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


Channel log from 8 December 2008   (all times are UTC)

00:05
<warren>
Ryan52: could you implement an optional "System default dmrc", if defined, use that, if not otherwised defined in the user's dmrc?
00:05
Ryan52: not everyone needs it, but I do. I currently have NO WAY of setting a system default that is non-GNOME.
00:07
<Ryan52>
hrm. doesn't ldm have some way of setting a default session? 0.o
00:08
anyway, ya, it'll be easy to do that.
00:10
<warren>
Ryan52: dmrc can contain language and session?
00:10
<Ryan52>
hrm. I dunno...
00:10* Ryan52 wonders if there's some specification for this
00:10
<Ryan52>
I was just doing session, and basing it off of what kdm put in my .dmrc..
00:13
<johnny>
hello warren
00:13
and Ryan52
00:14
<Ryan52>
hiya johnny
00:14
<johnny>
Ryan52, so i guess you know my friend joe..
00:14
or at least have met him
00:15
i forced him to finally get involved there a bit
00:15
<Ryan52>
who? what? huh??? :)
00:17* Ryan52 wonders who this joe person is...
00:18
<johnny>
he does tech support stuff
00:18
afaict
00:18
i guess he's not very vocal
00:18F-GT has joined #ltsp
00:18
<F-GT>
ONLY $319 AFTER $59 CASHBACK!
00:18
NEXT DAY DISPATCH
00:18
Acer Aspire One Linux Edition
00:18
Intel Atom Processor, 8.9” LED Backlit Screen
00:18
Ultraportable, Light and Compact. Weighs just 985 grams, under 3cm thick! Only available in Blue Colour.
00:18
doh
00:18
wrong chan
00:19
<Ryan52>
johnny, ohhhhhhh. I hide from tech support :)
00:19* Ryan52 might know who he is..
00:20
<johnny>
he knows his stuff about linux.. but he's not a programmer yet
00:20
<warren>
Ryan52: this is what my ~/.dmrc has
00:20
[Desktop]
00:20
Session=gnome
00:20
Language=en_US.utf8
00:20
Layout=us
00:21
keyboard layout too it seems
00:21
<Ryan52>
warren: oh, ok. I will do Language too. thanks.
00:21
<warren>
yeah, keyboard layout is HARDE
00:21
HARD
00:22catalystduck has joined #ltsp
00:22
<warren>
Ryan52: will your implementation handle writing a new ~/.dmrc if they choose something different?
00:23
<Ryan52>
yup
00:24
<warren>
Ryan52: will it ask if they want to save that as the new default?
00:24
<Ryan52>
yes.
00:24
<catalystduck>
anyone think they can help with a client display issue?
00:25
<warren>
Ryan52: if you use the exact strings as gdm or kdm, you could copy all of their translations for that string as well, making your job easier.
00:25
<johnny>
don't ask to ask..
00:25
<warren>
Ryan52: check out gdm, I suspect it has more translations
00:25
<johnny>
just ask your question catalystduck
00:26dirigeant is now known as MeW_away
00:27
<catalystduck>
ok I have a client computer formerly installed with ubuntu; I have the original xorg.conf from that installation and am trying to use it to configure the ltsp client
00:31
can I just use this file to configure the client's display?
00:35
<johnny>
do you have to?
00:35
it should work without such a thing
00:38
<catalystduck>
I had to edit the xorg.conf when I had ubuntu installed on the machine (it's old) to get the resoloution right
00:42
<johnny>
well.. does it work right with ltsp or not? :)
00:43
<catalystduck>
it works but the display is crappy, that's what I'm trying to fix
00:44
<johnny>
ok.. just making sure the problem didn't automatically resolve it self before you go hacking
00:44
well.. just copy the file into the chroot, run ltsp-update-image
00:45
and then edit lts.conf for it
00:45
[workstation ip or mac or hostname]
00:45
X_CONF=/path/to/conf
00:45
/var/lib/tftpboot/i386/
00:47
<catalystduck>
that's where I'm confused I've googled for this problem and been told to edit lts.conf files in various locations, some don't exist
00:49
<johnny>
there's 2 locations on ubuntu
00:49
one is deprecated
00:49
that is in /etc/lts.conf in the chroot
00:49
which tells you so
00:50
<catalystduck>
and how do I run ltsp-update-image?
00:50
is that the command itself?
00:50
<johnny>
yes
00:51siki has quit IRC
00:51
<catalystduck>
ok thanks
01:04Hyperbyte has joined #ltsp
01:05
<catalystduck>
johnny, where do I need to put the xorg.conf file?
01:27
<johnny>
wherever
01:28
in the chroot
01:35
it's 2:30.. i gotta attempt to hit the sack..bbl
01:39
<catalystduck>
Thanks for the help Johnny
02:01artista_frustrad has quit IRC
02:08polytan__ has joined #ltsp
02:16polytan has quit IRC
02:27polytan__ has quit IRC
02:34jammcq has quit IRC
02:43Eghie has joined #ltsp
02:43
<Appiah>
"Is being held until the list moderator can review it for approval. The reason it is being held: Message has a suspicious header"
02:43
What's so suspicious about my header =(
02:53Shingoshi has quit IRC
02:57hanthana has joined #ltsp
02:58EXP2 has quit IRC
03:05tjikkun_work has joined #ltsp
03:11eric has joined #ltsp
03:12
<eric>
hi!
03:12eric is now known as Guest98923
03:12
<Guest98923>
need help to make local devides work with NX-LTSP5
03:13
My ltsp server is functional (debian-etch). Local devices work fine with ldm.
03:14
But I need to plug 60 thin clients on a 100Mb network, so NX is indispensable
03:14
Login with NX works
03:16
I'm trying to write a script (nx-localdev) called by startnx in screen.d
03:18
It seems localdev managing is initialized by the ldm script. Which part of it must I keep ?
03:20randra has joined #ltsp
03:22mikkel has joined #ltsp
03:29
<Guest98923>
Is there anybody here who can answer my question ?
03:37
<Appiah>
in short you're trying to enable localdev in nx-ltsp5 ?
03:39
<Guest98923>
Appiah : yes, that's it
03:43
I've been on : https://wiki.ubuntu.com/DebugLocalDev page
03:44
Every step is OK until step 3. in step 4, they tell to type the command :
03:44
ssh -X -S /var/run/ldm_socket_vt7_192.168.0.254 192.168.0.254 "/usr/sbin/ltspfsmounter /tmp add"
03:45
and to post the error messages on irc
03:45
In my case, it first ask for root password (what it shouldn't do, of course)
03:46
then it says : Error: /tmp/.root-ltspfs/tmp is not mounted
03:47
(if it can help debugging)
03:51catalystduck has quit IRC
04:40gonzaloaf_laptop has quit IRC
04:47siki has joined #ltsp
05:23ubu3 has joined #ltsp
05:39russell_nash has joined #ltsp
05:41
<russell_nash>
release: Ubuntu Hardy. I have a problem with a blank screen and mouse pointer on my client after login. I have posted on mnay forums and reda much of the documenttaion
05:41
oops, typos
05:43
what i meant to say is that I have searched far and wide and nobody seems to have documented an error which is quite the same, can anybody offer any suggestions?
05:46
<Appiah>
that you force the X settings
05:46
in lts.conf
05:49
<russell_nash>
thanks for replying, what exactly do you mean, I have made many adjustments to lts.conf e.g. XSERVER auto, vesa , color depth 24, 16.
05:49
<Appiah>
try changing them
05:50
a lower resoultion , a forced mouse device , etc
05:50
see if it changes anything
05:50
Is this LTSP 4.x or 5 ?
05:55
<russell_nash>
incidentally, my ltsp worked fine for months, then one day stopped, i made no changes except for Ubuntu automatic updates. it is version 5.0.40 according to synaptic package manager. ok, i haven't tried resolution and mouse adjustments, although the mouse appears to be working perfectly, I just never get into the x session. My xsession-errors log says that "another SSH agent is running" and Fatal IO error 2 (no such file or
06:10alkisg has joined #ltsp
06:12generalsnus has joined #ltsp
06:12
<generalsnus>
Is there any way of viewing a LTSP session? like remote desktop into a thinclient session..
06:15russell_nash has left #ltsp
06:19
<Appiah>
just viewing or controling?
06:21
<generalsnus>
both..would be nice
06:22hanthana has quit IRC
06:22pmatulis has joined #ltsp
06:23
<Appiah>
maybe there's a feature in xdmcp?
