IRC chat logs for #ltsp on irc.libera.chat (webchat)


Channel log from 24 May 2007   (all times are UTC)

00:04Burgundavia has joined #ltsp
00:25cliebow has quit IRC
00:29cliebow has joined #ltsp
00:29Burgundavia has quit IRC
00:30Burgundavia has joined #ltsp
00:41privet has joined #ltsp
00:44gu has quit IRC
00:44oh207 has quit IRC
00:44Patina has quit IRC
00:44shogunx has quit IRC
00:44shogunx has joined #ltsp
00:45Patina has joined #ltsp
00:48gu has joined #ltsp
00:48oh207 has joined #LTSP
00:49Yo has joined #ltsp
00:51Yo has left #ltsp
00:55cliebow has quit IRC
00:55cliebow has joined #ltsp
01:01Yo has joined #ltsp
01:02Yo has quit IRC
01:03Burgundavia has quit IRC
01:05Burgundavia has joined #ltsp
01:14highvolt1ge is now known as highvoltage
01:28Burgundavia has quit IRC
01:32cliebow has quit IRC
01:37Burgundavia has joined #ltsp
01:54cliebow has joined #ltsp
02:03Burgundavia has quit IRC
02:22maaarek has joined #ltsp
02:25federicoco has joined #ltsp
02:25
<federicoco>
hello all
02:25
1 question.....apt is present on ltsp.iso??
02:45maaarek has quit IRC
02:46MoL0ToV has joined #ltsp
02:47
<federicoco>
hello...i whant install java on my ltsp client ....any idea????? helllllppppppp :-)
02:48vagrantc has joined #ltsp
02:52vanya has joined #ltsp
02:56cliebow has quit IRC
02:57cliebow has joined #ltsp
03:05privet has quit IRC
03:15federicoco has quit IRC
03:21privet has joined #ltsp
03:29jsgotangco has joined #ltsp
03:31cliebow has quit IRC
03:39meduxa has joined #ltsp
03:41Avatara has joined #ltsp
03:46ChinaForge has joined #ltsp
03:48
<ChinaForge>
Hello,I install ltsp 5 on Ubuntu 7.04,ltsp support usb device?
03:49
<vagrantc>
sure does
03:50
well
03:50
it supports usb drives and such
03:50
not cd-writers or scanners or anything like that
03:51
<ChinaForge>
but i can't,i plug a usb disk on my terminal but nothing by detect.
03:52
<vagrantc>
grep LOCALDEV /opt/ltsp/i386/etc/lts.conf
03:53
<ChinaForge>
LOCALDEV=True
03:54
<vagrantc>
dpkg --root=/opt/ltsp/i386 -l ltspfsd
03:54
dpkg -l ltspfs
03:55
<ChinaForge>
期望状态=未知(u)/安装(i)/删除(r)/清除(p)/保持(h)
03:55
| 当前状态=未(n)/已安装(i)/仅存配置(c)/仅解压缩(U)/配置失败(F)/不完全安装(H)
03:55
|/ 错误?=(无)/保持(?)/须重装(R)/两者兼有(#) (状态,错误:大写=故障)
03:55
||/ 名称 版本 简介
03:55
+++-==============-==============-============================================
03:55
ii ltspfsd 0.4.3-0ubuntu6 Fuse based remote filesystem daemon for LTSP
03:55
<vagrantc>
grep SCREEN /opt/ltsp/i386/etc/lts.conf
03:55cliebow has joined #ltsp
03:56
<ChinaForge>
nothing
03:56
<vagrantc>
ok.
03:56
dpkg -l ltspfs
03:56
<ChinaForge>
期望状态=未知(u)/安装(i)/删除(r)/清除(p)/保持(h)
03:56
| 当前状态=未(n)/已安装(i)/仅存配置(c)/仅解压缩(U)/配置失败(F)/不完全安装(H)
03:56
|/ 错误?=(无)/保持(?)/须重装(R)/两者兼有(#) (状态,错误:大写=故障)
03:56
||/ 名称 版本 简介
03:56
+++-==============-==============-============================================
03:56
un ltspfs <无> (无相关介绍)
03:56
<vagrantc>
sudo apt-get install ltspfs
03:56
and try again
03:57
<ChinaForge>
I install complete,need reboot?
03:58
<vagrantc>
i think you just need to log out of the thin client
03:58
but maybe reboot the thin client.
03:58
<ChinaForge>
ok.
03:58ChinaForge has quit IRC
03:58
<vagrantc>
server shouldnt need a reboot at all
04:07ChinaForge has joined #ltsp
04:08
<ChinaForge>
Can't!!
04:08
<vagrantc>
don't stress out so much :P
04:09
ChinaForge: is your user in the fuse group/
04:09
?
04:10
<ChinaForge>
no
04:11
<vagrantc>
add your user to the fuse group, log out, unplug the usb stick, log back in, plug the usb stick in, look in /media/USERNAME
04:11ChinaForge has quit IRC
04:11ChinaForge has joined #ltsp
04:12ogra has joined #ltsp
04:13
<vagrantc>
ChinaForge: you might find this page useful https://wiki.ubuntu.com/EnableLTSP5LocalDevices
04:13
<ChinaForge>
sorry,Xchat exit self,I can't see you message,
04:13
<vagrantc>
ChinaForge: you might find this page useful https://wiki.ubuntu.com/EnableLTSP5LocalDevices
04:13
ChinaForge: add your user to the fuse group, log out, unplug the usb stick, log back in, plug the usb stick in, look in /media/USERNAME
04:14
<ChinaForge>
ok,i try,thank you!
04:15meduxa has quit IRC
04:15meduxa has joined #ltsp
04:17ChinaForge has quit IRC
04:17Q-FUNK has joined #ltsp
04:18cliebow has quit IRC
04:19ChinaForge has joined #ltsp
04:20
<ChinaForge>
vagrantc:It right,Thank you very much!
04:20
<vagrantc>
ChinaForge: no problem :)
04:25meduxa has joined #ltsp
04:26
<ChinaForge>
vagrantc,I have other problem,before consult for you,I setting host by other tech document,
04:26
<vagrantc>
ChinaForge: sorry, i don't quite understand you ...
04:26
<ChinaForge>
like lts.conf,
04:26
I add LOCAL_DEVICE_01 = /dev/hdc:cdrom
04:26
LOCAL_DEVICE_02 = /dev/fd0:floppy
04:26
LOCAL_DEVICE_03 = /dev/sda1:usb
04:26
HOTPLUG = Y
04:26
ALLOW_UNKNOWN_USB = Y
04:27
Is must for LTSP 5?
04:27
<vagrantc>
i don't think any of those options are used in LTSP 5 ...
04:30
!localdev
04:30
<ltspbot>
vagrantc: "localdev" is Information regarding using pluggable devices and cdroms with LTSP can be found at http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev
04:30
<ChinaForge>
If I reinstall Ubuntu 7.04 and ltsp 5,What must to do?
04:30
<vagrantc>
ltspbot: learn localdev as for ubuntu: https://wiki.ubuntu.com/EnableLTSP5LocalDevices
04:30
<ltspbot>
vagrantc: The operation succeeded.
04:31
<vagrantc>
ltspbot: learn localdev as for debian-edu: http://wiki.debian.org/DebianEdu/HowTo/LocalDeviceLtspfs
04:31
<ltspbot>
vagrantc: The operation succeeded.
04:31
<vagrantc>
!localdev
04:31
<ltspbot>
vagrantc: "localdev" is (#1) Information regarding using pluggable devices and cdroms with LTSP can be found at http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev, or (#2) for ubuntu: https://wiki.ubuntu.com/EnableLTSP5LocalDevices, or (#3) for debian-edu: http://wiki.debian.org/DebianEdu/HowTo/LocalDeviceLtspfs
04:31
<vagrantc>
ChinaForge: follow the instructions in the wiki above for ubuntu
04:32
ltspbot: factoids search --values
04:32
<ltspbot>
vagrantc: 'ltsp', 'ubuntu', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'grayscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'serversizing', 'wireless', 'sound', 'topics', 'ubuntuquick', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', (1 more message)
04:32
<vagrantc>
!greyscreen
04:32
<ltspbot>
vagrantc: "greyscreen" is No login, and only a grey screen? See http://wiki.ltsp.org/twiki/bin/view/Ltsp/Troubleshooting-GrayScreen
04:32
<ChinaForge>
#1,#2#3 all to do?
04:32
<vagrantc>
!grayscreen
04:33
<ltspbot>
vagrantc: "grayscreen" is No login, and only a grey screen? See http://wiki.ltsp.org/twiki/bin/view/Ltsp/Troubleshooting-GrayScreen#GDM
04:33
<vagrantc>
ChinaForge: look at the one that says "for ubuntu"
04:33
<ChinaForge>
ok.
04:33
<vagrantc>
ltspbot: forget grayscreen
04:33
<ltspbot>
vagrantc: The operation succeeded.
04:33
<vagrantc>
ltspbot: factoids search --values
04:33
<ltspbot>
vagrantc: 'ltsp', 'ubuntu', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'serversizing', 'wireless', 'sound', 'topics', 'ubuntuquick', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', (1 more message)
04:33
<vagrantc>
ltspbot: factoids search --values --all
04:33
<ltspbot>
vagrantc: (factoids search [<channel>] [--values] [--{regexp} <value>] [<glob> ...]) -- Searches the keyspace for keys matching <glob>. If --regexp is given, it associated value is taken as a regexp and matched against the keys. If --values is given, search the value space instead of the keyspace.
04:33
<vagrantc>
ltspbot: factoids search --values *
04:33
<ltspbot>
vagrantc: 'ltsp', 'ubuntu', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'serversizing', 'wireless', 'sound', 'topics', 'ubuntuquick', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', (1 more message)
04:34
<vagrantc>
!localdev 2
04:34
<ltspbot>
vagrantc: Error: "localdev" is not a valid command.
