IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 20 February 2009 (all times are UTC)
| 00:01 | w0lter has joined #ltsp | |
| 00:01 | w0lter has left #ltsp | |
| 00:27 | shrek has joined #ltsp | |
| 00:39 | SDuensin has quit IRC | |
| 00:44 | vagrantc has joined #ltsp | |
| 00:48 | slashdotfx has quit IRC | |
| 00:48 | slashdotfx has joined #ltsp | |
| 00:49 | shrek has quit IRC | |
| 00:53 | alkisg has quit IRC | |
| 00:54 | talntid has quit IRC | |
| 01:09 | Gadi_eeepc has joined #ltsp | |
| 01:09 | Gadi_eeepc1 has quit IRC | |
| 01:18 | Egyptian[Home] has joined #ltsp | |
| 01:19 | vagrantc has quit IRC | |
| 01:21 | Egyptian[Home] has quit IRC | |
| 01:22 | gate_keeper_ has joined #ltsp | |
| 01:25 | Egyptian[Home] has joined #ltsp | |
| 01:33 | Egyptian[Home]1 has joined #ltsp | |
| 01:41 | Gadi_eeepc1 has joined #ltsp | |
| 01:41 | Gadi_eeepc has quit IRC | |
| 01:45 | Egyptian[Home] has quit IRC | |
| 02:00 | sepski has quit IRC | |
| 02:21 | slashdotfx has quit IRC | |
| 02:21 | slashdotfx has joined #ltsp | |
| 02:27 | tjikkun_work has joined #ltsp | |
| 02:52 | alkisg_work has joined #ltsp | |
| 02:54 | ogra has quit IRC | |
| 02:54 | ogra has joined #ltsp | |
| 03:00 | polytan has joined #ltsp | |
| 03:02 | polytan has quit IRC | |
| 03:02 | polytan has joined #ltsp | |
| 03:14 | Remaille has joined #ltsp | |
| 03:36 | synchris_odium is now known as synchris | |
| 03:38 | Egyptian[Home]1 is now known as Egyptian[Home] | |
| 03:56 | Gadi_eeepc has joined #ltsp | |
| 03:56 | Gadi_eeepc1 has quit IRC | |
| 03:57 | hanthana has quit IRC | |
| 04:09 | cascade has joined #ltsp | |
| 04:11 | cascade is now known as CaScAdE^FarAway | |
| 04:12 | alkisg_work has quit IRC | |
| 04:16 | _Shane has joined #ltsp | |
| 04:26 | polytan has quit IRC | |
| 04:42 | icefusion11 has joined #ltsp | |
| 04:42 | <icefusion11> morning
| |
| 04:43 | i have a problem with my ltsp, I make it with debian, but after boot, the login screen don't be showed, just a gray scrren with a X of mouse in center
| |
| 05:02 | pdjbarber has joined #ltsp | |
| 05:10 | polytan has joined #ltsp | |
| 05:11 | pdjbarber has quit IRC | |
| 05:12 | otavio has joined #ltsp | |
| 05:20 | alkisg has joined #ltsp | |
| 05:22 | polytan has quit IRC | |
| 05:23 | polytan has joined #ltsp | |
| 05:24 | polytan has quit IRC | |
| 05:24 | pdjbarber has joined #ltsp | |
| 05:31 | hanthana has joined #ltsp | |
| 05:35 | slashdotfx has quit IRC | |
| 05:37 | Remaille has quit IRC | |
| 05:40 | slashdotfx has joined #ltsp | |
| 05:45 | ilaiho has quit IRC | |
| 05:45 | ilaiho has joined #ltsp | |
| 05:50 | ilaiho has quit IRC | |
| 05:51 | w0lter has joined #ltsp | |
| 05:51 | w0lter has left #ltsp | |
| 05:52 | otavio has quit IRC | |
| 06:05 | icefusion11 has left #ltsp | |
| 06:17 | bobby_C has joined #ltsp | |
| 06:30 | _Shane has left #ltsp | |
| 06:33 | OvaKill has joined #ltsp | |
| 06:33 | F-GT has quit IRC | |
| 06:36 | Gadi has quit IRC | |
| 06:38 | alkisg has quit IRC | |
| 06:39 | vvinet has joined #ltsp | |
| 06:40 | Nubae has quit IRC | |
| 06:43 | Gadi_eeepc1 has joined #ltsp | |
| 06:43 | Gadi_eeepc has quit IRC | |
| 06:46 | OvaKill has quit IRC | |
| 06:46 | F-GT has joined #ltsp | |
| 06:51 | berriop has joined #ltsp | |
| 06:52 | <berriop> how can I change the screen resolution of the thin clients?
