and these guys saying "140 fps"
| |
| 16:09 | <slipttees> see i alter
| |
| 16:09 | <markvandenborre> ah, I should have read the man page first
| |
| 16:09 | <slipttees> happy new year for all...!!
| |
| 16:09 | slipttees has quit IRC | |
| 16:09 | <vagrantc> earge to start the new year, i guess
| |
| 16:10 | eagre
| |
| 16:19 | <markvandenborre> x11perf -display :0 -all works perfectly on my desktop machine
| |
| 16:19 | but I'm a bit lost running it o nmy thin clinet
| |
| 16:24 | and just a sudo x11perf -display 10.0.0.250:0 -all obviously is not going to work
| |
| 16:26 | Egyptian[Home1 has quit IRC | |
| 16:27 | <jammcq> are you logged into the thin client?
| |
| 16:27 | <markvandenborre> yes, I have root on the client
| |
| 16:27 | (tty1)
| |
| 16:27 | <jammcq> where do you want x11perf to run? on the client, displayed to the client? or on the server, displayted to the client?
| |
| 16:28 | <markvandenborre> I want to measure the performance of the client graphics hardware, obviously
| |
| 16:28 | <jammcq> why not just do a gui login from the client to the server, then run: x11perf --all
| |
| 16:28 | that will do that
| |
| 16:29 | <markvandenborre> ah, I was a bit confused here, thought I would be measuring server graphics hardware performance...
| |
| 16:29 | <jammcq> no, it'll measure the display, whatever it happens to be set to.
| |
| 16:30 | if you log in via the thin client, your display will be set in $DISPLAY
| |
| 16:30 | <markvandenborre> I see...
| |
| 16:30 | <jammcq> in fact, go ahead and log in that way, and echo $DISPLAY, see what it's set to
| |
| 16:30 | <markvandenborre> 10.0.0.250:6.0
| |
| 16:31 | (client ip:6.0)
| |
| 16:31 | <jammcq> hmm, ok
| |
| 16:31 | i'm wondering why it's 6.0
| |
| 16:31 | <markvandenborre> me too
| |
| 16:31 | <jammcq> shouldn't be a problem
| |
| 16:31 | maybe it's because it is running on the 7th screen
| |
| 16:31 | <markvandenborre> it's performing fairly well for it's age, and very stable too
| |
| 16:31 | yeah, probably
| |
| 16:31 | it's just that I need to squeeze out that last drip
| |
| 16:32 | <jammcq> but that tells us it's definately using the LDM_DIRECTX option
| |
| 16:32 | what video chipset is it?
| |
| 16:32 | <markvandenborre> yup, that's the way I set it, thanks to your hints
| |
| 16:32 | ati 3d rage pro agp 1x/2x
| |
| 16:32 | 8 Mb of ram
| |
| 16:33 | <jammcq> what video driver is being used?
| |
| 16:33 | <markvandenborre> I found out that I could probably squeeze out a little more by explicitly ordering it to use agp 2x mode
| |
| 16:33 | ati
| |
| 16:33 | not vesa of course ;)
| |
| 16:33 | <jammcq> good
| |
| 16:33 | <markvandenborre> the problem is really with that bastard binary only flash plugin
| |
| 16:33 | incredibly inefficient
| |
| 16:33 | all other online video is working really smoothly
| |
| 16:34 | and even most flash is working quite ok
| |
| 16:34 | just youtube, and guess what people want?
| |
| 16:34 | ... youtube
| |
| 16:36 | hope gnash will solve that in the not too distant future
| |
| 16:37 | <jammcq> I'm not convinced that gnash will be any better performance than the macromedia player
| |
| 16:38 | chup has joined #ltsp | |
| 16:39 | <markvandenborre> you're right on that one
| |
| 16:39 | but it will probably be much more network efficient
| |
| 16:39 | right?
| |
| 16:39 | <jammcq> why?
| |
| 16:40 | their first order of business is to function
| |
| 16:40 | chupacabra has quit IRC | |
| 16:40 | <jammcq> they'll be working on full compatibility for a very long time. performance improvements will happen after it works
| |
| 16:40 | <markvandenborre> of course, and they try to be portable between archs, which will always give them a kind of performance penalty
| |
| 16:41 | <jammcq> the goal of that project is to provide a "FREE" flash player, not a "FASTER" flash player
| |
| 16:41 | <markvandenborre> yeah, I should probably stop talking out of my ass
| |
| 16:42 | it's just me really really wanting to give these kids decent video performance, even with this youtube flash stuff
| |
| 16:43 | <jammcq> what kind of network card in that machine?