04:34
<vagrantc>
!localdev
04:34
<ltspbot>
vagrantc: "localdev" is (#1) Information regarding using pluggable devices and cdroms with LTSP can be found at http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev, or (#2) for ubuntu: https://wiki.ubuntu.com/EnableLTSP5LocalDevices, or (#3) for debian-edu: http://wiki.debian.org/DebianEdu/HowTo/LocalDeviceLtspfs
04:34
<vagrantc>
!serversize
04:34
<ltspbot>
vagrantc: "serversize" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/ServerSizing
04:34
<vagrantc>
!serversizing
04:34
<ltspbot>
vagrantc: "serversizing" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/ServerSizing
04:35
<vagrantc>
ltspbot: forget serversizing
04:35
<ltspbot>
vagrantc: The operation succeeded.
04:35
<vagrantc>
ltspbot: factoids search --values *
04:35
<ltspbot>
vagrantc: 'ltsp', 'ubuntu', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'wireless', 'sound', 'topics', 'ubuntuquick', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', 'js', 's', (1 more message)
04:35* vagrantc wonders how to get rid of the (1 more message)
04:36
<vagrantc>
!ubuntuquick
04:36
<ltspbot>
vagrantc: "ubuntuquick" is https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:37
<vagrantc>
!ubuntu
04:37
<ltspbot>
vagrantc: "ubuntu" is a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org
04:37
<vagrantc>
ltspbot: learn ubuntu as to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:37
<ltspbot>
vagrantc: The operation succeeded.
04:37
<vagrantc>
!ubuntu
04:37
<ltspbot>
vagrantc: "ubuntu" is (#1) a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org, or (#2) to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:38
<vagrantc>
ltspbot: forget ubuntuquick
04:38
<ltspbot>
vagrantc: The operation succeeded.
04:38
<vagrantc>
ltspbot: factoids search --values
04:38
<ltspbot>
vagrantc: 'ltsp', 'ubuntu', 'ubuntu', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'wireless', 'sound', 'topics', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', 'js', 's', (1 more message)
04:38
<vagrantc>
!debian
04:38
<ltspbot>
vagrantc: "debian" is is a GNU/Linux based operating system that makes an excellent LTSP server. You can find it at http://www.debian.org. for information about LTSP on debian see http://wiki.debian.org/LTSP
04:38
<vagrantc>
!ubuntu
04:38
<ltspbot>
vagrantc: "ubuntu" is (#1) a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org, or (#2) to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:39
<vagrantc>
ltspbot: learn ubuntu as a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to instal lltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:39
<ltspbot>
vagrantc: The operation succeeded.
04:39
<vagrantc>
!ubuntu
04:39
<ltspbot>
vagrantc: "ubuntu" is (#1) a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org, or (#2) to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall, or (#3) a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. (1 more message)
04:39
<vagrantc>
ltspbot: forget ubuntu 1
04:39
<ltspbot>
vagrantc: The operation succeeded.
04:39
<ChinaForge>
vagrant,Thank you, I slow slow study!
04:39
<vagrantc>
!ubuntu
04:39
<ltspbot>
vagrantc: "ubuntu" is (#1) to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall, or (#2) a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to instal lltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:39
<vagrantc>
ltspbot: forget ubuntu 1
04:39
<ltspbot>
vagrantc: The operation succeeded.
04:40
<vagrantc>
!ubuntu
04:40
<ltspbot>
vagrantc: "ubuntu" is a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to instal lltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:40
<vagrantc>
much simpler
04:40
<ogra>
lltsp?
04:40
lazyltsp ?
04:40
<vagrantc>
ogra: ?
04:40
oh
04:40
<ogra>
to instal lltsp on ubuntu
04:40
<vagrantc>
ltspbot: forget ubuntu
04:40
<ltspbot>
vagrantc: The operation succeeded.
04:41cliebow has joined #ltsp
04:41
<vagrantc>
ltspbot: learn ubuntu as a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:41
<ltspbot>
vagrantc: The operation succeeded.
04:41
<vagrantc>
!ubuntu
04:41
<ltspbot>
vagrantc: "ubuntu" is a Debian based GNU/Linux distrubution that includes a Ubuntu specific version of LTSP. It can be found at http://www.ubuntulinux.org. to install ltsp on ubuntu: https://help.ubuntu.com/community/UbuntuLTSP/LTSPQuickInstall
04:42
<ChinaForge>
I read a lot of ltsp tech documents,but most is for 4.XX,so ...
04:42
<ogra>
yeah
04:42
<vagrantc>
ChinaForge: yes, it's in transition from ltsp 4 to ltsp 5 right now ...
04:42
<ogra>
ChinaForge, there are some more on https://help.ubuntu.com/community/UbuntuLTSP
04:42
<vagrantc>
ltspbot: factoids search --values
04:42
<ltspbot>
vagrantc: 'ltsp', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'wireless', 'sound', 'topics', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', 'js', 's', 'troubleshooting', 'sbnet', (1 more message)
04:43
<vagrantc>
do we really need both !sbalneav and !s ?
04:43
!troubleshooting
04:43
<ltspbot>
vagrantc: "troubleshooting" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/TroubleShooting
04:55
<ChinaForge>
bye!
04:55ChinaForge has quit IRC
04:56
<vagrantc>
!edubuntu
04:56
<ltspbot>
vagrantc: "edubuntu" is an Ubuntu variant especially tailored for schools. You can find it at http://www.edubuntu.org
05:00
<ogra>
hmm, it should mention that ltsp comes preinstalled and preconfigureed with it
05:01
and where is debian-edu ? :)
05:02
<vagrantc>
well, i'm not one to list each and every variant of a distro in the ltspbot
05:02
<sep>
vagrantc, can we all teach the bot about factoid's (like dpkg?)
05:02cliebow has quit IRC
05:02
<vagrantc>
sep: yup ... syntax is "learn FOO as BAR"
05:02
<sep>
ack
05:02* vagrantc wonders if there's a relearn option
05:03* vagrantc doesn't know dpkg
05:03
<ogra>
haha
05:03cliebow has joined #ltsp
05:04
<vagrantc>
well, not the bot, anyways :)
05:05
<ogra>
http://www.googlefight.com/index.php?lang=en_GB&word1=classmate+pc&word2=playmate+pc
05:05
hmm
05:06
i wouldnt have expected the classmate to win
05:06muh2000_ is now known as muh2000
05:11ogra-classmate has joined #ltsp
05:26cliebow has quit IRC
05:33ogra-classmate has left #ltsp
05:43cliebow has joined #ltsp
05:54bobby_C has joined #ltsp
05:59meduxa has quit IRC
06:04ccherret1 has joined #ltsp
06:05nobios has joined #ltsp
06:05cliebow has quit IRC
06:09a5benwillis has quit IRC
06:10ogra-classmate has joined #ltsp
06:16ccherrett has quit IRC
06:19Guaraldo has joined #ltsp
06:21cliebow has joined #ltsp
06:30ogra-classmate has quit IRC
06:47muh2000 has quit IRC
06:47jammcq has quit IRC
06:48cliebow has quit IRC
06:49nobios has quit IRC
07:00fernando1 has quit IRC
07:03Gianluca has joined #ltsp
07:04cliebow has joined #ltsp
07:07fernando1 has joined #ltsp
07:19envite has joined #ltsp
07:20
<envite>
Hola a todos desde España
07:20
Hello all from Spain
07:20
<vagrantc>
hola
07:20
<envite>
hola vagrantc
07:21
<vagrantc>
envite: so, in the canaries you're using both debian and ubuntu ? what decides which to do?
07:22Thumann has joined #ltsp
07:23
<Thumann>
hi, we've been running a ltsp setup for about a year now, it currently holds around 50 concurrent users and for improving performance, and adding fail over, we'd like to make a "farm" of two ltsp servers. From what i can read from the net, this can be done using OpenSSI.. but this seems kinda.. outdated to say it mildly?
07:24jammcq has joined #ltsp
07:24cliebow has quit IRC
07:24
<vagrantc>
jammcq: morning
07:25
Thumann: probably the simplest thing to do that will partially address it is to do some sort of load-balancing in such a way that it provides cold failover
07:25
<Thumann>
Or is there any other type of setup you would recommend?
07:25cliebow has joined #ltsp
07:26
<vagrantc>
cold failover being that if a server goes down while someone's using it, they'll have to log back in
07:26
<Thumann>
vagrantc: yes, load-balancing was acutally the "only" feature we would be needing.. and from what i can read on the net, openssi would hook you up with that.
07:26
vagrantc: and i'm familiar with the term cold faildover :)
07:27Gianluca has quit IRC
07:27
<Thumann>
What do you guys use for load-balancing?
07:27
If there is an alternative to openssi
07:28
<vagrantc>
only thing i've worked with is manually choosing a server at login. but you could easily randomize that.
07:28
more dynamic load balancing is tricky
07:28
<Thumann>
yeah well, that wouldn't really be "load balancing" or i'd have to get some scripts made that would measure cpu and memory use etc.
07:29
<vagrantc>
yes, it wouldn't balance any load at all :P
07:30
Thumann: if you want "True" load-balancing, there hasn't been much done in that department
07:30
<Thumann>
vagrantc: Isn't that a.. hard blow for ltsp?
07:31
compared to citrix, ms term srv. etc. who all have this feature
07:31
<vagrantc>
when a simpler option works in most cases, no
07:32
<Thumann>
vagrantc: Do you have any knowledge of openssi?
07:32
<vagrantc>
i've heard about it a little. someone had worked on using it with ltsp.
07:33
<Thumann>
http://wiki.openssi.org/go/Talk:Main_Page , hehe.. defaced :S
07:33
<envite>
vagrantc: sorry, i was fighting with kde kiosk mode
07:33jsgotangco has quit IRC
07:33
<Thumann>
or.. i wouldn't really call defacing a wikisite defacing.