| |
| 06:53 | BrunoXLambert has joined #ltsp | |
| 06:53 | pmatulis has joined #ltsp | |
| 07:00 | <berriop> i guess it is on the lts.conf file but I dont know d exact parameter
| |
| 07:00 | <shamino> berriop: you have to add X_MODE_0=<resolution> and CONFIGURE_X=True to /var/lib/tftpboot/ltsp/i386/lts.conf
| |
| 07:02 | Gadi_eeepc1 has left #ltsp | |
| 07:05 | jammcq has quit IRC | |
| 07:11 | <berriop> shamino: thanks very much
| |
| 07:15 | F-GT has quit IRC | |
| 07:16 | ogra has quit IRC | |
| 07:17 | F-GT has joined #ltsp | |
| 07:19 | ogra has joined #ltsp | |
| 07:22 | dmaran has joined #ltsp | |
| 07:26 | vvinet has quit IRC | |
| 07:37 | F-GT has quit IRC | |
| 07:39 | jammcq has joined #ltsp | |
| 07:39 | <jammcq> Gooood morning #ltsp
| |
| 07:44 | <berriop> shamino: it doesn't work, idont know why but the thin clients have the wrong resolution
| |
| 07:44 | it is a fresh LTSP install on Ubuntu 8.04, i didnt change anything
| |
| 07:51 | RobertLaptop has quit IRC | |
| 07:53 | vvinet has joined #ltsp | |
| 07:53 | F-GT has joined #ltsp | |
| 07:58 | <shamino> berriop: did you reboot the thin client?
| |
| 08:00 | BrunoXLambert has quit IRC | |
| 08:15 | Gadi has joined #ltsp | |
| 08:26 | cliebow has joined #ltsp | |
| 08:30 | aledec has quit IRC | |
| 08:33 | RiXtEr has quit IRC | |
| 08:34 | vvinet has quit IRC | |
| 08:43 | <berriop> shamino: yep, I even did a ltsp-update-image
| |
| 08:44 | alkisg has joined #ltsp | |
| 08:44 | <berriop> the lts.conf didtn exist so i created it and added those parameters
| |
| 08:47 | its a bit strange, in one of the clients the resolution is ok, in the other is not, they are using the same hardware and monitor
| |
| 08:48 | aledec has joined #ltsp | |
| 09:02 | <alkisg> If I have two hard disks on the server and want to increase performance (something like RAID 0), what should I use? LVM?
| |
| 09:02 | aledec has quit IRC | |
| 09:03 | aledec has joined #ltsp | |
| 09:07 | bobby_C has quit IRC | |
| 09:07 | bobby_C has joined #ltsp | |
| 09:13 | vvinet has joined #ltsp | |
| 09:43 | <berriop> I have change the resolution of one of my clients using the lts.cfg and it doesnt work
| |
| 09:43 | the other client, which has the same hw and monitor has the resolution set properly, what is going on??
| |
| 09:46 | sepski has joined #ltsp | |
| 09:49 | pdjbarber has quit IRC | |
| 10:00 | gate_keeper_ has quit IRC | |
| 10:08 | zjzn has joined #ltsp | |
| 10:08 | |Paradox| has quit IRC | |
| 10:08 | zjzn is now known as |Paradox| | |
| 10:09 | Ahmuck has joined #ltsp | |
| 10:12 | BrunoXLambert has joined #ltsp | |
| 10:15 | hanthana has quit IRC | |
| 10:20 | vvinet has quit IRC | |
| 10:22 | vvinet has joined #ltsp | |
| 10:24 | johnb has joined #ltsp | |
| 10:26 | staffencasa has joined #ltsp | |
| 10:43 | tjikkun_work has quit IRC | |
| 10:48 | japerry has quit IRC | |
| 10:58 | SDuensin has joined #LTSP | |
| 10:58 | <SDuensin> Greetings at last.
| |
| 11:02 | japerry has joined #ltsp | |
| 11:07 | alkisg has quit IRC | |
| 11:07 | Appiah has quit IRC | |
| 11:07 | alkisg has joined #ltsp | |
| 11:14 | aledec has quit IRC | |
| 11:14 | phantom has joined #ltsp | |
| 11:14 | F-GT has quit IRC | |
| 11:15 | Remaille has joined #ltsp | |
| 11:21 | leio has quit IRC | |
| 11:24 | LaserJock has joined #ltsp | |
| 11:25 | leio has joined #ltsp | |
| 11:29 | Remaille has quit IRC | |
| 11:34 | johnb has quit IRC | |
| 11:36 | <shamino> berriop: did you add [default] tag to the beginning of the lts.conf file?
| |
| 11:36 | hanthana has joined #ltsp | |
| 11:37 | <ogra> jammcq, do you know whats up with scottie ? he didnt show up for a while
| |
| 11:38 | <jammcq> ogra: no worries about scotty. I chatted with him this morning
| |
| 11:38 | <ogra> ah, good
| |
| 11:39 | <jammcq> he's got someone in his office going live with some accounting project. it's got him real busy
| |
| 11:40 | <ogra> yeah, i suspected its work related
| |
| 11:42 | <jammcq> ogra: hey, I see the next UDS is May 25-29. that means it won't interfere with FISL on June 24, so you can come
| |
| 11:42 | <ogra> yeah, i might :)
| |
| 11:42 | * jammcq needs to get that going soon | |
| 11:42 | <ogra> depends how demanding that karmic koala is :)
| |
| 11:43 | <LaserJock> ogra: put in a lot of work now and karmic will reward you ;-)
| |
| 11:44 | <ogra> well, i still have a lot of work to do to make 30M 133MHz ARM crap work :P
| |
| 11:44 | with jaunty
| |
| 11:44 | <jammcq> hey LaserJock
| |
| 11:44 | did you see my comments about sbalneav?