| |
| 16:43 | <markvandenborre> intel 100 Mbit in the client, Gbit in the server
| |
| 16:43 | <jammcq> maybe it's a network bandwidth issue
| |
| 16:43 | <markvandenborre> full screen video gets through quite well...
| |
| 16:43 | correction:
| |
| 16:44 | a scaled up 1/4 screen video gets through quite well
| |
| 16:44 | except for flash video
| |
| 16:44 | when played through the flash player
| |
| 16:45 | works perfectly smooth with totem and gstreamer ffmpeg plugin
| |
| 16:46 | <vagrantc> otavio: i'm wondering exactly what to say with the removals to the Uploaders field ... ?
| |
| 16:46 | <markvandenborre> jammcq, would it be useful to use iptraf and stuff?
| |
| 16:46 | Alexandre_ has joined #ltsp | |
| 16:46 | <vagrantc> otavio: it's easy for mdz and stratus, as they said they didn't have time to be involved ...
| |
| 16:46 | otavio: but mvo ... maybe never even knew they were on the team ...
| |
| 16:47 | <markvandenborre> I'm asking a lot of questions here, will try to document what I learnt in the wiki of course
| |
| 16:47 | for as far as it isn't there already
| |
| 16:47 | Alexandre_ has quit IRC | |
| 16:53 | outofrange has joined #ltsp | |
| 16:55 | outofrange has left #ltsp | |
| 17:10 | <otavio> vagrantc: I work with mvo on apt
| |
| 17:10 | vagrantc: he's nice to play with
| |
| 17:10 | vagrantc: and I think it's ok to remove him.
| |
| 17:14 | vagrantc has quit IRC | |
| 17:16 | OvaKill has joined #ltsp | |
| 17:19 | F-GT has quit IRC | |
| 17:24 | vagrantc has joined #ltsp | |
| 17:28 | OvaKill has quit IRC | |
| 17:28 | F-GT has joined #ltsp | |
| 17:31 | <vagrantc> otavio: sent ltsp upload too.
| |
| 17:31 | <otavio> vagrantc: got them
| |
| 17:31 | vagrantc: will do them tomorrow ... and sorry by being late
| |
| 17:31 | <vagrantc> otavio: thanks :)
| |
| 17:31 | otavio: you're just trying to coerce me to finish NM :P
| |
| 17:32 | <otavio> ahahh sorry
| |
| 17:32 | wasn't my intention
| |
| 17:33 | i've been busy and ended forgetting it
| |
| 17:34 | <vagrantc> otavio: i figured ... not seeing you on irc for a few days ... figured you'd have lots of catch-up work
| |
| 17:36 | <markvandenborre> has anyone compared flash video performance on thin clients between the latest versions?
| |
| 17:36 | google seems to say that the latest version is worse than previous ones
| |
| 17:37 | http://www.adobe.com/cfusion/webforums/forum/messageview.cfm?forumid=72&catid=616&threadid=1317513&enterthread=y
| |
| 17:37 | (see at the bottom)
| |
| 17:39 | chupa has joined #ltsp | |
| 17:40 | vagrantc has quit IRC | |
| 17:40 | chup has quit IRC | |
| 17:44 | vagrantc has joined #ltsp | |
| 17:52 | otavio has quit IRC | |
| 18:15 | vagrantc_ has joined #ltsp | |
| 18:15 | <markvandenborre> hey all, I just found something that might be used as a stopgap solution for low performance
| |
| 18:15 | youtube clients
| |
| 18:15 | tribler.org
| |
| 18:16 | free software, from a dutch university
| |
| 18:16 | does seem to be able to use xv, contrary to adobe flash that only uses x11
| |
| 18:20 | hm, not too impressive at second sight
| |
| 18:20 | markvandenborre has left #ltsp | |
| 18:23 | sudonix has joined #ltsp | |
| 18:27 | vagrantc has quit IRC | |
| 18:30 | <sudonix> hello, can anyone give some advice with debugging ltspfs problems?