07:34
<envite>
vagrantc: which is the meaning of the question? I do not understand it.
07:34
<jammcq>
g'morning kids
07:35
<vagrantc>
envite: well, when "meduxa" talked at the ubuntu education summit, he talked about using ubuntu ... and the other night you were talking about using debian ... so i was curious which gets used for which
07:36
<envite>
well, these are different projects
07:37
<vagrantc>
ah, i see.
07:37
<envite>
meduxa (Agustín Benito) talked in Sevilla about a work we did for Canary Islands govt. It is a educational distro based on Kubuntu and ha nothing to do with LTSP
07:37
<vagrantc>
got it
07:38
<envite>
besides that, we at the company use Debian, we install Debian to our customers (if they don't ask specifically for another thing) and we do LTSP 4.2 over Debian
07:39rjune-isis has joined #ltsp
07:41
<envite>
the project we did for the govt. is called mEDUXa and was based in Kubuntu because we wanted the integrated feeling it gives to the Desktop, instead of the Debian way. And, moreover, because we wanted a stable branch and Debian stable, at the moment, was quite old
07:42
<vagrantc>
envite: what's the biggest thing holding you back from switching to ltsp5 ?
07:44
<envite>
well, running projects already using ltsp 4.2
07:44
<vagrantc>
heh :)
07:46
<envite>
I want to start learning ltsp5 but we need to keep running projects without big changes
07:46
btw
07:46
is ltsp5 as good as 4.2 in sound and lda?
07:46
<vagrantc>
mostly i'd say yes.
07:46Thumann has left #ltsp
07:47
<vagrantc>
we didn't get ltspfs into debian/etch, but the ltspfs packages from lenny still work in etch
07:48mhterres has joined #ltsp
07:48
<envite>
use the same fuse system?
07:48
<vagrantc>
sort of
07:48
parts of it are very different
07:51
<envite>
i see
07:51
and about sound? same esd way?
07:51
<vagrantc>
esd in debian. ubuntu switched to pulseaudio, and i suspect debian will follow
07:52
<envite>
i've never touched pulseaudio
07:53
<vagrantc>
i think it's intended to replace esd
07:53
<envite>
i'll try when i'll start with ltsp 5
07:57vagrantc has quit IRC
08:01
<rjune-isis>
does pulse use the same protocol as esd?
08:06
<ogra>
no, it has an own native protocol
08:06
but it can emulate esd
08:07
even simultaneously
08:08
but what is more important, alsa has a plugin that emulates a native alsa device on top of the pulse network connection, so all apps using alsa and esd just work ootb
08:10
<rjune-isis>
that part is badass
08:11Shane-S has joined #ltsp
08:11
<ogra>
the only problematic stuff stays oss crap using /dev/dsp directly but thats getting less and less
08:12
<Shane-S>
is this where i can get help for k12ltsp.org's project as well?
08:15
<Lumiere>
yes to the best of people's ability
08:16
<Shane-S>
ok, as per a note I found on the k12ltsp site, it says for NIC computers edit /opt/ltsp/i386/etc for video res and sound, I have no clue what options I have for video
08:16
right now my systems boot, but when the video is initialized for the GUI either the refresh rate or resolution is too high because the monitors just blink there power lights
08:16
their*
08:19
<envite>
Shane-S: you need to touch the lts.conf file
08:19
see the ltsp wiki page about lts.conf
08:20
<Shane-S>
ok
08:26Blinny has joined #ltsp
08:30Guaraldo has quit IRC
08:33Guaraldo has joined #ltsp
08:33sbalneav has joined #ltsp
08:33
<sbalneav>
Morning all
08:34Guaraldo has left #ltsp
08:34
<envite>
morning
08:34
<Shane-S>
ok I read the wiki on Xserver stuff, it says it should detect the video card if it is not specified, will it auto detect the resolution or base it off the servers GUI settings?
08:34
<ogra>
!s
08:34
<ltspbot>
ogra: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
08:34
<sbalneav>
hey ogra!
08:34
I was doing some work last night on LDM
08:34
<ogra>
yay
08:34
<jammcq>
!s
08:34
<ltspbot>
jammcq: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
08:35
<ogra>
on the c or pyhon side ?
08:35
<sbalneav>
Gadi's not here, but rather than his 4 variables for autologin, I came up with an even easier solution.
08:35Guaraldo has joined #ltsp
08:35
<sbalneav>
Morning jammcq
08:35
<envite>
Shane-S: no
08:35
<sbalneav>
python.
08:35
So, check this out: when we grab the IP, we grab the MAC as well.
08:35
<envite>
Shane-S: lts.conf is used to create an Xorg.conf file
08:35
<ogra>
Shane-S, if you use a distro with ltsp5 it will autodetect
08:35
<Shane-S>
envite: the wiki does show more then resolution and an full example, if there a simple resolution and refresh rate line?
08:35
<sbalneav>
We add a new lts.conf var: LDM_GREETER
08:36
path to a greeter.
08:36
<ogra>
sbalneav, fine
08:36
<envite>
so if you do not specify it, Xorg.conf will not have it configured
08:36
<sbalneav>
when running the greeter, pass it the IP and and MAC
08:36
<ogra>
it has the ip already
08:36
<sbalneav>
A standard "ask user for password" greeter will ignore it.
08:36
<ogra>
there is a function called get_ip()
08:36
<envite>
Shane-S: you can specify x_vert_refresh simply, or create a whole modeline
08:37
<sbalneav>
Right, but it's not passed to the greeter command line.
08:37
<Shane-S>
ty
08:37
<ogra>
sbalneav, why do you need it on the commandline ?
08:37
<sbalneav>
pass it to the greeter. Then, if you want autologin, you just need a shell script that looks like
08:37
#!/bin/sh
08:37
<envite>
Shane-S: all info you give to lts.conf will be used to the Xorg.conf
08:37
<ogra>
sbalneav, did you look at gadis autologin stuff ?
08:37
<sbalneav>
if [ip = something] ; echo "userid\npassword\n"
08:38
I did, but he specifies the password in the lts.conf
08:38
<envite>
but all info you do not give, will be inferred or detected or created or defaulted by X
08:38
<sbalneav>
which is mode 644
08:38
<Shane-S>
ohh found on wiki, was further down, is it X_VERTREFRESH or X_vert_refresh?
08:38
<sbalneav>
this way, you could make the autologin greeter mode 700, and protect the passwords.
08:38
<ogra>
sbalneav, you can just copy an sshkey in place instead
08:38
<envite>
i don't remember just new, let me see
08:39
<sbalneav>
How you gonna map an sshkey to a mac?
08:39
<ogra>
not at all
08:39
but i know the server ip
08:39
what do you need a mac for ?
08:40
<sbalneav>
Lets say I want terminal with mac 00:11:22:33:44:55 to autologin as "user1"
08:40
<ogra>
yeah
08:41
i put the key for user1 in the chroot and set LDM_USERNAME in lts.conf for that TC
08:41
<envite>
Shane-S: X_VERTREFRESH
08:42
<sbalneav>
Now you're going to force admins to set up individual ssh keys for all their users.
08:42
in addition to their regular user accounts.
08:43
I suppose for most autologin setups, it would be a small number of WS's
08:43
<ogra>
right
08:43
and usually one user bound to each
08:43shogunx has quit IRC
08:44
<sbalneav>
you'll need an LDM_GREETER variable anyway, because you want to have different greeters, depending on GTK or KDE looks.
08:44
You'll see my patches, I'll upload my tree tomorrow.
08:45
Have you already merged gadi's patches into your tree?
08:46
<ogra>
i want the greeter to be autodetected by the installed dependencies in the chroot
08:46
(i didnt plan to install QT alongside the gtk greeter)
08:46
<sbalneav>
ok, but you'll want the user to be able to specify the greeter manually as well.
08:47
i.e. you've got both KDE and GTK greeters installed. User A wants GTK, cuz he's a gnome guy, User B wants KDE because she's KDE
08:47
<ogra>
nope
08:47
not at all
08:47
<sbalneav>
buh?
08:47muh2000 has joined #ltsp
08:48
<sbalneav>
Why not?
08:48
<ogra>
because that gets horribly complicated
08:48
<sbalneav>
Well, default to autodetect, surely, but allow the option of an override.
08:48
<ogra>
there will be a default greeter which is either qt or gtk
08:49
<sbalneav>
[00:11:22...]
08:49
<ogra>
and there will be an optional low_ressources greeter (if someone writes one)
08:49
<sbalneav>
LDM_GREETER = /usr/lib/ltsp/greeters/kde
08:49
Don't see what's so complicated about that.
08:49
<ogra>
the admin will specify that at chroot build time
08:50
it introduce extra layers we dont need imho
08:50
<sbalneav>
It's, like, 5 lines of code in LDM
08:50
<ogra>
hrm
08:50
these start to stack up
08:51
gadi had 10 lines as well
08:51
and i might have ideas for 20 lines
08:51
<sbalneav>
Wasn't aware we were on a strict line budget for LDM.
08:52
<ogra>
well, you ar the guys complaining about >120second boots
08:52
<sbalneav>
The slowdown's not in LDM
08:53
Not that I can see, anyway.
08:53
<ogra>
and the bigger we pump up the python side of thngs the harder it will get to switch to something else at some point
08:54
<sbalneav>
For the LDM, I don't see we NEED to switch to something else. The greeter was what was taking up all the time, not the LDM itself. It's quite small.
08:54
<ogra>
well, if i pretend to have an ltsp4.2 chroot here (by setting a static xorg.conf and cutting down modules and raing udev a bit) i get nearly on par with bottimes ... ldm then still takes more than 2secs
08:54
*20
08:55
<sbalneav>
Is that LDM with the python greeter, or with the C greeter.
08:55
?