| |
| 11:44 | <LaserJock> jammcq: yep
| |
| 11:45 | <jammcq> he said he'd be around this afternoon
| |
| 11:45 | vagrantc has joined #ltsp | |
| 11:45 | <LaserJock> jammcq: I just wanted to make sure he was doing OK
| |
| 11:45 | <jammcq> it's lunch time here, so I'll see ya'll in a little while
| |
| 11:45 | * ogra needs to head out ... | |
| 11:45 | <jammcq> yeah, scotty disappears for weeks sometimes, but he always comes back
| |
| 11:51 | Appiah has joined #ltsp | |
| 11:59 | <jcastro> I was starting to believe that scotty had become a mythical creature
| |
| 12:06 | vagrantc has quit IRC | |
| 12:06 | <cliebow> Scotty IS a mythical creatue
| |
| 12:06 | creature
| |
| 12:09 | RobertLaptop has joined #ltsp | |
| 12:19 | spectra has joined #ltsp | |
| 12:27 | <staffencasa> can someone help me with using cron and the --kiosk plugin
| |
| 12:28 | It doesn't seem to run
| |
| 12:28 | aledec has joined #ltsp | |
| 12:41 | <Gadi> staffencasa: you installed it in the chroot?
| |
| 12:51 | dmaran has left #ltsp | |
| 12:53 | BrunoXLambert has quit IRC | |
| 12:56 | Lns has joined #ltsp | |
| 13:00 | johnny has left #ltsp | |
| 13:00 | johnny has joined #ltsp | |
| 13:01 | <Lns> hey all
| |
| 13:01 | ogra: ping?
| |
| 13:04 | <staffencasa> Gadi, yeah
| |
| 13:04 | <Lns> Anyone wanna talk to me about the command LDM uses to shutdown? I'm having issues with WakeOnLan after using this command (vs. 'halt' from the chroot shell)
| |
| 13:05 | <Gadi> staffencasa: according to my friend Lns, you have to unlock the root user in the chroot for cron jobs to run
| |
| 13:05 | you might want to give that a try: chroot /opt/ltsp/i386 passwd -u root
| |
| 13:05 | (and reroll the image)
| |
| 13:06 | <Lns> Gadi: i told you not to quote me on that.. =p
| |
| 13:06 | <Gadi> hehe
| |
| 13:06 | :)
| |
| 13:06 | <staffencasa> I'll give that a shot, thanks
| |
| 13:06 | <Gadi> staffencasa: and by Lns I mean my Longterm Nice Social friend
| |
| 13:06 | who is in no way related to the Lns that appears here
| |
| 13:06 | * Lns blushes... aww shucks | |
| 13:07 | <Gadi> :)
| |
| 13:07 | whew
| |
| 13:07 | <Lns> gg
| |
| 13:07 | <Gadi> Lns: wrt shutdown command, according to ogra, what we have in upstream is wrong
| |
| 13:07 | and he does not know where it came from
| |
| 13:07 | but, I dont either
| |
| 13:08 | so, I am loathe to change unless it fixes a bug
| |
| 13:08 | which it sounds like there is
| |
| 13:08 | :)
| |
| 13:08 | <Lns> wrt?
| |
| 13:08 | <Gadi> with respect to
| |
| 13:08 | <Lns> ah
| |
| 13:08 | <Gadi> sorry - Im younger than I feel
| |
| 13:08 | :)
| |
| 13:08 | otavio has joined #ltsp | |
| 13:08 | <Lns> lol
| |
| 13:08 | so, what *is* the cmd ldm uses to shut down?
| |
| 13:09 | <Gadi> currently: poeroff -p
| |
| 13:09 | <Lns> ah
| |
| 13:09 | <Gadi> *poweroff -p
| |
| 13:09 | which should be (according to ogra)
| |
| 13:09 | poweroff -fp
| |
| 13:09 | which makes more sense to me
| |
| 13:09 | dunno where we lost the "f"
| |
| 13:09 | <Lns> it does to me, but it seems to disable (at least on some systems) wol support
| |
| 13:10 | even with my spicy "wolfix" script that calls "ethtool -s eth0 wol g"
| |
| 13:10 | which turns wol "magic packet" support on after it boots
| |
| 13:10 | <laga> hum
| |
| 13:10 | back in the day, you had to edit this init script on ubuntu..
| |
| 13:10 | <Lns> again, just on some systems i've run into though..so it seems it could be h/w specific
| |
| 13:11 | <Gadi> some hardware will power down the NICs when the system is halted
| |
| 13:11 | ie complete power down
| |
| 13:12 | <Lns> laga: the runlevel 0 init ?