06:23
<generalsnus>
usage, would be to troubleshoot user problems.. so we dont have to run all over school.. just tell the students to press "ok" button..and such
06:24
xdmcp?
06:24
<Appiah>
ye i understand , just like in Windows Terminal services / citrix
06:24
actully I have no idea if such feature exists
06:24
would be great
06:24
<generalsnus>
hmm
06:26
<alkisg>
generalsnus: https://help.ubuntu.com/community/UbuntuLTSP/iTalc
06:27
<Appiah>
ok
06:28
<klausade>
generalsnus: you use ssh or xdmcp?
06:32otavio has quit IRC
06:33otavio has joined #ltsp
06:33Guest98923 has quit IRC
06:51ajohnson has joined #ltsp
06:52Eghie has quit IRC
06:58Egyptian[Home]1 has quit IRC
06:59Egyptian[Home] has joined #ltsp
06:59BrunoXLambert has joined #ltsp
07:03
<generalsnus>
looking at italc now..
07:05
<Appiah>
klausade: is there a way with xdmcp ?
07:08
<klausade>
Appiah: there is a way with both, it's just different.
07:08
<Appiah>
Could you explain both of them or do you have a link to some howto?
07:11
<klausade>
Appiah: have a look at teachertool and controlaula.
07:12
<Appiah>
thanks
07:35kriegaffe has joined #ltsp
07:36ubu3 has quit IRC
07:37
<kriegaffe>
hello, is this the correct channel for questions about ltsp?
07:38
I'm looking for a howto on the localapp side of ltsp5 (xubuntu 8.04.1 based)
07:39
And I found a lot of interesting stuff (been working on it for several weeks now)
07:39
<Appiah>
isent much of a howto...
07:40
<kriegaffe>
but still when I start any application on the thin client it doesn't show up in "ps" on the client
07:40
<Appiah>
copy the ltsp-localapps script to /usr/bin , chroot into the ltsp client envoirment
07:40
install apps , exit , update the image
07:40
<kriegaffe>
I am confused: do I need nfs with the ltsp5 system for localapps to work?
07:40
<Appiah>
start application with ltsp-localapps script
07:40
no
07:41
<kriegaffe>
ok, the ltsp-localapps script resides somewhere on the server I assume, I'll check if I can find it. Thanks! (the chroot install is ok)
07:41
<Appiah>
/usr/share/ltsp or somehting like that
07:41
<kriegaffe>
ok. thanks
07:47alkisg has quit IRC
07:48
<kriegaffe>
I found a few scripts in the /usr/share/ltsp/scripts directory but none seem related to localapps (it's 3 scripts about popularity-data and daemon wrapping)
07:48
my ltsp version is this one: 5.0.40~bzr20080212-0ubuntu7
07:50
the lts.conf in /var/lib/.../tftboot/i386/ has the localdev, local_apps, local_storage all set to True.
07:50
<generalsnus>
I have managed to get "student-control-panel" to work, i can see users desktop with the x11vnc.. but how can i control the desktops?
07:51nubae has joined #ltsp
07:54vvinet has joined #ltsp
07:56mikkel has quit IRC
08:01six2one has joined #ltsp
08:08pscheie_ is now known as pscheie
08:09
<pscheie>
what is the name of the unified rpm repository effort?
08:11EXP2 has joined #ltsp
08:15Blinny has joined #ltsp
08:17X0d_of_N0d has quit IRC
08:20Gadi has joined #ltsp
08:22X0d_of_N0d has joined #ltsp
08:23dmaran has joined #ltsp
08:25mikkel has joined #ltsp
08:30generalsnus has quit IRC
08:36polytan has joined #ltsp
08:36
<polytan>
hi
08:40CAN-o-SPAM has joined #ltsp
08:42
<Appiah>
kriegaffe hold on
08:43elisboa has quit IRC
08:43elisboa has joined #ltsp
08:45
<polytan>
I'm installing ltsp on a gentoo system
08:46
(one server and one netboot node currenlty)
08:47
<nubae>
johnny: is the guy to speak to... don't know if he's around....
08:47cliebow has joined #ltsp
08:47
<Appiah>
kriegaffe: /usr/share/doc/ltsp-server/examples/ltsp-localapps
08:48
<polytan>
is there any gentoo dev here ?
08:48
I just whant to know what how to follow once I've compiled and install all program to make the client using ltsp instead of normal X connections
08:48
want*
08:48
<Appiah>
you dont need a dev for that
08:49
<polytan>
lovely :)
08:49
<Appiah>
...
08:49
<polytan>
simply the ltsp official doc ?
08:49
<Appiah>
have you tried it?
08:50
<polytan>
no
08:50
<Appiah>
the "offical" is not LTSP 5 if im correct
08:50
<polytan>
I've just finished to install a netboot system (it runs)
08:51
and to compile the ltsp-server on the server and ltsp-client on the client
08:53
<nubae>
the offical IS LTSP 5
08:53
<sbalneav>
Morning all
08:53
<Appiah>
nubae: do you have a link to this offical document?
08:53
<sbalneav>
!doco
08:53
<ltspbot`>
sbalneav: "doco" is you can find the link to the latest, and most current upstream documentation at: http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
08:53
<nubae>
!docs
08:53
<ltspbot`>
nubae: "docs" is For the most current documentation, see http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
08:53
<Appiah>
for gentoo that is
08:54
<nubae>
talk to Johnny... he should know
08:54
<Appiah>
:D
08:54
<Gadi>
!s
08:54
<ltspbot`>
Gadi: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
08:54
<polytan>
the thing is that I've read the official doc
08:55
<Appiah>
which offical doc polytan ?
08:55
<polytan>
but there are a few specific things on gentoo (init script) and I don't want to destroy to many things
08:55
the doc in the topic : http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
08:55
<Appiah>
ok
08:56
<polytan>
for exemple, there are ltsp-client and ltsp-client-setup init script
08:56
and I don't really know which one need the other
08:56
<sbalneav>
Well, that's the UPSTREAM documentation
08:57
<CAN-o-SPAM>
hi all
08:57
has anyone ever purchased a pre-configured linux OS server from anywhere?
08:57
<sbalneav>
if gentoo's got specific stuff, it'll be up to them to either a: provide me with some bits so I can stick it in the upstream docs, or b) branch the upsteam and provide their own docs package.
08:57
<polytan>
yes
08:57
<sbalneav>
CAN-o-SPAM: Not I.
08:58
<Appiah>
http://www.gentoo.org/doc/en/ltsp.xml wonder if this is up to date
08:58
no its 4.1
08:58
"/opt/ltsp-4.1/i386"
08:58
<sbalneav>
No it's not.
08:58* Appiah takes a guess
09:01artista_frustrad has joined #ltsp
09:02
<nubae>
is debian-edu still based on 4.2?
09:02
or has it migrated to 5?
09:08
<polytan>
I don't understand what to do to use ltsp opn my client
09:09
I can netboot, the OS starts, I can have X and use xdmcp on the server
09:09alkisg has joined #ltsp
09:09
<polytan>
but I don't understand how to put ltsp upon that
09:09
<Appiah>
have the client connect to the xserver ?
09:10
ldm (?!)
09:10
<kriegaffe>
Appiah : thanks! I'm now reading your info
09:11
<polytan>
Appiah, no, only using gdm and xdmcp
09:11
I've read ldm in the doc this morning ;)
09:12
I'm installing ldm
09:13
<nubae>
ltsp 5 = ldm
09:14
<kriegaffe>
Apiah : the /usr/share/doc/ltsp-server/examples/ltsp-localapps is not on my system but I'm looking
09:14
<polytan>
ok
09:15
<Appiah>
kriegaffe: what are you running again?
09:17
<kriegaffe>
Appiah: I have the 5.0.40 on xubuntu ltsp-server installed with ltspfs etc. I found a full client docu in the /usr/share/doc/ltsp-server called "workstation"
09:19
but no ltsp-localapps script anywhere. I tried with find / -iname, also looked for the file using apt-file search
09:19
<Appiah>
>_>
09:21
<sbalneav>
kriegaffe: Which version of xubuntu?
09:21
<kriegaffe>
it is xubuntu 8.04.1 lts
09:21
<sbalneav>
Localapps wasn't in hardy
09:22
<kriegaffe>
ah, ok so I have to upgrade first
09:22
<sbalneav>
You'd have to go to Intrepid to get Localapps
09:23
<kriegaffe>
maybe I just make the ltsp clients fat with the "workstation" info. They can have that pressure since it is mostly firefox that will be used there.