08:55
<ogra>
moquist tried the C greeter and dint see any much difference in boottime on his slow clients btw
08:55
the C greeter
08:56
the major amount of time that we loose is still spent in the python part
08:56
<sbalneav>
OK, so we just need to rewrite LDM in C then.
08:56
<ogra>
well
08:56
*just*
08:56
:)
08:56
<sbalneav>
that's easy. It's really not doing anything that complicated.
08:56
Tell you what.
08:56
<Shane-S>
thanks guys that worked!
08:57
<ogra>
its handling passwords
08:57nicoAMG has joined #ltsp
08:57
<ogra>
it will take ages until we got full approval from the security team
08:57
an i have to run it past them for such a change
08:57
<Shane-S>
now I have a new question, I want to have a generic login, so I made a user called "student" and logged it in on one system...well the other warned I was logged in...
08:57
<sbalneav>
So, guess I better get busy then :)
08:57* sbalneav hacks.
08:58
<Shane-S>
well when I go to launch things like FireFox on both clients, one says FFx is open already
08:58
<ogra>
and indeed it makes contributing a *lot* harder
08:58
that were my two main cause to keep it in py until now
08:58
<sbalneav>
Not really. In C it really wouldn't be THAT much bigger.
08:58
<Shane-S>
how can I make a generic login that does not share a session??
08:58
<ogra>
Shane-S, yes, use one user per client
08:58
<sbalneav>
It doesn't need to be setuid, since it's already running as root.
08:59Q-FUNK has quit IRC
08:59
<Shane-S>
but I want to have a single user, not multiples
08:59
<sbalneav>
Essentially, it's job is just to: set up some environment variables, grab a userid and password from a greeter, and spawn an ssh session.
08:59
<ogra>
Shane-S, then you need to rewrite the world ... most applications and desktop environments cant handle that
09:00
<Shane-S>
alright
09:00
then is there a way to make it so I can use NULL passwords, it asks for 6 character minimum for users
09:00* ogra just had a freaky idea how it would be possible to use one user for all
09:01
<Shane-S>
since I dont want the kids to have to remember their logins
09:01
unless Fedora can import logins from an OS X server??
09:03
<ogra>
sbalneav, how about a pam-mount script that creates a tmpfs mount for /home/$user, that way we could use the same user on every login :) (indeed with the cost that all data is lost on logout)
09:03* highvoltage has done that before
09:04
<sbalneav>
Well, the mkhomdir code's already been done for that.
09:04
<ogra>
highvoltage, why do you keep such stuff secret :) thats awesome
09:04
<highvoltage>
I added a guest user home (/home/guest) to rw_dirs for diskless full machines
09:04
<ogra>
ah, no
09:04
<highvoltage>
ogra: I always asume everything I do is way to hacky for ubuntu
09:04
<ogra>
i'm talking about the server
09:04
<highvoltage>
aaah
09:05
<ogra>
make it create a tmpfs, mount /home/$user into it ... every new user gets a fresh home and no .lock files
09:05
i wonder how much dbus would get confused :)
09:06envite has quit IRC
09:08
<rjune-isis>
Shane-S: does your OS X server use LDAP? if so, then in theory you can use an ldap server with any linux distro
09:09
<Shane-S>
yeah it has LDAP, but I guess for now it will just be easier to search how to make linux users with no passwords
09:09
I am only testing it right now anyway
09:14mhterres has left #ltsp
09:15markgreene has left #ltsp
09:20Shane-S_ has joined #ltsp
09:22Shane-S_ has quit IRC
09:33vykarian has joined #ltsp
09:37Shane-S has quit IRC
09:38
<vykarian>
hi all
09:38
found the error "_X11TransSocketINETConnect Can't get address for localhost"
09:39
any tip?
09:41PMantis has joined #ltsp
09:42shogunx has joined #ltsp
09:42Guaraldo has left #ltsp
09:43
<vanya>
vykarian, check /etc/hosts for localhost record
09:44
<moquist>
ogra: interesting idea about the fresh /home/$user with no .lock files
09:44
I think FF's "You must close the other Firefox" message is one of the most amazingly terrible usability things in Ubuntu right now.
09:44
<vykarian>
thats message is on LTS client running freeNX client
09:44
<ogra>
well, it is right
09:45
it would break if two users use it at the same time
09:45
like openoffic
09:45
e
09:45
<moquist>
1. Usually (in my use cases) there is no other FF instance running. 2) The user has no idea how to figure out if there is one running or not, or where it is, or how to kill it. 3) The fix, 99% of the time in my use cases (LTSP), is to remove the lockfile(s), which the error message doesn't even mention.
09:46
It should *at least* give the option of using a different profile, which I believe it used to do.
09:46
<ogra>
indeed, that could be better
09:47vagrantc has joined #ltsp
09:48
<moquist>
I have a "firefox-profile-whack" script in the Applications->Internet menu that moves the user's ~/.mozilla directory over by appending a date to the name, creates a fresh .mozilla directory (from /etc/skel, IIRC), and copies the bookmarks back in.
09:48
ogra: People run into that problem *ALL THE TIME* on our Edubuntu networks. I guess I could've mentioned it before now...
09:50
I think people don't shut down properly b/c thin clients seem so resiliant, and this is a symptom of that. User education is a good start on addressing the problem, but surely at least the error message could be better at pointing the user toward likely realities.
09:51* vagrantc had to fight the ctrl-alt-backspace culture at freegeek
09:52
<vagrantc>
people kept doing that to "log out" and then wondered why their web browser would freak out once in a while
09:52ogra-classmate has joined #ltsp
09:53Q-FUNK has joined #ltsp
09:58
<ogra-classmate>
moquist: you coulkd add it to ldms session command
09:58
like we do with ltspfsmounter
09:58
look at the python code of 6the wrapper
09:58
<rjune-isis>
moquist: when We had that problem, we wrote a little program that let them reset firefox, OO.o, and a few other things to default configs.
09:59
<moquist>
rjune-isis: Right; I'm doing almost the same thing. We just haven't had so much trouble with OO.
09:59
Though resetting Gnome settings would be *really* valuable.
10:00bobby_C has quit IRC
10:00
<moquist>
Is there a recommended way to allow users to log into systems with different versions of Gnome, all while sharing one ~/ ?
10:00
<rjune-isis>
it was just a bash script using kdialog. no reason it couldn't be adapted
10:00
<ogra-classmate>
well, do you know the status of gnome-reset ? we talked about it plenty of times but nobody contacted upstream yet afaik
10:00
<rjune-isis>
use a different .gnome dir for each one would be my guess
10:00
aka ./gnome-<version>
10:00
<moquist>
ogra-classmate: gnome-reset is packaged in universe.
10:01
<ogra-classmate>
yes, but unmaintained afaik
10:01
<moquist>
rjune-isis: That's my guess, too. Not sure how to set that up.
10:01
ogra-classmate: Ah, OK.
10:01
ogra-classmate: I'm still interested in it (obviously), but it's lower in my priority stack than getting these particular TCs booting faster so the big cheese doesn't axe the edubuntu direction we're taking. 8-\
10:02
<vagrantc>
ouch
10:02
<moquist>
ogra-classmate: Though it's occurred to me that his big reason for saying we need to standardize on one distro was so that people in the training facility would see the same desktop in the schools... I say we develop our own Gnome theme, and then we can run whatever the hell we want under the covers.
10:02
<ogra-classmate>
did you try the right udev changes
10:03
<moquist>
ogra-classmate: Yeah; the udev changes bought 20 seconds to bring boot-to-login time down to 4:40 instead of 5 min.
10:03* ogra-classmate cant belive that ....
10:03
<moquist>
ogra-classmate: I still have the nbd& and static xorg.conf to try.
10:04
<ogra-classmate>
not waiting for udev should shove down more than 50%
10:04
<moquist>
ogra-classmate: It's *just* these laptops, of which we only have 12, out of probably more than 1000 TCs in the district. Surely the problem can be addressed. :)
10:04
<ogra-classmate>
udev and xorg autoconfig are the two big delays
10:05
well, there must be some kind of kernel issue
10:05
<moquist>
ogra-classmate: Yes.
10:05
<ogra-classmate>
probably acpi/apm related
10:05
<moquist>
ogra-classmate: Next on my list is changing the TC kernel config. Do you have any analysis methods to recommend other than just looking at the timestamps(right term?) in the TC syslog?
10:06
<ogra-classmate>
the kernel has only the bare minimum setup, i'd rather experiment ewith bootoptions
10:06
<moquist>
ogra-classmate: Sure. I'm happy to do so.
10:07
<ogra-classmate>
nolapic,noacpi, pci=noacpi, noapm
10:07
etc etc
10:07
<vagrantc>
acpi=off
10:07
<ogra-classmate>
i bet one of them solves it
10:07
or a combo
10:07
<vagrantc>
and lots of them look like typos
10:07
like pci=noapic
10:08
<ogra-classmate>
yeah
10:10vanya has quit IRC
10:10
<moquist>
ogra-classmate: here's the data I've collected so far: http://n01se.net/paste/VS2?pretty=yes
10:11
<ogra-classmate>
heh
10:11
oo.o only tkes foru secs
10:11
four
10:12
<moquist>
ogra-classmate: Yeah, it's on the gnome panel. In Edubuntu I have to wiggle the mouse-nub with my left hand to get into the menu, etc. My RH is ready to hit return as soon as it opens...it would be faster if I just moved my RH, or if I put a launcher on the panel... Hmm. :) That's not the crucial test, though.
10:13
ogra-classmate: I don't have numbers here for the 250Mhz 64MB netier TCs, but with Edubuntu those are ~2min to login, which I can sell (I think).
10:13
<ogra-classmate>
hmm, all toshiba ?
10:14
nolapic noacpi
10:14
<moquist>
ogra-classmate: I'm just working with this one slow laptop for now, until I get it booting reasonably. After that I'm going to tour all our schools with my testing setup and test every different type of TC, so when we (colleagues & boss) discuss this later we have all the data.