| |
| 13:12 | <Gadi> to draw practically no power
| |
| 13:12 | <Lns> Gadi: yes, but when i use /sbin/halt, it works
| |
| 13:12 | <Gadi> for wol to work, there has to be at least power to the NICs
| |
| 13:12 | ah, so you would be a candidate for: poweroff -fp
| |
| 13:12 | :)
| |
| 13:13 | too bad its compiled in the greeter
| |
| 13:13 | :P
| |
| 13:13 | <Lns> maybe? I'm sorta confused as to the differences
| |
| 13:13 | Gadi: yeah, i saw that..i wish it was more configurable
| |
| 13:13 | <Gadi> -fp says : run halt, damnit!
| |
| 13:13 | <Lns> would be nice to customize those entries too to add/remove some
| |
| 13:13 | Gadi: ah
| |
| 13:13 | <Gadi> -p says: run shutdown in a halt kind of way
| |
| 13:13 | dan_young has joined #ltsp | |
| 13:14 | <Lns> so technically, what are the diffs ?? are these all atx specific call differences?
| |
| 13:14 | seems that some turn off "more" than others
| |
| 13:14 | nm, you already said =p
| |
| 13:14 | dan_young has left #ltsp | |
| 13:15 | <Lns> just weird how other systems behave differently when using the same methods
| |
| 13:15 | <Gadi> from man shutdown: -h Requests that the system be either halted or powered off after it has been brought down, with the choice as
| |
| 13:15 | to which left up to the system.
| |
| 13:15 | <Lns> even the same models, just slightly different versions
| |
| 13:16 | <Gadi> ie, there is a difference between halt and poweroff
| |
| 13:17 | LaserJock has left #ltsp | |
| 13:18 | <Lns> yeah..i remember reading that on sparc workstations 'halt' (iirc) will bring it back to openfirmware boot menu or some such
| |
| 13:20 | <alkisg> "-p Instructs the halt command to instead behave as poweroff." => then why doesn't one just use the poweroff command???
| |
| 13:21 | <Lns> alkisg: that'd be too easy =p
| |
| 13:21 | <alkisg> Heh, lol :)
| |
| 13:23 | <cliebow> weee!
| |
| 13:25 | <Lns> cliebow: please step away from the laughing gas
| |
| 13:27 | <laga> Lns: i was talking about /etc/init.d/halt and modify NETDOWN
| |
| 13:27 | <Gadi> Lns: if you dont want to change the code, you could make /opt/ltsp/i386/sbin/poweroff a symlink to your own script
| |
| 13:27 | <alkisg> Hrm... I don't get it... other SDL apps run fine in TCs (e.g. supertux2 => about 15 fps) and other, similar ones, really slow (e.g. secret mario chronicles, smc => about 0.1 fps)
| |
| 13:27 | <Gadi> alkisg: have you been following this xcb bug
| |
| 13:28 | ?
| |
| 13:28 | <alkisg> Gadi, I've seen about it, but I don't know what xcb is :)
| |
| 13:28 | Do you think it also affects SDL?
| |
| 13:28 | <Gadi> it is something that is supposed to boost performance of X, but winds up crippling performance of remote X
| |
| 13:28 | dunno if it is related
| |
| 13:29 | so far, it seems to improve such things as: java apps, right-click context menus in firefox, other
| |
| 13:29 | *others
| |
| 13:29 | improve meaning improve is you remove xcb
| |
| 13:29 | :)
| |
| 13:29 | <alkisg> And to try the workaround I just have to overwrite some .so with the gutsy version? Or completely remove a .so?
| |
| 13:29 | <Gadi> well, I did it differently
| |
| 13:30 | I apt-get'd the source for libx11-6
| |
| 13:30 | changed the configure flags and such
| |
| 13:31 | and rebuilt a deb with libs that dont depend on xcb
| |
| 13:31 | so far, I definitely see some improvements
| |
| 13:31 | <alkisg> Ah! That sounds better! :)
| |
| 13:31 | <Gadi> I added: --with-xcb=no
| |
| 13:31 | to the configure flags
| |
| 13:32 | <alkisg> Gadi: what's your .deb for? Intrepid? :)
| |
| 13:32 | <Gadi> yeah
| |
| 13:32 | i386
| |
| 13:32 | <Lns> Gadi: that's what I was thinking too (symlink)
| |
| 13:32 | <Gadi> not 64 bit
| |
| 13:32 | <alkisg> Yeah!!! :) Could you please pretty please upload it somewhere? :)
| |
| 13:32 | <Gadi> do you want me to post it someplace?
| |
| 13:32 | heh
| |
| 13:32 | sure
| |
| 13:33 | <alkisg> It would save me from learning how to build a .deb... :P
| |
| 13:33 | <Gadi> the update manager will try to replace it when you install it
| |
| 13:33 | so, as long as you don't apt-get upgrade
| |
| 13:33 | it will be there
| |
| 13:33 | one sec, I'll post it someplace
| |
| 13:33 | <Lns> woo, testers!
| |
| 13:33 | <alkisg> Fine, just trying to see if this is my problem
| |
| 13:33 | <Lns> laga: i'll look into NETDOWN that might be my issue
| |
| 13:35 | <Gadi> ftp://symbio-technologies.com/libx11-6_1.1.5-2ltsp1.1_i386.deb
| |
| 13:35 | there ya go
| |
| 13:36 | let me know if it fixes things or offers any problems
| |
| 13:36 | <alkisg> Thanks!! I sure will.