09:23
intrepid is not an option since I want/need Longtermsupport
09:24
I recently did a very painfull upgrade from 6.06 to 8.04 systems (client and server) due to my limited knowledge. A "Support-forever" distro will certainly catch my interest!
09:26
Thanks for the help.
09:28zwiep`` has joined #ltsp
09:28zwiep`` has left #ltsp
09:31
<kriegaffe>
Does anybody know if there exists a chance that ubuntu hardy will support localapps? Does this just need an upgrade of the current 5.0.40 ltsp-server package?
09:31
or is there a specific kernel needed?
09:32
<sbalneav>
kriegaffe: "Support-forever"'s going to be kind of difficult.
09:33
kriegaffe: There might, if there's enough interest and/or time by some of the developers, be a backport of a more recent LTSP version to Hardy.
09:34
<nubae>
anyone know the numbers for worldwide ltsp deployments?
09:35
<sbalneav>
nubae: Not lateley
09:35
We used to think it was somewhere in the millions range, back in the 4.2 days.
09:35
That was when people came to LTSP to get it.
09:35
Now that it's in distro's, it's going to be much harder.
09:36
To figure out who all's running it, I mean.
09:36
<cliebow>
let's do a Gates..and forcew it to report home
09:36
<nubae>
hmmm I'm writing an article for olpcnews.com, I'd like a realistic number :-)
09:36
spain and Japan have the largest deployments right=
09:36
?
09:37
<cliebow>
Japan??
09:37
<nubae>
yeah big Fedora deployment iirc
09:37
<sbalneav>
Telecentros in Brasil has like 150,000 or so
09:38
I can personally speak for 185 :)
09:38
<CAN-o-SPAM>
nubae: check this out: http://www.frappr.com/k12ltsp
09:38
<sbalneav>
They have a "Linux Counter" project, we need an "LTSP counter" :)
09:39
University of Amazonia has several thousand.
09:39
<nubae>
yeah... Brazil must be a biggie...
09:39
I hope I can afford to attend the meet there...
09:39gonzaloaf_laptop has joined #ltsp
09:40cliebow has quit IRC
09:48cliebow has joined #ltsp
09:51Pascal_1_ has joined #ltsp
10:00yanu has quit IRC
10:01
<sbalneav>
I just posted a question to -develop, I'd like to hear the opinions of anyone and everyone.
10:01
About the scope of the docs
10:01yanu has joined #ltsp
10:05MeW_away is now known as dirigeant
10:20staffencasa has joined #ltsp
10:22jammcq has joined #ltsp
10:22
<jammcq>
good morning #ltsp
10:22
<rjune>
howdy jammcq
10:23
<jammcq>
hey
10:24
<Blinny>
Morning.
10:24
How does one 'fix' the time displayed by LDM? Do I need to adjust the tz rules in the chroot?
10:24
Or is it dependent upon the hardware clock of the individual client?
10:25
<jammcq>
ummm... LDM displays the time?
10:26
<sbalneav>
Blinny: It's dependent upon the hardware clock in the client.
10:26
Morning jammcq
10:26
<jammcq>
sbalneav: hey
10:26
<rjune>
!s
10:26
<ltspbot`>
rjune: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
10:26
<Blinny>
sbalneav: Normally you'd set an offset, specifying whether the hardware clock is set to UTC. Do you know how to do so in the chroot?
10:27
<rjune>
ltspbot`: help
10:27
<ltspbot`>
rjune: (help [<plugin>] [<command>]) -- This command gives a useful description of what <command> does. <plugin> is only necessary if the command is in more than one plugin.
10:29
<sbalneav>
Blinny: Same whay you'd do so on the server. tzconfig or something like that.
10:29
<Blinny>
I think in RH-based distros it'd be a boolean in /etc/sysconfig/clock (UTC=false)
10:29
sbalneav: OK thanks I'll check it out.
10:29
<sbalneav>
rjune: Looking for something in particular with ltspbot?
10:29
<rjune>
yeah, how to make a command
10:30
sbalneav: jim needs a !j :-)
10:31mccann has joined #ltsp
10:33Blinny has quit IRC
10:33
<sbalneav>
rjune: ltspbot: learn j as blahblahblah
10:36
ltspbot`: help learn
10:36
<ltspbot`>
sbalneav: (learn [<channel>] <key> as <value>) -- Associates <key> with <value>. <channel> is only necessary if the message isn't sent on the channel itself. The word 'as' is necessary to separate the key from the value. It can be changed to another word via the learnSeparator registry value.
10:37
<rjune>
!j
10:37
<ltspbot`>
rjune: "j" is jamcq!!!!!!!!!!!!!!!!!!!!!!!
10:37
<rjune>
!j
10:37
<ltspbot`>
rjune: "j" is (#1) jamcq!!!!!!!!!!!!!!!!!!!!!!!, or (#2) jammcq!!!!!!!!!!!!!!!!!!!!!!!
10:38
<rjune>
!sigh
10:38
<ltspbot`>
rjune: Error: "sigh" is not a valid command.
10:38
<sbalneav>
Now you've got it in there twice
10:39
<rjune>
!j
10:39
<ltspbot`>
rjune: "j" is jammcq!!!!!!!!!!!!!!!!!!!!!!!
10:39
<rjune>
there.
10:39
sheesh
10:39
<sbalneav>
!topics
10:39
<ltspbot`>
sbalneav: "topics" is To get a list of topics, type ltspbot: factoids search --values
10:39
<rjune>
sbalneav: I got oit worked out. finally
10:40
<sbalneav>
ltspbot`: factoids search --values
10:40
<ltspbot`>
sbalneav: 'ltsp', 'sbalneav', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', 'localdev', 'localdev', 'localdev', 'checklist', 'muekow', 'bestltspdistro', 'serversize', 'wireless', 'sound', 'topics', 'integration', 'lts.conf', 'pastebot', 'bootfloppy', 'ltsp5', 'tarball', 'tarball', 'download', 'monkeys', 'ogra', 'nfs', 'nfsnotresp', 'js', 's', (1 more message)
10:40
<rjune>
!sbalneav
10:40
<ltspbot`>
rjune: "sbalneav" is Scotty !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
10:40
<rjune>
neat
10:40
!ogra
10:40
<ltspbot`>
rjune: "ogra" is ogra!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
10:40
<sbalneav>
ltspbot`: forget sbalneav
10:40
<ltspbot`>
sbalneav: The operation succeeded.
10:41
<sbalneav>
ltspbot`: !tarball
10:41
<ltspbot`>
sbalneav: Error: "!tarball" is not a valid command.
10:41
<sbalneav>
!tarball
10:41
<ltspbot`>
sbalneav: "tarball" is (#1) http://wiki.ltsp.org/twiki/bin/view/Ltsp/Ltsp5TarballInstructions, or (#2) while the tarballs may work for you, most of the tarballs as of 2008-03-19 are extremely old and out of date.
10:42tjikkun_work has quit IRC
10:43
<sbalneav>
ltspbot`: forget tarball
10:43
<ltspbot`>
sbalneav: Error: 2 factoids have that key. Please specify which one to remove, or use * to designate all of them.
10:43
<sbalneav>
ltspbot`: forget tarball 2
10:43
<ltspbot`>
sbalneav: The operation succeeded.
10:43
<sbalneav>
ltspbot`: factoids search --values
10:43
<ltspbot`>
sbalneav: 'ltsp', 'icewm', 'frappr', 'wiki', 'debian', 'edubuntu', 'dhcpd', 'greyscreen', 'ltsp42', 'localdev', '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)
10:44
<sbalneav>
!localdev
10:44
<ltspbot`>
sbalneav: "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, or (#4) http://wiki.ubuntu.com/DebugLocalDev
10:44
<sbalneav>
Hmm, factoids list scrolls off the screen (1 more message)
10:45
I should write a plugin for supybot that fixes that.
10:45
<johnny>
hmm..
10:45
that was a lot of questions going on..
10:48
<sbalneav>
!download
10:48
<ltspbot`>
sbalneav: "download" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/DownLoads
10:48
<sbalneav>
!docs
10:48
<ltspbot`>
sbalneav: "docs" is For the most current documentation, see http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
10:48
<sbalneav>
doco
10:48
!doco
10:48
<ltspbot`>
sbalneav: "doco" is you can find the link to the latest, and most current upstream documentation at: http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
10:48
<sbalneav>
ltspbot`: forget doco
10:48
<ltspbot`>
sbalneav: The operation succeeded.