10:14
<ogra-classmate>
i'm oretty sure these two will solve it
10:14
pretty
10:15
<moquist>
ogra-classmate: Will give it a shot tomorrow and let you know. (Thanks.)
10:15irule has joined #ltsp
10:15Q-FUNK has joined #ltsp
10:15
<moquist>
ogra-classmate: any reason not to include those boot options in general for TCs?
10:16
<ogra-classmate>
yes, anything can be a thin client
10:16
and all newer models will work fine with acpi
10:16
<moquist>
ogra-classmate: K.
10:17
<ogra-classmate>
these toshibas are very likely from the time where the apm->acpi switch of the hw vendors took place
10:17
<moquist>
ogra-classmate: Yes.
10:17
<ogra-classmate>
i remember having some toshibas at that time where i needed to set these options
10:17* moquist wishes he had one at home to test
10:18
<ogra-classmate>
its a laptop, you can carry it?
10:18
<moquist>
ogra-classmate: Yeah, but I leave work at work (most of the time). I've got other responsibilities I'm supposed to be meeting now... and now I have no choice, b/c I cannot test the laptop. :)
10:19
<ogra-classmate>
go to your meeting then :)
10:24SlackDweber has joined #ltsp
10:24
<SlackDweber>
hii
10:25
excusme i have trouble about ltsp in client with pxe card
10:25
:(
10:25
<vagrantc>
what happens?
10:25
<SlackDweber>
i confused because on other pc with same pxe it can load n login on ltsp server :(
10:25J45p3r_ has joined #ltsp
10:26
<SlackDweber>
im try to installing ltsp 4.2u2 in zenwalk
10:26
n successfully configure
10:26
but on client, i got searching dhcp :(
10:27
<vykarian>
guys
10:27
XIO: Fatal IO error 104 (connectoin reset by peer) on X server ":0.0"
10:27
<SlackDweber>
note: in client pc i use compaq build up with pxe card
10:27
<vykarian>
any parametters needed?
10:29Egyptian[Home] has joined #ltsp
10:31
<SlackDweber>
ping
10:32
met malem om :)
10:34prpplague has joined #ltsp
10:38jsgotangco has joined #ltsp
10:48Blinny has quit IRC
10:56SlackDweber has quit IRC
10:59Q-FUNK has quit IRC
11:08PMantis has quit IRC
11:10
<sbalneav>
ogra-classmate: "ldm.c" [Modified] 253 lines --96%-- 245,30 97%
11:11
<vagrantc>
sbalneav: mornin
11:11
<sbalneav>
vagrantc: hello
11:11
vagrantc: ogra thinks ldm itself in python is slow.
11:12
So, it's getting re-written in c :)
11:12
<vagrantc>
so y'all are re-writing the whole thing in C ?
11:12
<sbalneav>
yep
11:12
<vagrantc>
so, while you're doing that
11:12
maybe this is a good time to think about putting a significant portion of the add-on features into other scripts
11:13
i.e. sound, localdev
11:14
<sbalneav>
Which, re-write them in C?
11:15
<vagrantc>
well, i think starting local device stuff shouldn't be handled by ldm, for example
11:16
oh, that stuff was in the ltsp-client init scripts
11:16
<sbalneav>
right
11:17kaminski-ltsp-br has quit IRC
11:17
<sbalneav>
esd's started, so we could move that out, and just deal with making sure the environment's set up correctly.
11:17
sure.
11:18
<vagrantc>
i also don't understand why we try to get the ip address ... that should be handled server-side and pulled from SSH_CLIENT
11:19
<sbalneav>
Probably could be.
11:19
I'll look into it.
11:19
<vagrantc>
same for setting all the detailed information for PULSE_SERVER, ESPEAKER, etc.
11:19
this would, of course, require the ldm-server package :)
11:20
<sbalneav>
The ssh command line's getting a little long. yeah, you're just echoing what i'm thinking
11:20
<vagrantc>
yup
11:20maaarek has joined #ltsp
11:20
<sbalneav>
what we need's a small shell script on the server that can be .'d, and sets up the env.
11:20
that's not that bad.
11:20
<vagrantc>
we can just drop files into /etc/X11/Xsession.d
11:20
<sbalneav>
that too.
11:21
hm
11:21
<vagrantc>
well, on debian-based systems
11:21
<sbalneav>
what if we have SOUND=n, LOCALDEV=n
11:21kaminski-ltsp-br has joined #ltsp
11:21* sbalneav thinks
11:21Burgundavia has joined #ltsp
11:21
<sbalneav>
meh, first, I need to get something working :)
11:21
<vagrantc>
heh
11:22
sbalneav: well, i guess what i'm saying is don't re-implement all the features- just get the basic premise functional first
11:22
<sbalneav>
right.
11:23
ATM, I'm just going after getting the X session plumbed.
11:24
<vagrantc>
sbalneav: you could re-use an existing ssh-askpass to handle the password side of things
11:25
sbalneav: are you trying to re-use the greeter?
11:25
<sbalneav>
Well, we're going to want a greeter of some kind, yes?
11:25
<vagrantc>
right
11:27
just trying to guage where you're going with this
11:27
<Burgundavia>
hey vagrantc
11:27
<vagrantc>
Burgundavia: greetings. how's the homeland treating you?
11:27
<Burgundavia>
not bad
11:28
<sbalneav>
lets aim to use something existing if we can.
11:28
<Burgundavia>
yesterday I went to work at 11:30, had a two hour lunch and then went home at 4
11:28
<ogra-classmate>
vagrantc: i asked mdz about server sided stuff and he nearly got angry at me back then
11:28
<vagrantc>
sbalneav: i guess, i'm questioning some of the existing assumptions
11:28
ogra-classmate: back when?
11:29
<ogra-classmate>
i dont think its a good ide to rely to much on that, it breaks teh default compatibility with standard ssh logins
11:29
<vagrantc>
sbalneav: and if there's a lot of re-writing going on...
11:29
<ogra-classmate>
vagrantc: when i started with the first sound implementation
11:29
<vagrantc>
ogra-classmate: what do you mean it breaks default compatibility with standard ssh logins?
11:29
<ogra-classmate>
we shouldnt clutter /etc/X11/Xsession.d
11:29
<sbalneav>
Well, it's wide open now. I'm going to hack on it most of today. Hopefully I'll have something going and in bzr later today.
11:30
<ogra-classmate>
vagrantc: its supposed to work on any setup with ssh -X capability
11:30
without extra server side stuff
11:31
<vagrantc>
ogra-classmate: well, if they handle all the variables, we won't be breaking anything
11:31
<ogra-classmate>
the sound setup for pulse already breaks that rule and i'm planning to fix that before gutsy
11:31
no server sided hacks if avoidable ....
11:31
<vagrantc>
yes, let's do the most hair-brained client-side setup possible instead.
11:32
<ogra-classmate>
no, lets just make more usage of our socket ;)
11:33
we should be able to pump all data we need through it without and server sided stuff
11:35
<vagrantc>
this is one of those moments where i think about starting up lessdisks again
11:35
<ogra-classmate>
everything you would determine from SSH_CONNECTION would require server sided scripting
11:35
<vagrantc>
and as far as i'm concerned, that's a good thing.
11:36
<ogra-classmate>
we already have the tunnel, lets just use it and keep the scripts on one side
11:36
whats wrong with cutting down mainteneance by 50%?
11:36
<vagrantc>
what are we really gaining by doing that?
11:37
<ogra-classmate>
you only need to maintain one side
11:37
all is centralized on the client
11:37
you dont need to patch around in desktop apps to respect variables they dont know
11:37
<vagrantc>
except if the client is executing commands on the server, it needs to somehow handle dependency resolution server-side
11:38
<ogra-classmate>
right, thats what the -standalone metapackage does
11:38
beyond that the scripts should have enough error handling to not break everything if an app is missing
11:39sepski has joined #ltsp
11:39
<vagrantc>
i guess, i'm just tired of a monolithic LDM that every time you add a new feature you have to hack ldm
11:39
ditto for the init scripts
11:39
we somewhat fixed that for ltsp-build-client
11:40
<ogra-classmate>
i'm fine with splitting them, i told you that before
11:40
i'm just opposing to do server side scripts as well
11:41
having a server sided script to determine ESPEAKER from SSH_CONNEFCTION just makes no sense imho
11:41
we should rather have ltsp-client-setup export the ip once in a global var so we have it in all subsequent initscripts
11:42
<vagrantc>
sure, that'd be an improvement.
11:42
<ogra-classmate>
i think we have the get_ip function three times in the bootprocess
11:42
two times in the initscripts and once in teh greeter
11:44
if we have an ltsp-client-sound package i see no reason not to use the ssh socket to execute asoundconfig pluse-enable on the server instead of having an ltsp-sound script in the session startup scripts
11:44
liekwise for everything else we need to set on the server side
11:45
i'm totally happy to modularize
11:47
but i'm all against to hack up the users session
11:48* vagrantc thinks some
11:48
<vagrantc>
ogra-classmate: i think the only play we use the IP address is ldm
11:49
<ogra-classmate>
PULSESERVER needs it and ESPEAKER ... and the localdev scripts
11:49
<vagrantc>
yes, i know
11:49
well, not the localdev scripts
11:50
<ogra-classmate>
the ltspfsmounter needs it, but it gets it from teh client side scripts
11:50
through the socket
11:51
we should also think about a fallback mechanism for systems without Xsession
11:51
<vagrantc>
or put it in a different location
11:51
<ogra-classmate>
i.e. rely on xterm being in path and falling back to at least start that (with a popup message)
11:52
put it in a different location ?