| |
| 13:36 | <Lns> What's up with the xcb people not responding to stuff like what we're experiencing
| |
| 13:36 | <Gadi> at some point, I want to try the firefox patch to cache resized pixmaps instead of original size
| |
| 13:37 | anyone already try that ?
| |
| 13:38 | <alkisg> Gadi, erm, this should go into the chroot, right?
| |
| 13:38 | <Gadi> no
| |
| 13:38 | on the server
| |
| 13:38 | its an Xclient library
| |
| 13:38 | <alkisg> Darn... I still don't get the X staff
| |
| 13:38 | Ah, client, ok
| |
| 13:39 | <Gadi> in other words, apps that can use xcb function calls will use them if they are present
| |
| 13:39 | this makes them not present
| |
| 13:39 | so they need to use standard x11 calls
| |
| 13:39 | the way I see it, if devs keep developing this way, remote X11 is pretty much dead
| |
| 13:39 | and we'll have to do everything by proxy
| |
| 13:40 | * Lns is NOT happy about that | |
| 13:40 | <alkisg> That would be silly, X was network-aware from the start
| |
| 13:40 | network-transparent (more correctly)
| |
| 13:40 | * Gadi agrees | |
| 13:41 | hanthana has quit IRC | |
| 13:41 | <Gadi> of course, back then there were fewer images being sloshed around
| |
| 13:41 | <Lns> We need to get a formal group together for LTSP to counter this trend
| |
| 13:41 | <Gadi> :)
| |
| 13:42 | <Lns> Gadi: is XCB contributing to this mostly?
| |
| 13:42 | <Gadi> not the image thing
| |
| 13:42 | from what I can gather, xcb just has a bug in it
| |
| 13:42 | where it does too much "checking" of the xserver memory
| |
| 13:43 | which, locally, has no ill effect
| |
| 13:43 | but, remotely, causes a lot of traffic and latency when waiting for an answer
| |
| 13:43 | in other words, say the app wants to do something
| |
| 13:43 | it goes:
| |
| 13:44 | ok xserver, how much memory can I use
| |
| 13:44 | start action
| |
| 13:44 | wait, is this memory address available?
| |
| 13:44 | ok, continue
| |
| 13:44 | wait, so you're saying Im good if I write to this memory address?
| |
| 13:44 | <Lns> wow
| |
| 13:44 | <Gadi> ok, continue
| |
| 13:44 | :)
| |
| 13:44 | <Lns> too chatty
| |
| 13:44 | <Gadi> thats my read on it
| |
| 13:45 | exactly
| |
| 13:45 | x11 by itself is very chatty
| |
| 13:45 | and it sounds like xcb just chats it up some more
| |
| 13:45 | <Lns> heh
| |
| 13:45 | <Gadi> thats really NX's big win
| |
| 13:46 | <Lns> would there be any benefit for an x11-ltsp type extension/modification to the normal proto that would reduce this stuff as much as possible?
| |
| 13:46 | <Gadi> its not so much in the compression
| |
| 13:46 | its in keeping the chatting to a minimum
| |
| 13:46 | <Lns> hmm
| |
| 13:46 | <Gadi> well, I'm no X guru, but it seems to me that if you have a protocol that is both chatty and network transparent,
| |
| 13:47 | you might want to give some flexibility to tradeoff
| |
| 13:47 | very chatty on laptop, good
| |
| 13:47 | not so chatty on thin client good
| |
| 13:47 | other way around, not good
| |
| 13:48 | <Lns> XSERVER_ISLOCAL = ?
| |
| 13:48 | <Gadi> well, X already knows where the Xserver is
| |
| 13:48 | echo $DISPLAY
| |
| 13:48 | * alkisg is rebooting to test xcb... | |
| 13:48 | <Lns> but to specifically modify the 'chattiness'
| |
| 13:49 | <Gadi> well, there should definitely be a dial for the chattiness
| |
| 13:49 | <Lns> yeah
| |
| 13:49 | that's what i was going for
| |
| 13:49 | alkisg has quit IRC | |
| 13:49 | <Lns> what would have to be sacrificed for less chattiness? Can we learn from nx?
| |
| 13:49 | w/o proxy style modifications that is
| |
| 13:49 | <Gadi> nothing wrong with a proxy
| |
| 13:50 | as long as it is designed for what you want to achieve
| |
| 13:50 | <Lns> chroot integration?
| |
| 13:51 | <Gadi> I meant more like accelerated video
| |
| 13:51 | <Lns> oh
| |
| 13:51 | <Gadi> ssh has a proxy
| |
| 13:51 | but it sucks
| |
| 13:51 | because it is not really designed for accelerated video over LAN
| |
| 13:51 | it is more of an afterthought
| |
| 13:51 | alkisg has joined #ltsp | |
| 13:51 | <Lns> right
| |
| 13:52 | <Gadi> to tunnel Xclients
| |
| 13:52 | alkisg has quit IRC | |
| 13:52 | * Lns uses that for his clients | |
| 13:52 | <Lns> w/vnc
| |
| 13:52 | <Gadi> right
| |
| 13:52 | alkisg has joined #ltsp | |
| 13:52 | <Gadi> Lns: you should look into turbovnc
| |
| 13:53 | I really like its performance
| |
| 13:53 | * Lns googles | |
| 13:53 | <Gadi> www.virtualgl.org
| |
| 13:54 | atxppl has joined #ltsp | |
| 13:54 | <alkisg> Hey, it fixed the ssh -Y localhost firefox right click delay! :)
| |
| 13:54 | cliebow has quit IRC | |
| 13:54 | <Gadi> :)
| |
| 13:54 | yeah
| |
| 13:54 | <alkisg> trying remotely...