10:49japerry has quit IRC
10:54mistik1 has quit IRC
11:05japerry has joined #ltsp
11:14
<polytan>
re
11:14
<johnny>
hello polytan
11:14
i tried to catch you before you left last time
11:14
but i just missed you
11:14
<polytan>
I'm here !
11:15
<johnny>
if you want ltsp for gentoo, you must use layman
11:15
<polytan>
I did it
11:15
<johnny>
and add layman -a ltsp
11:15
<polytan>
and i've installed ltsp-server on the server and ltsp-client on the client chroot wich I netboot
11:16
(the netboot works but I don't know how to activate ltsp)
11:16
<johnny>
you're not supposed to manually install ltsp-client
11:16
you need to run ltsp-build-client
11:16
it sets things up properly
11:18
<polytan>
ok
11:18
so I enter the chroot client
11:18
and I run ltsqp-buikld-client
11:18
-k
11:19
there is no command ltsp-build-client in the gentoo chroot environment
11:19
is it into the server package ?
11:19
yes
11:20
johnny, does this command create its own environment ?
11:21
because I've already build a gentoo in a folder that can be netbooted
11:21
<johnny>
sure, but it also installs ltspfs, ldm, and removes stuff that gets in the way
11:22
take a look at /etc/ltsp/quickstart/profile.qs
11:22
that gives an overview.. altho in quickstart profile style
11:22Eghie has joined #ltsp
11:22
<johnny>
the important thing is to have ltspfs and ldm, and make sure ltsp-client and ltsp-client-setup are added as in that quickstart profile
11:23
<polytan>
ok
11:23
but how to keep a gentoo system on the client ?
11:23japerry has quit IRC
11:23japerry has joined #ltsp
11:23vagrantc has joined #ltsp
11:23
<polytan>
because I don't want ubuntu or something else, I know how to administrate gentoo
11:25
I see
11:25
(I'm reading the profile.qs file
11:25
)
11:25
and actually, it does the work I've done
11:25
<johnny>
ok.. so add what i mentioned
11:25
and it might just work
11:27
<vagrantc>
Ryan52: ldminfod in ldm-trunk definitely supports xsession, as well as the ldminfod shipped with ltsp in debian experimental :P
11:27
<polytan>
c'est super
11:27
exactement ce que j'ai finalement
11:27
sorry
11:27ogra has joined #ltsp
11:28
<polytan>
johnny, this script is fantastic
11:28
the netboot node creation as been totally automatised
11:28
<johnny>
i can't wait to remove the ~arch stuff
11:28
<polytan>
I did the same
11:28
<johnny>
waiting for some of it to be finalized
11:28
<polytan>
on both my server and my ndoe
11:28
node*
11:28
<johnny>
ltsp-build-client is the standard wrapper to install ltsp, so i just wrapped it around quickstart
11:28
<polytan>
and next, to add an apps, can I use emerge ?
11:29
<johnny>
you add apps to the server
11:29
in most cases
11:29
<vagrantc>
Ryan52: though i see you've made some additional changes...
11:30Q-FUNK has joined #ltsp
11:30
<vagrantc>
Ryan52: it looks like you removed support for finding an Xsession script :P
11:33Q-FUNK has quit IRC
11:33gonzaloaf_laptop has quit IRC
11:38
<vagrantc>
Ryan52: unlike most ldminfod values, the ldminfod xsession stuff was added to src/ldm.c
11:38
now i see why you thought it was missing
11:51
<polytan>
in the ltsp-build-client script, is the default distribution the one running on the server ?
11:52
<sbalneav>
off for lunch
11:52
<johnny>
that is the only distribution supported
11:52
to be installed
11:54
<polytan>
ltsp-build-client --dist gentoo --base /netboot --chroot GentooLTSP --locale fr_FR.UTF-8 --prompt-rootpass
11:54
is that ok ?
11:55
hum
11:55
<vagrantc>
sbalneav: all i can muster regarding the ltsp docs is "yeah, scotty. sounds good!"
11:55
<polytan>
/etc/ltsp/quickstart/profile.qs: line 42: /etc/conf.d/clock: No file or folder (I'm using openrc on my server)
11:55
<johnny>
yeah.. so am i..
11:55
i thought i had the logic right
11:55
ignore --dist
11:55
it doesn't do anything on gentoo atm
11:56
if anything, it would be --dist 2008.1
11:56
or whatever
11:56
<warren>
vagrantc: what exactly did ryan52 do?
11:56
<johnny>
you should just stay with the structure imo
11:56
and not use base or chroot
11:56
<polytan>
ok
11:56
<johnny>
the chroot option is actually useful tho
11:57
in case you already have a chroot
11:57
and building another one
11:57
<vagrantc>
warren: removed support for the Xsession detection while adding support for selecting sessions by name (rather than the binary)
11:57
<johnny>
lemme look into the clock thing
11:57
<warren>
vagrantc: oh yeah, we need that back
11:57
<johnny>
polytan, you didn't follow the openrc instructions then!
11:58
<polytan>
johnny, I still have the clock problem even if I remove the --dist,base,chroot option
11:58
<johnny>
# For OpenRC
11:58
if [ -e /etc/timezone ]; then
11:58
TIMEZONE="$(</etc/timezone)"
11:58
else
11:58
. /etc/conf.d/clock
11:58
fi
11:58
fi
11:58
<vagrantc>
warren: i already committed and pushed a fix
11:58
<johnny>
oops.. too many lines
11:58
<warren>
vagrantc: both ltsp and ldm or only ldm?
11:58
<polytan>
hum
11:58
<johnny>
openrc should require /etc/timezone
11:58
<polytan>
strange
11:58
<vagrantc>
warren: i think ldm only
11:58
<johnny>
maybe i should change the check for /etc/conf.d/hwclock instead
11:58
err /etc/init.d/hwclock rather
12:00
<polytan>
johnny, is that a problem if my files are not in /opt/ltsp ?
12:00
<johnny>
it just helps consistantcy
12:00
and other people know how to help easier :)
12:01
no.. it's not really a problem..
12:16spectra has joined #ltsp
12:31mikkel has quit IRC
12:42polytan has quit IRC
12:43alkisg has quit IRC
12:46ogra has quit IRC
12:50ogra has joined #ltsp
12:53
<rjune>
!ogra
12:53
<ltspbot`>
rjune: "ogra" is ogra!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
12:53alkisg has joined #ltsp
12:54
<rjune>
ogra: Ich habe etwas fur dich
12:54Lns has joined #ltsp
13:16randra has quit IRC
13:21
<johnny>
cds with 101 songs are fun..
13:21
especially when cds can only have 99 songs.. :)
13:21
99,100,101 are all on one track with a few seconds spacing
13:22ogra has quit IRC
13:27ogra has joined #ltsp
13:40
<rjune>
ogra: you about?
13:42mccann has left #ltsp
13:43lbruno has joined #ltsp
13:43* lbruno waves
13:43
<lbruno>
should LTSP use the support in Debian's rcS for "/var/run and /var/lock in tmpfs"?
13:44
or does that happen too late?
13:45mistik1 has joined #ltsp
13:45
<vagrantc>
lbruno: i think i looked into it in the past, and it interfered with something
13:46
<lbruno>
vagrantc: got it
13:46
<sbalneav>
How's it different from what we're doing now, anyway? /var's on a ramdrive
13:46
<lbruno>
sbalneav: the complete /var?
13:46
<vagrantc>
no
13:46
just what's needed
13:46
<lbruno>
vagrantc: I'll have to look at this myself
13:46
<vagrantc>
and it's configurable
13:47
<lbruno>
vagrantc: ah, ok; I thought that maybe Debian's integration was not using the latest LTSP
13:47
<ogra>
vagrantc, i never disabled it in ubuntu ... should just work
13:48
<vagrantc>
lbruno: well, the version in lenny isn't the latest, but it's not absurdly behind
13:48
ogra: it's different in debian
13:49
<lbruno>
./var/run and /var/lock are mounted in /etc/rcS.d/S02; early enough, I take it? my previous LTSP (5weeks old lenny) is complaining of /var/run being hidden
13:49
<vagrantc>
lbruno: the version in etch is quite old... but the /var/run, /var/lock stuff is ancient.
13:49
<lbruno>
vagrantc: true! that's what catapulted me into lenny
13:50
.oO(for which I'm thankful, btw)
13:52
I spoke at first with imperfect knowledge; mounting tmpfs in all rw_dirs happens much later in the boot process, so I guess it can work like that
13:52
<vagrantc>
lbruno: i haven't done a clean lenny install in a while... i've mostly been testing debian experimental packages for ltsp
13:52bobby_C has joined #ltsp
13:52
<lbruno>
and I'm running fat-clients, not thin-
13:52
is this uncommon?