11:52
<vagrantc>
just discovered on centos it's in /etc/X11/xinit/Xsession
11:52
<ogra-classmate>
ah, well we could walk the known paths
11:52maaarek has quit IRC
11:53maaarek has joined #ltsp
11:53
<vagrantc>
ogra-classmate: that would really suck on errors
11:54
<ogra-classmate>
?
11:54
<vagrantc>
well, unless we test for the file not found error
11:54
<ogra-classmate>
if /etc/X11?Xsession doesnt exist we try again with /etc/X11/xinit/Xsession if that doesnt exist either we fail
11:55
<vagrantc>
how you going to do that client-side ?
11:55
<ogra-classmate>
checking stderr of ssh-client ;)
11:55
<vagrantc>
and what if the connection fails for other reasons?
11:56
anyways ...
11:58
<ogra-classmate>
ssh -S /tmp/.ltspfs_socket test -x /etc/X11/Xsession || echo False
11:59
easy test :)
11:59
<vagrantc>
so we create the socket before actually logging the user in?
11:59
<ogra-classmate>
we could have a control connection that we establish first
12:00Egyptian[Home] has quit IRC
12:00
<ogra-classmate>
it doesnt need to be through the socket
12:00
<vagrantc>
the slowest part of an ssh connection is the initialization ... so we don't want to start over
12:00prpplague has quit IRC
12:01
<jammcq>
can't we just install a script on the server that always goes in the same place? Then, the script would take care of finding the xsession file
12:01
<ogra-classmate>
how about creating the connection first and then attaching process by process to the socket
12:01
<vagrantc>
jammcq: but that would be installing something server-side!
12:02
jammcq: apparently, that's to be avoided at all costs
12:02
<jammcq>
umm, yeah, that's the point. do this step server-side, where the cpu is nice and fast
12:02
<ogra-classmate>
jammcq: not really, its very ugly to poke around in teh defaukts on teh server siide
12:02
<vagrantc>
jammcq: i'm with you here
12:02
<jammcq>
how is this "poking around in the defaults" ?
12:03
<vagrantc>
all this code will be infinitely simpler to write server-side.
12:03
<ogra-classmate>
/etc/X11/Xsessio.d is the system X defaukts dir
12:03
\its not intended to be used like we do all teh time
12:03
<vagrantc>
it will also make it work for both XDMCP and ssh based logins.
12:03
<jammcq>
did I say it should go in /etc/X11/Xsession.d ?
12:03
<ogra-classmate>
where elsed would you put it ...
12:03
<vagrantc>
/usr/bin/ldm-session
12:03
<jammcq>
umm, how about /usr/bin ?
12:03
<ogra-classmate>
bah
12:04
so the great feature of 'log in to any ssh server' is briken then ?
12:04
*broken
12:04
<vagrantc>
that great feature never existed
12:04
<jammcq>
"great feature" ?
12:04
hmm, /me never saw that spec
12:04
<ogra-classmate>
it always did
12:04
i do it all the time here
12:04
<vagrantc>
log into any ssh server that has all the files in the right places
12:05
<ogra-classmate>
thats the initital design i took over from mdz
12:05
<vagrantc>
+ as long as the ssh keys are known to the thin client
12:05
<jammcq>
umm, was that really a design criteria ? or was did it "just so happen to work" ?
12:06
<ogra-classmate>
he was very clear that nothing has to be changed server side to not break anything of the defaults and to not depend on extra software
12:06
<jammcq>
hmm, 'was did' ? dunno what I said there
12:06
<ogra-classmate>
right
12:06
<jammcq>
so now he's god, eh?
12:06
<ogra-classmate>
log in from any ubuntu TC to any ubuntu system with ssh server
12:06
<jammcq>
sorry, couldn't resist poking :)
12:07
I'd add: "As long as the server has been configured to allow that"
12:07
<ogra-classmate>
jammcq: nope, but matt had a reason for all decisions he made in teh initial design ...
12:07
<vagrantc>
+ x11-common + a window manager of some form or fashion
12:07
<jammcq>
some of those reasons were formulated in a vacuum
12:08
<ogra-classmate>
imho adding server sided deps is as bad as adding and removing stuff in user homedirs
12:08
<vagrantc>
why?
12:08
<jammcq>
so to fix it, we'll depend on some goofy trial and error between the client and the server every time a user logs in.
12:08
<ogra-classmate>
if you force me to accept it i will have no choice but i would expect us to do more proper solutions
12:09
<jammcq>
this isn't about forcing. this is about discussing possibly solutions, to decide what makes sense
12:09
<vagrantc>
at the moment, i think we've got some pretty improper solutions.
12:09
<jammcq>
agreed
12:09
we just need to find the lesser of the evils
12:09
<vagrantc>
and i'd like to fix that.
12:09
<ogra-classmate>
well, you wont get me into the add more deps and scripts to teh server side camp, imho we do already way to much on that side
12:10
<vagrantc>
what is so wrong with that?
12:10
<jammcq>
I think there's a certain set of things that need to be installed server-side, just to expect a thin client to work.
12:10
above and beyond the whole chroot thing
12:11
LTSP is becoming an "API". There needs to be a set of services and scripts available on the server, in a known location, for the thin client to connect
12:11
<ogra-classmate>
vagrantc: you tweak defaults (/etc/X11/Xsession.d), you override system settings /etc/ltsp introduce open ports (all the inetd stuff) etc etc thats just as wrong as poking around in userdata
12:12meduxa has joined #ltsp
12:12
<vagrantc>
ogra-classmate: i'll not that all that stuff was added by you, not me :P
12:12
note
12:12
<ogra-classmate>
especially since you always need to make sure your tweaks dont clash wint any other system sesttings probably used by a desktop env you have never seen
12:12
vagrantc: i didnt say it was right to add
12:13
<jammcq>
if you dont' want to "open ports, override system things, etc", then possibly the right approach is more like a RDP, ICA, NX way, where ALL communication is served by a single protocol/daemon
12:13
<ogra-classmate>
i think i'm saying since 30 mi9n i want tio get rid of it for more proper solutions ;)
12:13
and dont want you to introduce more if the evil :)
12:14
<dan__t>
'morning.
12:14
<vagrantc>
ok. so in what way does setitng stuff client-side prevent clashing with other system settings?
12:14
<ogra-classmate>
we have a proper mechanism in place we dont fully use yet
12:14* jammcq is lost
12:14
<ogra-classmate>
it will only happen from a thin client login
12:14
<jammcq>
dunno what "proper mechanism" is being refered to
12:15
<ogra-classmate>
jammcq: ssh and its capabilities, we dont use 10% of what we have
12:15
<jammcq>
so fine, use ssh where it shines, but I don't think it makes sense to use ssh to go searching for a session script
12:16
<ogra-classmate>
sure it does
12:16
why wouldnt it
12:16
<jammcq>
you say tomato, I say tomato
12:16
<ogra-classmate>
the greeter is the place wher we want that info
12:16
<jammcq>
it shouldn't because it's EXPENSIVE to do that
12:16
<ogra-classmate>
??
12:17
its not more expensive than what we do now, not a byte
12:17
well, probably a byute or two :)
12:18
<vagrantc>
right now, it's essentially a hard-coded default. opening another ssh connection would slow login down considerably.
12:18
<jammcq>
if ! ssh host /etc/X11/Xsession then if ! ssh host /etc/X11/xession then if ! ssh host /usr/bin/start_my_session then if ! ....
12:18
<ogra-classmate>
heh, no
12:18
<jammcq>
heh, ok
12:18
<ogra-classmate>
i'm taliking about changing the general design here ...
12:19
<jammcq>
I guess I haven't seen that talking yet
12:19
<ogra-classmate>
you establish a session without any X
12:19
then you have your socket
12:19
<vagrantc>
(at which point you'll need to establish a second session with X)
12:19
<ogra-classmate>
now you can execute anything you like through that socket, no extra costs
12:19
<jammcq>
like how?
12:19Guaraldo has joined #ltsp
12:20
<jammcq>
how do I say "if [ -x /etc/X11/Xsession ]" over a socket ?
12:20
<ogra-classmate>
vagrantc: why ? ssh -S <socket> Xsession will just start the session in the existing tunnel
12:20
jammcq: you put ssh -S <socketname> in front
12:20
<jammcq>
so invoking an additional 'ssh .... 'doesn't cost anything? try telling the e2300 that
12:21
<ogra-classmate>
an additional ssh ?
12:21
<jammcq>
maybe it doesn't create a new socket
12:21
<ogra-classmate>
no, all is done through one socket in one place
12:21
<jammcq>
but it's still a binary that has to be loaded and run
12:21
I understand the "communication" goes over the same channel
12:21
<ogra-classmate>
its like it is now, just way more effectively used
12:21
<jammcq>
but there's still a cost ty running 'ssh -S <socketnname>' it doesn't come for free
12:22
<ogra-classmate>
no, indeed, but you are executing something that is already ion ram and hands over to the existing process after invoking
12:22
thats marginal
12:22
<jammcq>
we can debate that mirginal-ness all day, i'm not sure who's right there
12:23
there may be other things you are trying to solve with this <socketname> thing, I'm just thinking that finding the session script isn't one of them
12:23
<ogra-classmate>
well, i know it only attaches the server side process to the socket for IO and it wont live longer than that in my processlist
12:25
i think having a non Xsession connection established first and then using the socket here is the best thing we could do ...
12:26
<jammcq>
this really all boils down to who is gonna do the work. at this point, you are more likely to do it than I am, so naturally, you'll do it the way you think is best. I've said my opinions about where it should be done. the balls in your court
12:26
<ogra-classmate>
even better would beto build that with a bidirectional socket
12:27
that way we could redirectpower events ;)
12:28* jammcq remembers several people saying we set the bar too high with ltsp-4.2. apparently we did something right
12:28
<vagrantc>
ok
12:28
"ssh -S <socket> -X" doesn't work after the fact
12:29
<ogra-classmate>
jammcq: well, actually the ball is more in vagrants court here :)
12:29
<jammcq>
heh
12:29Burgundavia has quit IRC
12:29
<jammcq>
we could all be in for a surprise
12:29
<ogra-classmate>
we scared Burgundavia already :)
12:30
vagrantc: thats how ldm invokes it atm
12:30
<jammcq>
ogra-classmate: did you get that wireless driver for the classmate?