| |
| 13:54 | <Gadi> brb
| |
| 13:55 | <Lns> damn you xcb!!
| |
| 13:55 | lol
| |
| 13:55 | <alkisg> Dah, no, remotely the delay is still big.
| |
| 13:57 | I don't get it. I did `ssh -Y server` and right click was slow. Then I did `ssh -Y client` _from_ the server with the modified xcb, and right click was fast!!!
| |
| 13:57 | (client = a standalone pc, non ltsp related)
| |
| 13:58 | <Lns> Gadi: wow, virtualgl looks nice!! I'm assuming ltsp could use that
| |
| 13:58 | <Gadi> Lns: well, the virtualgl part, for sure
| |
| 13:58 | the turbovnc, not as much
| |
| 13:59 | <Lns> right
| |
| 13:59 | <Gadi> but turbovnc is quite good in its own right
| |
| 14:00 | <Lns> "They appear to the application as a normal X server, but they perform X11 rendering to a virtual framebuffer in main memory rather than to a real framebuffer on a graphics card. This allows the X proxy to send only images to the client machine rather than chatty X-Windows rendering commands." --- sending images < x-windows rendering commands?
| |
| 14:00 | is that 3d specific?
| |
| 14:01 | <Gadi> are you reading the virtualgl or the turbovnc?
| |
| 14:01 | <Lns> virtualgl
| |
| 14:02 | <Gadi> basically what it does is this:
| |
| 14:02 | takes the 3D rendering part and sends it to a running Xserver on the *server-side*
| |
| 14:02 | so, pop your most powerful graphics card in the server
| |
| 14:03 | then, what it gets back are 2D frames
| |
| 14:03 | which it merges in with the 2D frames sent to the client
| |
| 14:03 | so, the client gets all 2D info
| |
| 14:04 | but the 3D stuff is rendered by the server (or farm of graphics servers)
| |
| 14:04 | :)
| |
| 14:04 | its neat
| |
| 14:04 | its an answer to not using the graphics power of the desktop, but that of the server
| |
| 14:05 | <Lns> so sending 2d "frames" = 2d "images" ? or x11 commands to draw 2d images?
| |
| 14:05 | <Gadi> the latter, I believe
| |
| 14:05 | <Lns> ah ok
| |
| 14:05 | <Gadi> tho, not 100% sure
| |
| 14:05 | <Lns> i guess that would only apply to window frames & friends
| |
| 14:06 | <Gadi> basically, if you had no GL support on the desktop, it would still work
| |
| 14:06 | its mainly for accelerating video
| |
| 14:06 | as in movies
| |
| 14:06 | or weather models
| |
| 14:06 | :)
| |
| 14:10 | vagrantc has joined #ltsp | |
| 14:10 | <Lns> that's awesome
| |
| 14:10 | |Paradox| has quit IRC | |
| 14:10 | atxppl is now known as |Paradox| | |
| 14:12 | <alkisg> Secret Mario Chronicles and supertuxkart went from 0.1 fps to 4 fps, impressive speedup! But in glxgears, even though it reports the double fps, the gears stay still...
| |
| 14:16 | <Gadi> alkisg: so what does that mean?
| |
| 14:16 | are the games playable?
| |
| 14:17 | <alkisg> Erm, almost playable, it would be better with > 10 fps
| |
| 14:18 | But it's a great speedup... let's see the other apps (I was making a list with apps that play well and others that don't)
| |
| 14:19 | <Gadi> http://www.freelists.org/post/uae/SDL-speedup
| |
| 14:19 | alkisg: ^^ interesting
| |
| 14:20 | <alkisg> Ahm, this needs compiling... a summer task :)
| |
| 14:23 | Video playback with totem has a little lag now...
| |
| 14:30 | <Lns> alkisg: your'e talking about emulated snes games?
| |
| 14:30 | <alkisg> Lns, no, mostly SDL games
| |
| 14:31 | <Lns> alkisg: you konw about the --nosound and --window[ed] switches for a lot of the tux games right?
| |
| 14:31 | <alkisg> 10 of the ones I tried showed no improvement, but the ones that were really slow (<1fps) improved a lot (=4-5 fps)
| |
| 14:32 | Lns, no, but right now I'm trying with plain ssh -Y, so no sound :)
| |
| 14:33 | tuxpaint runs fine with plain ssh -Y, even without the modified xcb lib
| |
| 14:37 | <Lns> alkisg: use --nosound if you're not getting sound anyway. That switch alone greatly improved sdl app performance
| |
| 14:38 | alkisg: https://bugs.launchpad.net/ubuntu/+source/tuxmath/+bug/269082 for reference
| |
| 14:39 | spectra has quit IRC | |
| 14:40 | <Gadi> it really is amazing how simple operations cause tens of megabytes of x11 traffic to go ver the wire
| |
| 14:46 | <Lns> Gadi: in your opinion, do you think X11 will still be a viable as a network-transparent GUI framework in the next 10 years?