13:52
<vagrantc>
not unheard of, but not common either
13:54
<lbruno>
vagrantc: i'm going to do a couple more clean installations (and lenny's up to date on most packages); I guess those who are running fat-terminals don't document much
13:54
I've stumbled through a couple of undocumented gotchas already
13:56
<vagrantc>
yeah, we've definitely focused more on genuinely thin clients
13:56
<lbruno>
then I'll ask your opinion on capacity issues; let me check the machine specs
13:57
<vagrantc>
shouldn't be any different from a disked system for capacity, except maybe more ram, as they either have no swap or networked swap
13:57* lbruno nods
13:57
<lbruno>
my LTSP server is a Intel(R) Pentium(R) Dual CPU E2180 @ 2.00GHz
13:58
if using thin-terminals, about how many clients would you put on this?
13:59
<vagrantc>
!serversize
13:59
<ltspbot`>
vagrantc: "serversize" is http://wiki.ltsp.org/twiki/bin/view/Ltsp/ServerSizing
13:59
<rjune>
depends on what you want to do with it
13:59
and how much RAM is in it
13:59
<lbruno>
I guess "it depends" is the answer to most of my questions
13:59
<vagrantc>
ram is almost more important than CPU, really.
14:00
lbruno: check out the serversizing document above... that gives some loose guidelines
14:00
<rjune>
vagrantc: that's a vantastically good article
14:01
<vagrantc>
haven't looked at it in a while...
14:01
<rjune>
vagrantc: whatever happened the php script I wrote?
14:01
<lbruno>
I can stick either 4G or 8G RAM on that machine; that's what drove me to fat-clients: I was too lazy to crack open some of the terminals and wanted to use their RAM without cutting my fingers in the process
14:01* lbruno reads
14:01
<vagrantc>
rjune: i know not of this php script :)
14:02
<rjune>
it was kinda need.
14:02
<vagrantc>
lbruno: also, though it's not integrated into lenny, you could use localapps... running individual applications on the thin client but mostly on the server.
14:02
<rjune>
neat
14:02
basically it just asked you # users, # terminals, then gave you a rough estimate of what you needed for RAM / Hard Drive / CPU
14:02* lbruno nods to vagrantc
14:03
<vagrantc>
though localapps is still in it's infancy
14:05
<lbruno>
those loose guidelines were quite nice, thank you
14:06
<vagrantc>
though, if in doubt, add more ram.
14:06
definitely err on the side of too much ram rather than too little.
14:06
<lbruno>
right now, I think my ceiling is the network
14:07
<vagrantc>
how so?
14:07
<lbruno>
I have two switches in the client->server path, and a 100mbps link between the switches...
14:07
...yep, dumb
14:07
<vagrantc>
probably not a huge deal.
14:08
<lbruno>
even considering I'm doing 30 fat-clients (10 per such switch)?
14:08
<vagrantc>
my main deploment has 15-20 users at a time, and it's all 100-base
14:08
<lbruno>
oh, ok
14:09
<vagrantc>
ah, fat clients may be different
14:09
<lbruno>
especially at the start
14:10
what boxes do you use for clients? I went for fat-clients because it seemed a waste: the boxes I get for EUR90 have 2GHz processors and 1GB RAM
14:13
<vagrantc>
those should make good diskless workstations, yes.
14:13* vagrantc doesn't like the term "fat clients"
14:14
<lbruno>
you have an alternative for the F-word?
14:15
<vagrantc>
i like "diskless workstations"
14:15* nubae likes low fat clients :-)
14:15
<vagrantc>
that makes my stomach churn
14:16
<lbruno>
ah, but the traditional thin-clients are diskless too! yes, nomenclature is difficult
14:16
<warren>
high fat, high cholesterol clients
14:16
<vagrantc>
though in the end, i guess we just need to know what we're talking about :)
14:16
lbruno: well, that's why i call them workstations rather than clients...
14:16
<warren>
100% bacon
14:16* lbruno nods
14:16
<lbruno>
got it
14:16
<nubae>
transfat
14:17
<lbruno>
must be dinner time at your TZ
14:17
<vagrantc>
i've heard fat clients referred to when talking about machines with disks vs. thin clients ...
14:17
<lbruno>
vagrantc: oh? that I didn't expect
14:18
<vagrantc>
though really, i've heard then all called all sorts of things, and none of the language is perfect. :)
14:18
<Gadi>
yo mama's thin client's so fat...
14:19
<vagrantc>
and ou get Gadi on a roll by discussing it
14:19
<Gadi>
:)
14:19* vagrantc seems to have a broken "y" key today
14:20
<Gadi>
vagrantc: I saw you perused by session-script/ stuff
14:20
<vagrantc>
Gadi: yes
14:20
<Gadi>
numbering aside, any thoughts?
14:21
stgraber and I were discussing a clean architecture for dropping in stuff
14:21
and this was kind of the result
14:21
tho, I dont like what he pushed upstream for an example script
14:22
so, I worked his script into the larger framework
14:23
<vagrantc>
Gadi: you know, what i don't like about this is this gets run with every screen session, X related or not ...
14:24
<Gadi>
hmm
14:24
well, thats easy to fix
14:24
(I think)
14:24
<vagrantc>
heh
14:24
<Gadi>
we could run it if and only if there are X_ params in the env
14:25
(which is a start)
14:25
it should run through quickly regardless
14:25
<vagrantc>
Gadi: and i don't see why you've got all those function definitions in separate files ... why not have a single file with all the functions defined and a place to insert customizations inbetween generating xorg.conf
14:26
<Gadi>
stgraber wanted individual files to make it easier to package
14:26
<vagrantc>
i guess it's easier for distro's to not include certain parts, then.
14:27* Gadi is not a packager, but I imagine so
14:27
<Gadi>
:)
14:28
<vagrantc>
but this really seems more appropriate to screen-x-common
14:28
<Gadi>
well, it has to happen before the Xserver
14:28
but, yeah, I agree with that
14:28
<vagrantc>
screen-x-common is sourced at the top of all screen scripts that use X
14:29
<Gadi>
the question is - are there any non-X related hacks that need to come before/after screen scripts
14:29
<vagrantc>
right, there might be a need for both.
14:29
<Gadi>
I suppose we could have X* files
14:29
;)
14:29
S*, K*, X*
14:30
<vagrantc>
XS XK
14:30
<Gadi>
and X* only get executed from screen-x-common
14:31
<vagrantc>
ou know about our wack /var/run/ltsp-xorg.conf ?
14:31
yyyyyy
14:31
<Gadi>
your wack?
14:31
<vagrantc>
s,wack,wacky,g
14:31
<Gadi>
no - we have a new xorg.conf?
14:32
<vagrantc>
heh
14:32
i'm not sure if it's used...
14:32
<Gadi>
not in upstream , yet, right?
14:32
<vagrantc>
been there a long, long time
14:32
<Gadi>
really?
14:32
oh boy
14:32* lbruno is using the /etc/X11 one
14:32
<lbruno>
unless I didn't get it
14:33
<vagrantc>
Gadi: since march
14:33
<lbruno>
I was under the impression that X -configure > /etc/X11/xorg.conf was the way to go
14:33
<vagrantc>
warren and i came up with /var/run/ltsp-xorg.conf at some point...
14:34
<Gadi>
to get around rw issues?
14:34
<vagrantc>
i suspect so
14:34dmaran has quit IRC
14:34
<Gadi>
fools!
14:34
:)
14:34
<vagrantc>
it's pretty simple, reall.
14:34
y
14:34alkisg has quit IRC
14:34
<Gadi>
ou dont sa
14:34
hehe
14:34
<vagrantc>
just when i get a decent laptop... the y key has to fail on me...
14:34
<Gadi>
couldnt resist
14:35
<vagrantc>
basically, it sets X_CONF, so if the screen-x-common.d scripts respect that, we should be fine ... if it's the kind of thing that can be set after that stuff
14:35
<rjune>
Gadi: you're mean
14:36* lbruno hates looking back on old commits because of this
14:37
<Gadi>
vagrantc: yeah, we'd have to check into that
14:37alkisg has joined #ltsp
14:37
<Gadi>
right now I think I have them writing to /etc/X11/xorg.conf
14:37
<vagrantc>
Gadi: basically, if /var/run/ltsp-xorg.conf exists, it sets X_CONF to that ... but we can't currently assume X_CONF is writeable, as it uses it directly otherwise.