| |
| 18:30 | I have followed https://wiki.ubuntu.com/DebugLocalDev, got stuck on step 4 (running gutsy)
| |
| 18:31 | the workaround/fix mentioned in https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/160420 did not help
| |
| 18:35 | vagrantc has joined #ltsp | |
| 18:39 | sudonix has quit IRC | |
| 18:41 | sudonix has joined #ltsp | |
| 18:43 | vagrantc_ has quit IRC | |
| 18:52 | joebaker has quit IRC | |
| 18:52 | joebaker has joined #ltsp | |
| 19:01 | sudonix has quit IRC | |
| 19:50 | F-GT has quit IRC | |
| 19:51 | chup has joined #ltsp | |
| 19:52 | chupa has quit IRC | |
| 20:37 | markvandenborre has joined #ltsp | |
| 20:37 | <markvandenborre> any idea what a normal score is for
| |
| 20:37 | x11perf -copywinpix500
| |
| 20:37 | on a thin client?
| |
| 20:38 | I'm getting 6.9 per second, and something tells me that's just not righ
| |
| 20:38 | t
| |
| 20:40 | <vagrantc> try it on a desktop system?
| |
| 20:43 | <markvandenborre> I tried it on my laptop
| |
| 20:43 | got 96.4
| |
| 20:43 | F-GT has joined #ltsp | |
| 20:43 | <markvandenborre> but I'm curious if this is _really_ really bad or not
| |
| 20:43 | because it sure does look like that
| |
| 20:44 | so if you could run that on something not too beefy, I'd be grateful
| |
| 20:47 | mistik1_ has joined #ltsp | |
| 20:47 | <markvandenborre> oh man, this is strange
| |
| 20:47 | vesa is working better than ati
| |
| 20:47 | !
| |
| 20:48 | mistik1 has quit IRC | |
| 20:48 | mistik1_ is now known as mistik1 | |
| 20:49 | <vagrantc> markvandenborre: what numbers do you want?
| |
| 20:50 | <markvandenborre> vagrantc, I've solved the issue after hours and hours
| |
| 20:50 | thx for offering to help, but it's done
| |
| 20:50 | really happy
| |
| 20:50 | <vagrantc> markvandenborre: ok.
| |
| 20:50 | <markvandenborre> I just wonder if I should file a bug report
| |
| 20:51 | on such old hardware
| |
| 20:51 | or at least put this info up somewhere in the wiki
| |
| 20:51 | <vagrantc> surely you should say something about how you fixed it :P
| |
| 20:52 | <markvandenborre> vagrantc, the problem was that on my hardware
| |
| 20:52 | (old recycled compaq deskpro machine with ati 3d rage pro agp 1x 2x)
| |
| 20:52 | the autodetected ati driver performed _really_ badly on certain actions
| |
| 20:53 | that made flash really choppy
| |
| 20:53 | and that made some x11 performance tests go really bad
| |
| 20:53 | (x11perf -copywinpix500)
| |
| 20:54 | that's a test that copies 500x500 squares from window to pixmap
| |
| 20:54 | something similar happens with the flash plugin
| |
| 20:54 | using the vesa driver somehow solved the issue
| |
| 20:55 | performance on this test went up from 6.9/sec to 140/sec
| |
| 20:55 | <vagrantc> well good :)
| |
| 20:56 | gotta run
| |
| 20:56 | vagrantc has quit IRC | |
| 20:56 | <markvandenborre> since this was quite common hardware, it might be interesting for others to know...
| |
| 20:58 | markvandenborre has quit IRC | |
| 21:00 | npman has left #ltsp | |
| 21:09 | joebaker has quit IRC | |
| 21:28 | primeministerp has quit IRC | |
| 22:02 | zodman has joined #ltsp | |
| 22:03 | zodman has quit IRC | |
| 22:08 | Top|OLPC has quit IRC | |
| 22:09 | joebaker has joined #ltsp | |
| 22:25 | open_lad has joined #ltsp | |
| 22:36 | joebaker has quit IRC | |
| 22:37 | joebaker has joined #ltsp | |
| 22:48 | open_lad has quit IRC | |
| 22:48 | open_lad has joined #ltsp | |
| 23:03 | zodman has joined #ltsp | |
| 23:07 | J45p3r has quit IRC | |
| 23:10 | joebaker has quit IRC | |
| 23:11 | zodman has quit IRC | |
| 23:11 | zodman has joined #ltsp | |
| 23:17 | basanta has joined #ltsp | |
| 23:26 | subir has joined #ltsp | |
| 23:29 | basanta has quit IRC | |
| 23:29 | chup has quit IRC | |
| 23:30 | chup has joined #ltsp | |
| 23:31 | joebaker has joined #ltsp | |
| 23:31 | joebaker has left #ltsp | |
| 23:39 | Egyptian[Home] has joined #ltsp | |
| 23:48 | chup has quit IRC | |
| 23:49 | chup has joined #ltsp | |