| |
| 14:46 | <Gadi> thats really up to upstream Xorg
| |
| 14:46 | <johnny> Lns, i think so
| |
| 14:46 | <Gadi> and how much they value the network transparency
| |
| 14:46 | <johnny> i think it will be streamlined
| |
| 14:46 | more work done on the client
| |
| 14:47 | as far as display is concerned
| |
| 14:47 | <Gadi> johnny: unfortunately, what app devs are doing is they see greater efficiencies to let the Xserver do the processing
| |
| 14:47 | <johnny> Lns, perhaps not viable on old stuff :)
| |
| 14:47 | isn't that up to the toolkit?
| |
| 14:47 | more than the app?
| |
| 14:48 | <Gadi> depends, I suppose
| |
| 14:49 | alkisg has quit IRC | |
| 14:50 | <Lns> What can we do to keep in constant contact w/Xorg upstream with our work, to keep them on the ball in regards to improving network transparency in X?
| |
| 14:50 | <johnny> talk to them?
| |
| 14:50 | <Lns> and not let them disregard it as "old school mainframe" tech
| |
| 14:51 | <johnny> Lns, go meet a dev
| |
| 14:51 | Lns, generally greater efficiencies are just around the corner..
| |
| 14:51 | it's like hal/pulseaudio all this new stuff
| |
| 14:51 | still evolving
| |
| 14:52 | sad but true.. sometimes things get stuck not working for awhile
| |
| 14:52 | berriop has left #ltsp | |
| 14:52 | <Lns> johnny: so they're actually improving network xparency and not just leaving it as is
| |
| 14:52 | and improving local xserver/client stuff
| |
| 14:53 | <johnny> i'm hoping to see less data transfered
| |
| 14:53 | overall
| |
| 14:53 | <Lns> yeah
| |
| 14:53 | <johnny> which should help everybody
| |
| 14:53 | Lns, are you famliar with all this new gem and kms stuff?
| |
| 14:53 | <Lns> johnny: nope, not at all
| |
| 14:54 | <johnny> it sounds like you should be
| |
| 14:54 | otherwise you're not going to be able to get the help you need
| |
| 14:55 | Lns, there was a good lost half decade
| |
| 14:55 | at the very least..
| |
| 14:55 | it's only in the very recent past that we're starting to recover from xfree86
| |
| 14:56 | Egyptian[Home] has quit IRC | |
| 14:58 | Egyptian[Home] has joined #ltsp | |
| 14:58 | aledec has quit IRC | |
| 14:58 | aledec has joined #ltsp | |
| 14:59 | * Lns wishes Apple would share some of their great toys with the world, since they built upon toys that were around before they were | |
| 15:12 | <Lns> i still can't rename my own wikipages on https://help.ubuntu.com/community/UbuntuLTSP :( that shutdown typo is bugging the crap out of me
| |
| 15:19 | <yotux> I am getting and error on my thin client: File transfer error: Image file too large for low memory.
| |
| 15:19 | can anyone offer some assistance on this one
| |
| 15:19 | <Gadi> yotux: sounds like you are trying to feed an etherboot image to a PXE client
| |
| 15:20 | have you been changin your dhcpd.conf?
| |
| 15:20 | *changing
| |
| 15:21 | <yotux> Gadi: no stock install of ubuntu
| |
| 15:22 | <Gadi> what is the name of the file it attempts to download?
| |
| 15:23 | <yotux> I don't get the name of the fie
| |
| 15:25 | Gadi: http://pastebin.com/d4bcfc759 this is what my thin client displays
| |
| 15:26 | <Lns> yotux: just curious, how much mem does your TC have?
| |
| 15:26 | (to eliminate the obvious error)
| |
| 15:26 | redspike_ has quit IRC | |
| 15:27 | <Lns> 639k lowmem? ;)
| |
| 15:27 | <yotux> 320MB ram
| |
| 15:28 | I'm using an old IBM Netvista with a 3com network card
| |
| 15:29 | I used LTSP back when Ubutnu was 6.04 if I recall correctly
| |
| 15:30 | <Gadi> I think the netvistas have their own sets of issues
| |
| 15:30 | maybe cliebow knows?
| |
| 15:30 | hmm... not here atm
| |
| 15:31 | <Lns> !seen cliebow
| |
| 15:31 | <ltspbot> Lns: cliebow was last seen in #ltsp 2 hours, 7 minutes, and 21 seconds ago: <cliebow> weee!
| |
| 15:31 | <Lns> lol
| |
| 15:31 | too much laughing gas
| |
| 15:32 | redspike has joined #ltsp | |
| 15:33 | litlebuda has joined #ltsp | |
| 15:34 | rjune has quit IRC | |
| 15:37 | wwx has quit IRC | |
| 15:39 | wwx has joined #ltsp | |
| 15:40 | Gadi has left #ltsp | |
| 15:42 | vvinet has quit IRC | |
| 15:44 | Shingoshi has quit IRC | |
| 15:46 | yotux has left #ltsp | |
| 16:01 | <johnny> yum is so lame
| |
| 16:01 | it breaks all the time..