14:38
i.e. X_CONF=/path/to/some/custom/xorg.conf
14:38
<Gadi>
right
14:38
so, if we wrote the hacks to ltsp-xorg.conf, then X_CONF would be set
14:39
where does X_CONF get set?
14:39
ltsp_config?
14:39
<vagrantc>
in screen-x-common
14:39
<Gadi>
or later?
14:39
ah
14:39
<vagrantc>
or lts.conf
14:39
it's the former XF86_CONF option or whatever it was called
14:39
<Gadi>
so, I might as well just change my default to ltsp-xorg.conf
14:40
and let that other code do the rest
14:40
<vagrantc>
if we put the screen-x-common.d stuff before it checks for /var/run/ltsp-xorg.conf
14:40
<Gadi>
right
14:41
I think stgraber needed a screen-session.d for his cluster stuff
14:41
<vagrantc>
i think we need to do a whole audit of where customizations to X can happen from, and clean it up a bit.
14:41
<Gadi>
so it could grab any changes to lts.conf
14:41
from the configurator
14:41
between sessions
14:42
<stgraber>
Gadi: not really for the cluster stuff but to apply some little changes on the xorg.conf like Virtual (that's depending on a lts.conf variable that can be changed at any time)
14:42
<Gadi>
stgraber: ah - are all of the hacks for X?
14:43
<stgraber>
hmm, not sure about that, we might also add some hacks for alsa and even move our current ltsp-cluster change to screen_session to a screen-session.d/ script
14:43
<Gadi>
well, then lets have an XS* and XK* that get called only from screen-x-common
14:44
OR have a screen-x-common.d in addition
14:44
not sure which you packagers prefer
14:44
:)
14:45
<vagrantc>
we're starting to get a lot of these .d directories ...
14:45
<Gadi>
its nice
14:45
<stgraber>
why having one more .d ? sorry I'm at UDS so haven't followed the whole discussion ?
14:45
<Gadi>
we have nice control of: before and after screen scripts, X initialization and before and after LDM logins
14:46
<vagrantc>
stgraber: so you don't reconfigure X whenever you start a shell screen session
14:46MeW_ has joined #ltsp
14:46
<Gadi>
stgraber: so that the screen-session hacks don't happen for non-Xorg scripts
14:46dirigeant has quit IRC
14:46
<stgraber>
vagrantc: can't we just do a "if" in the screen-session.d scripts ?
14:46
<vagrantc>
stgraber: that seems a whole lot clumsier
14:47
stgraber: if "$WHAT" ?
14:47
i think the XS*, XS* prefix could keep it simple enough
14:47
<Gadi>
right - we dont have a good identifier of a graphical screen script or otherwise
14:47
<stgraber>
so XS* and XK* would be called from screen-x-common ?
14:48
<Gadi>
right
14:48
<vagrantc>
maybe we even split some of screen-x-common into XS* scripts
14:48
<Gadi>
alternativelt
14:49
we can have screen-x-common set an environment var indicating it is an X script
14:49
and then we can is it out
14:49
*if it out
14:50
so, if boolean_is_true "$I_AM_AN_X_SCRIPT"; then
14:50
and that way, its clean
14:50spectra has quit IRC
14:51
<vagrantc>
i think it's easier with a different prefix
14:51
<Gadi>
okey dokey
14:51
<stgraber>
Gadi: screen-session.d is executed before screen-x-common is sourced so wouldn't work
14:51
<vagrantc>
then we don't have to wrap the whole script snippet into an if ; then ... which properly should have 4 indents
14:51
<Gadi>
right
14:52
gotcha
14:52
XS* and XK* it is
14:52
<vagrantc>
or, not indents, but spaces.
14:52
wheee!
14:52
then we can split out the X_CONF handling/setting into screen-session.d snippets
14:53
<Gadi>
right
14:54
and everything gets a bit cleaner
14:54cliebow has quit IRC
14:54
<rjune>
see, you guys scared cliebow off
14:54pmatulis has quit IRC
15:02
<warren>
hey what is ragnar's IRC name?
15:02
<jammcq>
laprag
15:02
<warren>
thx
15:04MeW_ has quit IRC
15:04
<rjune>
how's ragnar doing anyway?
15:04
<jammcq>
seems to be doing very well
15:05
<warren>
He didn't end up face down in the bushes outside this past year.
15:05
So I'd call that an improvement.
15:07dirigeant has joined #ltsp
15:07artista_frustrad has quit IRC
15:08lbruno has left #ltsp
15:08
<vagrantc>
Ryan52: the session-with-name patch is great!
15:08
<Gadi>
vagrantc, stgraber: something like this? lp:~gideon/ltsp/ltsp-trunk-screensessiond
15:09
<rjune>
warren: he didn't end up in the bushes face down when I went either.
15:13
<vagrantc>
Gadi: did you just delete and add all those scripts? "bzr mv" is a good freind :)
15:13
a.k.a. bzr rename
15:13
<Gadi>
thx
15:14
I like to have a few tools I can remember - that's why I use a hammer as a screwdriver
15:14
;)
15:14
<johnny>
except delete and add kills history
15:14
bzr can't possibly track across that
15:14
that's why mv is important
15:14
<Gadi>
gotcha
15:15
<vagrantc>
and saves space, over time.
15:15
<Gadi>
I assume it would mv the files for me, too
15:15
right?
15:15
<johnny>
why would it do that?
15:15
<warren>
Gadi: I'm glad you have fresh quotes. You were almost replaced.
15:15
<vagrantc>
Gadi: yes, it moves them.
15:15
<johnny>
oh.. bzr mv
15:15
yes
15:16ogra has quit IRC
15:16
<Gadi>
cool
15:16
<vagrantc>
Gadi: though i think it can't handle renaming multiple files, which is an annoying drawback
15:16
<rjune>
Gadi: I thought you used a hammer as a screwdriver because you were smart and added the phillips head to the bottom of the handle
15:16
<Gadi>
well, thats why God invented for loops
15:17
rjune: nah, every time I tried, I whacked myself in the head
15:17
those things really smart!
15:17
<rjune>
God invented for loops? when did he do that?
15:18
<Gadi>
rjune: creating the universe can get repetative - always helpful to write a script or two
15:18
<rjune>
Ahh
15:18
I see now
15:18
a big array to hold coordinates, etc. then a for loop to cycle through and create all the stars
15:19
good thinking. I hope he spun that off in its own thread
15:19
<warren>
Gadi: that isn't the purpose of delegation and outsourcing to cheap labor?
15:19
<rjune>
warren: why do you think the angels revolted?
15:20
<warren>
you mean when they unionized?
15:22
<Gadi>
hehe
15:22
<rjune>
warren: so unions are the work of the devil? that does explain a bit
15:23
<Gadi>
suddenly we all work for GM
15:23CAN-o-SPAM has quit IRC
15:24otavio has quit IRC
15:28otavio has joined #ltsp
15:32
<vagrantc>
well, i've got a hack to support ~/.dmrc parsing server-side for debian/ubuntu systems, at least.
15:33
well, i guess it doesn't parse ~/.dmrc ... but the premise holds...
15:34
<warren>
how far did Ryan52 get?
15:34
<ltsppbot>
"vagrantc" pasted "/etc/X11/Xsession.d/10x11-session-selection-hacks" (25 lines) at http://pastebot.ltsp.org/137
15:34
<vagrantc>
haven't seen any code for that
15:34
<warren>
Ryan52 was working on this last night
15:35
<vagrantc>
warren: what he did do was add support for using the more human-freindly names rather than /usr/bin/gnome-session or whatever for session selection
15:35
<warren>
vagrantc: no, beyond that he began working on ~/.dmrc
15:35
<vagrantc>
warren: cool.
15:35
<warren>
parsing it, writing a new one if the user picks something
15:35
<vagrantc>
warren: probably will be done soon, then. :)
15:35
<warren>
asking hte user if they want to save it or not
15:36
<vagrantc>
i'll wait to upload, then. :)
15:38
i stumbled through it for hours yesterday with no luck. trying to abuse ldminfo.c with a script in rc.d to parse ~/.dmrc and generate an ldminfo-like file to parse...
15:39
<warren>
let's see where Ryan52 is
15:40* vagrantc suspects a hostage of the state
15:42artista_frustrad has joined #ltsp
15:50BrunoXLambert has quit IRC
15:56otavio has quit IRC
15:57
<vagrantc>
[1489030.826963] clocksource/0: Time went backwards: ret=2fbb5e13a5 delta=-1982858490996942 shadow=2fbb12da1c offset=4b9f86
15:58polytan_ has joined #ltsp
15:58
<polytan_>
hi
15:58
<vagrantc>
oops.