12:30
<ogra-classmate>
it could be that the socket file needs to exist or something
12:30
<Lumiere>
ogra-classmate: if I wanted to start a discussion about when edubuntu releases... where should I go?
12:30
<vagrantc>
ogra-classmate: no, it's invoked -M -S <socket>
12:30
<ogra-classmate>
jammcq: we got something two days ago, baenc is testing today
12:30
looks like a firmware issue
12:31
<jammcq>
cool, so it's Opensource, and hopefully going into the kernel?
12:31
<ogra-classmate>
vagrantc: ohy, indeed, you need to create a master socket
12:31
<vagrantc>
ogra-classmate: you have to enable the -X when you create the master socket, according to my quick experiment.
12:31
<ogra-classmate>
jammcq: if i understood ben right we already have everything but the firmware ... and the driver uses an hardcoded path for searching the frimware
12:32
the firmware is free and shippable
12:32
only the searchpath needs to be fioxed
12:32
<jammcq>
cool.
12:32
<ogra-classmate>
vagrantc: yes thats how i planned to do it
12:33
vagrantc: but dont start an Xsession
12:33
<vagrantc>
ogra-classmate: right ...
12:33
<ogra-classmate>
if the tunnel is established we start to do all wotrk through the socket
12:33
and start the found Xsession sriot from there
12:33
script
12:34
<vagrantc>
still going to be a hell of a lot more overhead than a server-side script.
12:34
<ogra-classmate>
that has the advantage that we can hand over global vars etc easily
12:34
but we have the proper communication going so errors can be forwarded, smartcdards will become an option etc
12:35
and i dont think its that much overhead
12:35
some ram for the second ssh binary
12:35
mosto of it should be shared with the first one running anyway
12:36
i will try to write a proof of concept if i'm done with the classmate
12:38
(which will surely still take some days ... i'm just starting to write an installer that operates from initramfs, punches the partitions on the flash and untars the system image)
12:40vagrantc has quit IRC
12:53bengoa has joined #ltsp
12:53muh2000 has quit IRC
12:54
<bengoa>
jammcq, happy birthday buddy
12:54
<jammcq>
heh, thanks bengoa
12:54
<bengoa>
jammcq, :-)
12:55
<ogra-classmate>
jammcq: oh
12:55
ahppy birthday
12:55
or happy :)
12:55
<jammcq>
gee thanks ogra-classmate
12:57
<mistik1>
For he's a jolly good fellow, for he's a jolly fellow, for he's a jolly good fellooooooooooooow, which nobody can deny.
12:57
Happy Birthday Jim
12:58Avatara has quit IRC
12:58
<jammcq>
mistik1: thanks, i'm getting tears in my eyes now
12:58
<mistik1>
;-)
12:59
<Guaraldo>
hi, jammcq... Happy Birthday to you!!!
13:00
<jammcq>
jeez, they're crawling out of the woodwork
13:00
<mistik1>
You're lucky we're not living closer, the Red Stripe would be on me today
13:01
<whiprush>
jammcq: happy birthday dude
13:01
<cliebow_>
yay!!!
13:02
<jammcq>
yeah, finally hit 29
13:02
<cliebow_>
heh
13:03sepski has quit IRC
13:04vdittgen has joined #LTSP
13:05vdittgen has quit IRC
13:05a5benwillis has joined #ltsp
13:09spectra has joined #ltsp
13:10maximi89 has joined #ltsp
13:10
<maximi89>
hi
13:10
recently i release to LTSP 5.0
13:10
few minuts ago i was using 4.2
13:10
but now with all changes i have a problem with sh: /etc/rc.early_sysinit: nout found  
13:11
about that file /etc/rc.early_sysinit
13:11
what mean? that is because i dont have the tftpboot correctly? vmlinuz
13:12
/tftpboot/lts/2.6.16.1-ltsp-2
13:12
or /tftpboot/lts/2.6.17.8-ltsp-1
13:12
??
13:14
<jammcq>
maximi89: sounds like you are using LTSP-5 with a ltsp-4.2 kernel
13:15gepatino has joined #ltsp
13:15
<maximi89>
uhmm how can i fix this problem?
13:15
i don't understand at all this problem
13:16
<jammcq>
look at your /etc/dhcpd.conf file, and see what 'filename' is set to
13:16
it might be in /etc/ltsp/dhcpd.conf
13:16
<maximi89>
filename "/tftpboot/lts/vmlinuz-2.6.16.1-ltsp-2";
13:16
<jammcq>
yeah, that's wrong
13:16
what distro?
13:16
<maximi89>
ok only like this:
13:16
filename "/etc/ltsp/dhcpd.conf";
13:16
Debian 4.0
13:17
uhmm?
13:17
where is that file /etc/ltsp/dhcpd.conf?
13:18
ltsp i think can be in my case ltsp-5.0
13:18
<jammcq>
you'll need to change it to '/var/lib/tftpboot/ltsp/i386/pxelinux.0'
13:18
but that's the PXE kernel
13:18
I think you might need an Etherboot kernel
13:18
<maximi89>
yes
13:18
<jammcq>
so, go look in /var/lib/tftpboot/ltsp/i386 and tell me what you see there
13:19
<maximi89>
ok
13:20
no, i don't have that directory
13:20
i have /tftboot/
13:20
<jammcq>
hmm
13:20
hmm, looks like we lost vagrantc, the guru in debian/ltsp
13:21
<maximi89>
xD
13:21
"que mala onda!"
13:23
http://pastebin.ca/507707
13:24
that is the /etc/dhcpd.conf
13:24
<jammcq>
maximi89: you installed from the tarball ?
13:24
<maximi89>
http://pastebin.ca/507715
13:25
uhmm the installers come from a community called "EducaLibre" in www.educalibre.cl
13:25
<jammcq>
hmm, maybe you need to talk to them.
13:31privet has quit IRC
13:35SBNet has joined #ltsp
13:35bobby_C has joined #ltsp
13:41joebaker has joined #ltsp
13:45meduxa has quit IRC
13:46
<joebaker>
ogra: Remember the parallel port printer problem I had under Ubuntu 6.10 LTSP? It's working ok now under 7.04. THANK YOU!
13:47
<ogra-classmate>
yay, great :)
13:48* ogra-classmate would love to fix serial printers this round, but doent know anyone owning one
13:49
<ogra-classmate>
maximi89: if you use debian you should use debians ltsp, not a third party one
13:49
!debian
13:49
<ltspbot>
ogra-classmate: "debian" is is a GNU/Linux based operating system that makes an excellent LTSP server. You can find it at http://www.debian.org. for information about LTSP on debian see http://wiki.debian.org/LTSP
13:51
<maximi89>
thanks guys
13:52
<dan__t>
Aaaalrighty. Anyone know how I might be able to change the desktop background on a client?
13:52
Using gnome
13:52
<jammcq>
sure, right click on the desktop, and select 'Change desktop background'
13:53
<dan__t>
Well, remotely heheh.
13:53
Through a config file or gconf2 maybe
13:54
<ogra-classmate>
in recent gnomes you should be able to drop a config file in /usr/share/gconf/defaults/
13:55
if you pick a sequence number around 25 your setting will override the system default
13:55gepatino has quit IRC
13:55
<dan__t>
sequence number?
13:56
<ogra-classmate>
look i9n that dir, your diswtro should have put some defqaults in there already
13:56
<dan__t>
ok.
13:56
thanks
13:56
<ogra-classmate>
you want to set /desktop/gnome/background/picture_filename
13:58
<jammcq>
ogra-classmate: is that a pathname? or a gconf key?
13:58
<ogra-classmate>
pathname
13:58
err
13:58
a gconf key for a pathname variable *g*
13:58
<dan__t>
That path is relative to what?
13:59
<ogra-classmate>
teacher@edubuntu:~$ grep background /usr/share/gconf/defaults/20-edubuntu
13:59
/desktop/gnome/background/picture_filename /usr/share/backgrounds/edubuntu_default.png
13:59
<dan__t>
ahh
13:59
<ogra-classmate>
make a file called /usr/share/gconf/defaults/25-dans-custom-artwork
13:59
and add a line like the above
14:00joebaker has quit IRC
14:00
<dan__t>
Ah, I just want it to change for a single client.
14:00
I would imagine I'd find a similar structure in the user's homedir huh
14:00
<ogra-classmate>
client or user ?
14:00
<dan__t>
user, actually
14:00
Here, they are identified as one and the same.
14:01monteslu has quit IRC
14:02
<ogra-classmate>
as root: sudo -u $USERNAME gconftool-2 -s -t string /desktop/gnome/background/picture_filename /path/to/tour/file
14:02maximi89 has quit IRC
14:02
<ogra-classmate>
s/tour/your/
14:02
i will get used to this tiy keyboard one day, damned
14:03
gah
14:03
*tiny
14:03
<dan__t>
haha
14:03
Ok, where is that data stored?
14:03
<ogra-classmate>
in the gcxonf database of the user
14:03
gconf
14:04
in ~/.gconf or so
14:04
<dan__t>
~/.gconf?
14:04
Alright, cool. Let me give that a shot.
14:04
Thank you btw.
14:04
How did you know which key was the one to be changed?
14:04
<ogra-classmate>
dont touch that manually, use gconftool
14:04
<whiprush>
http://www.gnome.org/learn/admin-guide/latest/gconf-9.html
14:04
follow that
14:04
<dan__t>
Thanks.