| |
| 16:01 | or packagekit breaks it..
| |
| 16:02 | get thread died in berkeleydb library at least once every week
| |
| 16:02 | and you can't even ctrl+c it
| |
| 16:02 | you have to go kill -9 it elswhere
| |
| 16:08 | <Lns> gross
| |
| 16:08 | * Lns has never had luck with RPM-based pkg mgmt, even newer ones | |
| 16:20 | redspike has quit IRC | |
| 16:41 | Shingoshi has joined #ltsp | |
| 16:54 | RobertLaptop has quit IRC | |
| 17:00 | Lns has quit IRC | |
| 17:09 | bobby_C has quit IRC | |
| 17:12 | mih_d has quit IRC | |
| 17:14 | redspike has joined #ltsp | |
| 17:15 | jammcq has quit IRC | |
| 17:30 | phantom has quit IRC | |
| 17:30 | phantom has joined #ltsp | |
| 18:19 | gentgeen__ has joined #ltsp | |
| 18:25 | vagrantc has quit IRC | |
| 18:32 | jammcq has joined #ltsp | |
| 18:34 | cyberorg has quit IRC | |
| 18:37 | SDuensin has quit IRC | |
| 18:45 | litlebuda has quit IRC | |
| 18:49 | litlebuda has joined #ltsp | |
| 18:50 | SDuensin has joined #LTSP | |
| 18:56 | otavio has quit IRC | |
| 18:59 | staffencasa has quit IRC | |
| 19:05 | otavio has joined #ltsp | |
| 19:12 | RobertLaptop has joined #ltsp | |
| 19:19 | japerry has quit IRC | |
| 19:33 | otavio has quit IRC | |
| 20:06 | nothingman has joined #ltsp | |
| 20:11 | aledec is now known as aledec90 | |
| 20:11 | aledec90 is now known as aledec | |
| 20:18 | pmatulis has quit IRC | |
| 20:26 | litlebuda has quit IRC | |
| 20:30 | japerry has joined #ltsp | |
| 20:31 | johnb has joined #ltsp | |
| 21:04 | johnb has quit IRC | |
| 21:11 | mistik1 has quit IRC | |
| 21:28 | mistik1 has joined #ltsp | |
| 21:43 | <nothingman> has anyone here used a vm as a pxe client without a server running as a guest vm?
| |
| 21:45 | in other words, to a pxe server running on the host?
| |
| 21:45 | <Ryan52> yes.
| |
| 21:45 | which distro?
| |
| 21:45 | <nothingman> Ubuntu
| |
| 21:46 | <Ryan52> hmm. I dunno then.
| |
| 21:47 | <nothingman> which vm software have you used?
| |
| 21:47 | I have Virtualbox and qemu
| |
| 21:47 | <Ryan52> kvm (which is basically qemu)
| |
| 21:47 | <nothingman> every howto I find is either really vague or gets about halfway through before telling me I need to run the server in another vm
| |
| 21:47 | yeah
| |
| 21:47 | that I have
| |
| 21:48 | <Ryan52> Fedora's ltsp comes with a script to run a thin client vm.
| |
| 21:48 | tho it's easier on Fedora because all the daemons are already listening on a bridge. not sure how that would work with normal networking..
| |
| 21:57 | alkisg has joined #ltsp | |
| 21:58 | <nothingman> Ryan52: I get "No valid PXE ROM found for network device."
| |
| 21:59 | <Ryan52> not enough information provided, and I'm not good with the qemu stuff.
| |
| 22:00 | * Ryan52 still has nightmares :P | |
| 22:00 | CaScAdE^1arAway has joined #ltsp | |
| 22:05 | <johnny> nothingman, sure.. that's fine
| |
| 22:05 | it works
| |
| 22:05 | as long as you setup briding
| |
| 22:05 | bridging*
| |
| 22:06 | ltsp app stuff on host.. client in virtualbox..
| |
| 22:06 | mih has joined #ltsp | |
| 22:06 | <johnny> worked fine
| |
| 22:08 | CrazyPotato has joined #ltsp | |
| 22:08 | _wigwam_ has joined #ltsp | |
| 22:09 | <CrazyPotato> Hello.
| |
| 22:17 | CaScAdE^FarAway has quit IRC | |
| 22:17 | CrazyPotato has quit IRC | |
| 22:21 | wigwam has quit IRC | |
| 22:33 | <alkisg> So, about the sdl-sound performance problems in ltsp clients, if I installed libsdl1.2debian-pulseaudio instead of libsdl1.2debian-alsa, I might get better performance?
| |
| 22:57 | nothingman has quit IRC | |
| 23:37 | ScorpKing has joined #ltsp | |
| 23:37 | ajohnson has quit IRC | |
| 23:47 | ScorpKing has quit IRC | |
| 23:47 | Ahmuck has quit IRC | |