15:58
<polytan_>
is there any video card driver discovering with ltsp or only vesa is used ?
15:58
<johnny>
huh?
15:58
it does whatever X can do
15:59
and X can detect the driver
15:59
it's nothing ltsp specific at all
15:59
<polytan_>
ok
15:59
is it automatically done ?
16:00
I've always configure my X manually, I don't really know the automatic things
16:00
<johnny>
oh.. that's right.. there's a thing in the quickstart profile that i thought i committed :(
16:00
it sets VIDEO_CARDS=vesa
16:00
<polytan_>
yes
16:00
<johnny>
so.. you're gonna have to install the others manually
16:00
easiest way
16:01
remove that line
16:01
<polytan_>
and how X choose the right driver ?
16:01
<johnny>
and then emerge xorg-server
16:01
magic
16:01
no.. it looks at what hardware is in the system
16:01
and uses it
16:01
lemme remove that line for future usages
16:01
<polytan_>
ok
16:01
so I will do it too tomorrow
16:02
is it the same with type of keyboard, mouse (no pb I think with the mouse ;) ) and screen resolution ?
16:02
<johnny>
generally yes
16:03
<polytan_>
ok
16:03
so I will have pb with that ;)
16:05
<johnny>
i just committed the removal of VIDEO_CARDS from the default make.conf
16:05
<polytan_>
ok
16:12
<stgraber>
/query nand
16:12
oops
16:13
<vagrantc>
Gadi: i'd also like to use run_parts_list instead of more find calls, so we use a consistant namespace for everything
16:13
Gadi: that's a function in ltsp-trunk/ltsp-common-functions
16:14
probably need a "run_parts_list_with_prefix" wrapper function
16:15ogra has joined #ltsp
16:16davidj has joined #ltsp
16:21six2one has quit IRC
16:29
<Gadi>
vagrantc: can I simply add an optional second argument to run_parts_list?
16:30
[ -n "$2" ] && find_args="$find_args -name $2"
16:30
or some such
16:30
<vagrantc>
i was adding a separate function for it ...
16:30
<Gadi>
seems like overkill, no?
16:31
you just repeat the code
16:31
<vagrantc>
although, it looks like run_parts_list exits if there's no directory...
16:31
might be nice to have a more tolerant function, that simply doesn't make any output
16:31
<ogra>
launchpad is going opensource in june :)
16:32
just was officially announced
16:32* vagrantc wonders why june
16:32
<ogra>
code cleanups
16:32
<vagrantc>
well, that'll be nice. :)
16:32
<Gadi>
wow - opensource and cleaner code!
16:32
<ogra>
and licensing/copyright rewiew
16:33
<Gadi>
that should be part of the announcement
16:33
<ogra>
also signed PPAs next week
16:34
Gadi: it was ... i'm sitting in the announcement talk
16:34
<Gadi>
ah
16:34
:)
16:34
<warren>
Does this make it easier for everyone to make launchpad their upstream?
16:34artista_frustrad has quit IRC
16:34
<Gadi>
ogra: start a wave
16:34
<ogra>
it will probabl take a moment for it to show up in written form
16:35artista_frustrad has joined #ltsp
16:35
<ogra>
warren: you are free to enhance it so it doexs what you miss
16:36
i assume we'll see lots of launchpads in the future ... and probably also forks
16:36
<warren>
1) Make Ubuntu submit all patches back to upstream projects no matter where they are located. "Making available" is not sufficient.
16:37
2) Make Ubuntu stop shipping non-free components you can enable in the default install.
16:37
I doubt launchpad opening will solve these.
16:37* ogra wondrs what either has to do with LP
16:38
<ogra>
and ii also hear a lot of complaints recently that RH isnt actually acting better in projectswher they have no employees actively in the upstream projects
16:39
<warren>
That is true to a certain degree, but far less than Ubuntu.
16:39
We welcome people to point out those cases
16:39
<ogra>
your 2) would simply be again st the ubuntu philosophy ...
16:40
<warren>
which is "Make users today happy at any cost."
16:40
<ogra>
though i assume you are aware that ubuntu has a "fully free" install mode since brrzy
16:40
<warren>
http://lwn.net/Articles/294891/
16:40
<ogra>
breezy
16:40
<warren>
Yes, I am aware.
16:41Egyptian[Home]1 has joined #ltsp
16:42
<ogra>
you pasted the aticle 100 times here ... the 101th time wont change anything
16:42
<warren>
it still remains true.
16:43
Launchpad opening is a positive development, but it is not core to the problems of Ubuntu.
16:43
<ogra>
depends on your view
16:43bobby_C has quit IRC
16:43
<ogra>
these are not problems, its a definition
16:44
you are free to disagree with the definition ... but that wont change ubuntus focus
16:45
<warren>
"The World Trade Organization abusing the developing world through abusive concessions that threaten long-term sovereignty in exchange for loan guarantees and grants. This is not a problem, but rather a definition of the WTO."
16:45* ogra goes for a smoke
16:46twinprism has joined #ltsp
16:50alkisg1 has joined #ltsp
16:51
<johnny>
warren, non free components? seriously?
16:52
<warren>
johnny: Fedora ships zero non-free components.
16:52
(we just have a different definition of Free than the FSF)
16:52
<nubae>
I doubt many people can live with everything free
16:52
<johnny>
warren, actually.. linux mint does more of a job at making people happy than ubuntu
16:52
<vagrantc>
Gadi: got the prefix handling as second argument
16:52
<johnny>
it includes all the non free components
16:52
that ubuntu does not
16:52
<warren>
then why aren't fanboys flocking to them?
16:52
<johnny>
name recognition
16:52
<nubae>
right, I usually install that for my parents/relatives/friends
16:53
they do... mint has more users than fedora according to distrowatch :-)
16:53
<johnny>
warren, i'd really appreciate it if open development came with freedom
16:53
you guys are doing terrible when you are the upstream in the eyes of many external folks
16:53
especially in regards to hal and devicekit
16:53
so get off your high horse
16:53
<nubae>
ubuntu, opensuse, then Mint...
16:54
<johnny>
and policykit
16:54
<warren>
johnny: That is not a common theme across most of Fedora. We are terrible in specific areas, even other Fedora and RH devs are upset at them.
16:54
<johnny>
sure.. but those specific areas are really important to the gnu/linux infrastructure
16:55
i even got yelled at for questioning the wisdom of the mighty davidz
16:55
saying that he was a hero
16:55
and whatnot
16:55alkisg has quit IRC
16:55
<nubae>
davidz?
16:55
<johnny>
david zeuthen
16:55
policykit, hal
16:55
devicekit
16:55
<nubae>
ah
16:55
<johnny>
he works on all those important things
16:56
<warren>
I do wish he would be less of an ass.
16:56davidj has quit IRC
16:57
<johnny>
it's not just him.. sounds like he has a clique...
16:57Egyptian[Home] has quit IRC
16:57
<warren>
vagrantc: I'm pushing my alternate LDM theme as an IFDEF to ldm-trunk. It can be ripped out when ogra's more flexible ldm layout rewrite happens.
16:57
<johnny>
warren, i really appreciate your interest in keeping things free
16:58
<vagrantc>
warren: it's fairly simply?
16:58
<warren>
vagrantc: yes, simple
16:58
<johnny>
but choose your battles
16:58
and don't piss of valuable contributors
16:58
<vagrantc>
warren: fine by me
16:58
<warren>
johnny: yes, his entire department has a few difficult personalities
16:58
<johnny>
otherwise i'm going to jump in
16:58
and ruff you up
16:58
lol
16:59
<ogra>
warren: as i said, you guys go on as usual with LDM, my rewrite will still take a while
17:00
<warren>
ogra: ok
17:00* warren adds IFDEF's for 30 different layouts
17:00
<johnny>
warren, what i appreciate about fedora and ubuntu over debian is the integrated experience
17:00
<ogra>
heh
17:00
<johnny>
altho i really like what ubuntu is about to do in regards to dropping src debs
17:00
probably the best thing to ever happen
17:00
in distro technolog
17:01
<warren>
Gadi: any progress on more examples of X_* options?
17:01Lunks has quit IRC
17:02
<johnny>
perhaps i am too empathetic..
17:03
i just enjoyed hanging out with geeks/nerds for once in PDX and don't like seeing this kind of conflict
17:04alkisg1 has quit IRC
17:05Lunks has joined #ltsp
17:06brendan0powers has joined #ltsp
17:07brendan0powers has left #ltsp
17:08
<nubae>
Edubuntu can also refer to the Classroom LTSP server, which is a thin client based system.