14:04
<ogra-classmate>
well, i had to change it when building edubuntu ;)
14:05
<dan__t>
No, I just wanted to see what all was available to me heh
14:05
<ogra-classmate>
to see the available keys gconf-editor is handy btw
14:05
<dan__t>
And gconftool-2 knows enough to only attempt setting that BG for a single client, ogra-classmate?
14:06
<ogra-classmate>
if you sudo -u username it will run as the desired user and has no write access to anything else than ~/
14:07
just use the command i pasted above and replace $USERNAME with the name of the user and give a valid filename
14:07
<dan__t>
Well, yeah - just didn't know how it worked.
14:08
Thanks, I appreciate it.
14:08
<ogra-classmate>
:)
14:09
<dan__t>
One more thing here then
14:09vagrantc has joined #ltsp
14:09
<dan__t>
Can I force a client to log out
14:11
<Lumiere>
kill his or her wm session
14:12
<ogra-classmate>
or use a tool like thin client manager (ltsp5 ) or teachertool (ltsp4)
14:15joebaker has joined #ltsp
14:29vykarian has quit IRC
14:43cliebow_ has quit IRC
14:46* kaminski-ltsp-br is away: ohh yeahh
14:47jsgotangco has quit IRC
14:49jsgotangco has joined #ltsp
14:59
<ogra-classmate>
vagrantc: in case you are intrested, there is the distro team meeting in #ubuntu-meeting starting now
15:03* vagrantc is too weary
15:04
<ogra-classmate>
well, thats your actual future team :)
15:08irule has quit IRC
15:08a5benwillis has quit IRC
15:19bobby_C has quit IRC
15:19bobby__C has joined #ltsp
15:21sepski has joined #ltsp
15:22gentgeen__ has joined #ltsp
15:28
<lns>
hey does this channel get logged anywhere online?
15:28
I'm looking through it and there's a TON of invaluable information here that I bet a lot of people could benefit from =)
15:29
<sbalneav>
I log the channel, but the logs aren't posted anywhere, at the moment.
15:30
<lns>
well at least they're getting logged somehwere. =)
15:39
<jammcq>
http://blog.washingtonpost.com/fasterforward/2007/05/dell_does_linux.html
15:39
congrats and thank you to everyone who made that possible
15:42
<ogra-classmate>
jammcq: there is a ditro team meeting going on, you could thank them directly ;)
15:43vagrantc has quit IRC
15:45sbalneav has quit IRC
15:54SBNet has quit IRC
15:58maaarek has quit IRC
16:14J45p3r_ has quit IRC
16:18jammcq has quit IRC
16:22bengoa has quit IRC
16:36Guaraldo has left #ltsp
16:37
<cliebow>
good press there1!!
16:58Burgwork has joined #ltsp
17:04highvolt1ge has joined #ltsp
17:05highvoltage has quit IRC
17:12sepski has quit IRC
17:12bobby__C has quit IRC
17:19bricode has joined #ltsp
17:32nicoAMG has quit IRC
17:38muh2000 has joined #ltsp
17:39bricode has quit IRC
17:46happywithed has joined #ltsp
17:46
<happywithed>
Hello there
17:47
Can anyone direct me to the documentation need to read to enable the USB ports on my thin clients?
17:53
Anyone?
17:53
I am using Edubuntu Feisty
17:55
<Burgwork>
in what sense?
18:02
<happywithed>
I would like the students to be able to save files to their USB sticks
18:02
<Burgwork>
ahh, local devices
18:04
<happywithed>
yep
18:04
I am reading several places now and I think it has to be done in the ltps.conf file? right?
18:11
OK, I got it. I think I need to add the line RCFILE_01 = usb to lts.conf
18:24ccherret1 is now known as ccherrett
18:36happywithed has quit IRC
18:37lns_ has joined #ltsp
18:37
<lns_>
ugh...this is driving me crazy
18:39
I've got an HP Proliant ML370 G5 server - dual quadcore 3.0ghz cpu, 8.0gb ram running 2.6.20-15-server and different types of thin clients (neoware e100 & older compaq ipaq workstations)
18:39
i can easily get the thin clients to crash (hard-lock, no mouse movement) if i do things like log out with a bunch of applications open
18:40
it seems like the video chipsets are being detected correctly, as there are two different types of thin clients (thus video chipsets) but they behave the same
18:41a5benwillis has joined #ltsp
18:41
<lns_>
ubuntu Feisty on the server btw
18:42
w/ltsp5 installed
18:42
and 2.6.20-15-386 kernel on the chroot
18:43
can anyone give me any clues as to investigation?
18:49
also, i get this weird dotted line toward the upper-left corner of the screen when X is starting
18:49
almost like video corruption..but but that only seems to be on the neoware boxes
18:50ogra-classmate has quit IRC
19:05lns_ has quit IRC
19:09sbalneav_ has joined #ltsp
19:09
<sbalneav_>
Evening all
19:09sbalneav_ is now known as sbalneav
19:30youngun has joined #ltsp
19:31
<youngun>
is therre someon who can assit me in sound capture problems
19:32
anyone there
19:33
<sbalneav>
youngun: Hello, what can I help you with?
19:33
<youngun>
i have ltsp 5 (standalone)
19:33
i have redirected the sound
19:33
and when a client logs in it sounds off gud
19:34
but i can't get the web browsers to create sound like movie/news to work
19:34
ugh
19:34
I have downloaded flash
19:34
but still no sound
19:34
<sbalneav>
What distro are you running?
19:35
<youngun>
ubuntu ultimate
19:35
<sbalneav>
Which version?
19:35
<youngun>
7.04
19:35
<sbalneav>
So you're running feisty fawn?
19:35
<youngun>
with firefox 2.03 (which locks up on the clients)
19:35
yes
19:36
<sbalneav>
Well, check and see if the pulseaudio alsa redirector is loaded.
19:36
<youngun>
how?
19:37
<sbalneav>
lsmod should tell you. Paste the results to the pastebot
19:37
!pastebot
19:37
<ltspbot>
sbalneav: "pastebot" is The LTSP pastebot is at http://pastebot.ltsp.org. Please paste all text longer than a line or two to the pastebot, as it helps to reduce traffic in the channel. A link to the content will be pasted in the channel.
19:38
<youngun>
not found
19:38
my bad
19:38
<sbalneav>
What's not found?
19:39
<youngun>
i missed your first line i got the results
19:39
how do i paste to the pastemod
19:39
<sbalneav>
paste the results to http://pastebot.ltsp.org
19:40
<ltsppbot>
"youngun" pasted "results for lsmod" (111 lines) at http://pastebot.ltsp.org/142
19:40
<youngun>
k
19:40
they are there
19:41jsgotangco has quit IRC
19:41
<sbalneav>
Hmm, don't see anything wrong off the top of my head.
19:42
Sound's not really my specialty, you might want to try the ltsp-discuss mailing list.
19:42
<youngun>
all other programs work ie rythmbox cept the web browsers
19:42
k
19:43
<sbalneav>
Well, everything else uses a proper audio stack, i.e. gstreamer.
19:43
<youngun>
k
19:43
<sbalneav>
flash... kinda does it's own thing.
19:43
<youngun>
k
19:43
so no real gud way to fix it huh
19:43
<sbalneav>
Next version of Ubuntu will probably have one of the Free flashes in it, either gnash or libswfdec, and they'll work better anyway.
19:44
<youngun>
well should i uninstall flash and go with one of them?
19:44
<sbalneav>
They're not official packages yet, so getting them going can be a bit of a chore.
19:45
<youngun>
gnash is in package manager
19:45
<sbalneav>
I'd post to the list, and see if anyone's got an idea.
19:45
<youngun>
k
19:45
well thanks for your help
19:45youngun has left #ltsp
19:45
<sbalneav>
As well, pop back during the day, and look for either ogra or vagrantc, they'll have a better idea.
19:48Burgundavia has joined #ltsp
20:12jammcq has joined #ltsp
20:12
<jammcq>
hello
20:14
<sbalneav>
hey
20:14
Well, I've already got ldm to the point of setting up the xauth key, spawning X, and launching the greeter.
20:14
Now I just have to plumb the ssh connection :)
20:16
<cliebow>
you rock!!
20:17
<sbalneav>
wc ldm.c: 266 654 5820 ldm.c
20:25
ls
20:49
<mistik1>
sup fellas
20:53
<jammcq>
just watching the game
20:53
<mistik1>
I had forgotten about the game
20:53
I was having a Red Stripe for you Jim
21:22ogra has quit IRC
21:23ogra has joined #ltsp
21:33
<mistik1>
jammcq: I think your boys gonna win this game, for one thing Cleveland still has not learned to move without the ball
21:34
<jammcq>
heh
21:34
<mistik1>
Untill they learn that Labron will be useless
21:34
<jammcq>
pistons do their best in the 4th qtr
21:35
<mistik1>
We know about the detroit defense, but the point is to find the way to offset that
21:35
The only way I know is to MOVE without the ball
21:35
joke
21:36
This may well be a sweep
21:36
<jammcq>
hmm, cleveland is putting up a very tough fight
21:36
<mistik1>
yea, Just not a very smart one
21:39
I promise one thing, If Cle. cant move and space the floor they wont win a single game against Det.
21:40
As long as they stand there Detroit will just cut off all the passing lanes and then calapse 2 or 3 guys on the guy with the ball
21:45dan__t has quit IRC
21:49highvolt1ge has quit IRC
21:53highvoltage has joined #ltsp
21:58
<jammcq>
heh
22:02
<mistik1>
jammcq: trust me ;)
22:03
<jammcq>
yeah, but you gotta admit, that was edge-of-your-seat bball
22:04
<mistik1>
for sure
22:04
I also know that the pistons dont care if they win by 1 or 40
22:04
Its all a mental game to those guys
22:05
That's why I say Cleveland with thier youth should be running all over the place
22:05
That is the only thing that I see will give Detroit problems
23:21kaos01 has quit IRC
23:21kaos01 has joined #ltsp