17:08
thats still in edubuntu faq... is that incorrect now?
17:08
should I take that sentence out
17:12polytan_ has quit IRC
17:12
<warren>
oooh
17:12
I found a better way to make mkdst rpm work without adding my alt layout
17:20ogra_ has joined #ltsp
17:20
<Gadi>
warren: lp:~gideon/ltsp/ltsp-trunk-screensessiond
17:20
<warren>
oh
17:21
Gadi: how well tested?
17:21
<Gadi>
you'll see a bunch more scripts
17:21
not very well, yet, I am afraid - but it gives you an idea of what X_ options
17:22
in terms of xorg.conf, mostly: XSERVER, X_VERTREFRESH, X_HORZSYNC, X_VIRTUAL, X_VIRTUAL_AUTO
17:22
oh and
17:22
X_OPTION_** and X_MONITOR_OPTION_**
17:23
these are all things that *may* need to be specified, regardless of the xrandr and xorg.conf-less nature of things
17:24
I think, however, when all is said and done, we will have all the functionality of configure-x.sh in a much saner way
17:24
and with drop-in scripts that can be overidden or not packaged if it conflicts with the way a distro handles things
17:33
<warren>
wait, not understading
17:33
you are saying they NEED to be specified?
17:34
you can't rely on specifying nothing?
17:34* vagrantc notes *may*
17:34jammcq has quit IRC
17:35
<warren>
I don't understand why it may
17:38leio has quit IRC
17:38
<Gadi>
ok, here's an example: you want to force certain sync frequencies because your monitor lacks DDC support and the driver does not choose frequencies that allow the mode you want to be valid
17:39
another: you have hardware with a BIOS that reports having an internal LVDS panel when none exists and you need to tell the driver that there is no panel
17:39
another: you need to force the use of the "geode" driver ;)
17:39X0d_of_N0d has quit IRC
17:40
<Gadi>
there may be many reasons why no xorg.conf does not work in some cases
17:40
obviously, in an ideal world, one never needs xorg.conf
17:40Eghie has quit IRC
17:40
<johnny>
yet there's still no mouse tuning gui in gnome by default
17:40
or for configuring a trackpad
17:41
<vagrantc>
i am soooo glad i don't use gnome, then. :)
17:41X0d_of_N0d has joined #ltsp
17:41
<Gadi>
which reminds me to ask someone about how serial x input devices work in Xorg 1.5
17:41
<johnny>
you can change sensitivity and click threshold
17:41
in gnome
17:41
<Gadi>
are there inputattach files for every serial input device?
17:41
<johnny>
but you need gsyntapics to do a touchpad
17:42
but SHMConfig is not enabled by default for synaptics
17:42
without an xorg.conf
17:42
and nvidia drivers are especially heinous
17:42
no xorg.conf.. means you get a logo
17:42
yuck
17:42
they should drop that option altogether
17:42
o cam
17:42
i can't imagine a single person wanting to see it
17:43
<Gadi>
yeah - I think there are enough of these simple configs that we should continue to support and not throw them away and tell folks to use a static xorg.conf file
17:44
of course, we will need to teach people to set refresh rate with XRANDR_RATE_0 instead of X_VERTREFRESH
17:44
<johnny>
but what about nvidia?
17:44
does it take those parameters?
17:44
or ati proprietary for that matter
17:44
<Gadi>
like which?
17:44
<johnny>
XRANDR
17:44
iirc.. it takes some.. but not all
17:44
i'm not sure which
17:45
<Gadi>
ah - I support 1.2 and pre-1.2 params
17:45
rate is 1.2
17:45
<johnny>
it's sad that nvidia has the BEST implementation of anything
17:45
<Gadi>
but size is pre-1.2
17:45
<johnny>
for 3d
17:46staffencasa has quit IRC
17:47
<Gadi>
vagrantc: dropping the default 16-bit color?
17:52
<vagrantc>
Gadi: yeah.
17:53
<Gadi>
somebody's got bandwidth to burn
17:53
:)
17:53
<vagrantc>
heh.
17:54
i guess i hadn't done comparisons of 16 vs. 24 bit for network traffic...
17:54
<Gadi>
listen - you do what you gotta do
17:54
:)
17:55
they make some beautiful 24-bit wallpapers out there
17:55
32-bit is even nicer
17:58hanthana has joined #ltsp
18:00Gadi has left #ltsp
18:03
<warren>
PATTERN='*.patch'
18:03
echo $PATTERN
18:04
how do you echo $PATTERN without it expanding if .patch files are in the current directory?
18:12oh207_ has quit IRC
18:15
<vagrantc>
echo ${PATTERN} ?
18:15
nevermind
18:16
echo "${PATTERN}" seems to work!
18:19ogra_ has quit IRC
18:22ogra has quit IRC
18:23ogra has joined #ltsp
18:28* warren writing a few bug fixes for mkdst
18:28lbruno has joined #ltsp
18:29
<lbruno>
well, that was a nice one
18:29
I mean, hi again
18:30Loto__ has quit IRC
18:30
<lbruno>
so. I never stumbled on this one: I have an eth1, but not an eth0
18:30
that plays some havoc
18:30
<vagrantc>
oh yeah.
18:32
<lbruno>
I'm testing if moving some lines of $LTSP/etc/initramfs-tools/modules around "solves" that
18:32Loto__ has joined #ltsp
18:33
<Ryan52>
oh, hi people :)
18:34
vagrantc: I removed it?
18:34
that could explain it :p
18:34
vagrantc: were there other weird things I removed?
18:35* Ryan52 used a horrible workflow for changing that..
18:35* Ryan52 goes to look at his diff
18:36
<vagrantc>
Ryan52: no, your patch is great!
18:36
Ryan52: with that one fix, that is :)
18:36staffencasa has joined #ltsp
18:37
<lbruno>
ok, this must be "basic" Linux, but I'm stumped: my onboard LAN (atl1) now thinks it's on eth1
18:37
<Ryan52>
okay, other than that I've got the dmrc done and working, I think (I did a bunch of cleanup after I tested, so I need to test again).
18:38
<vagrantc>
Ryan52: excellent!
18:38
<lbruno>
"I didn't change anything" isn't quite true, but I only rebuilt the initrd.
18:38
<warren>
vagrantc: hmm... find doesn't like cases in .bzrignore like "po/messages.mo"
18:38
<lbruno>
thoughts?
18:38
<warren>
vagrantc: find: warning: Unix filenames usually don't contain slashes (though pathnames do). That means that '-name `po/messages.mo'' will probably evaluate to false all the time on this system. You might find the '-wholename' test more useful, or perhaps '-samefile'. Alternatively, if you are using GNU grep, you could use 'find ... -print0 | grep -FzZ `po/messages.mo''.
18:39
<Ryan52>
and, if I can find some nice person to test if I didn't break localdev and sound later, then I can merge it :)
18:39
<vagrantc>
Ryan52: i can test it tomorrow, as it's a feature needed for freegeek
18:39
or maybe even tonight
18:40
<warren>
Ryan52: I can test it soon
18:40
<Ryan52>
okay. I'm gonna do some more testing right now, then I'll put up a bzr repo.
18:40
<warren>
hmm
18:40
autogen.sh is failing for ldm for me.
18:41
<Ryan52>
weird. works for me.
18:41
it says "Please add the files" but it overall suceeds..
18:41
<warren>
oh wait, no
18:41
it has bogus lines in .bzrignore
18:41
<Ryan52>
oh.
18:41
our autogen.sh isn't set -e.
18:41* Ryan52 fixes
18:42
<vagrantc>
it should definitely be set -e
18:42
heck, i wouldn
18:42
't even be opposed to set -x
18:42
<warren>
mkdst-trunk now removes anything listed in .bzrignore before tarballing
18:42
<vagrantc>
ah, slick.
18:42
<warren>
thus .bzrignore needs some cleanup
18:42
then there's this problem:
18:42
find: warning: Unix filenames usually don't contain slashes (though pathnames do). That means that '-name `po/messages.mo'' will probably evaluate to false all the time on this system. You might find the '-wholename' test more useful, or perhaps '-samefile'. Alternatively, if you are using GNU grep, you could use 'find ... -print0 | grep -FzZ `po/messages.mo''.
18:43
<lbruno>
:w
18:43* vagrantc suspects the .po files could really use updating
18:43
<warren>
smoke some .pot
18:43
=)
18:44
<vagrantc>
how does the .pot get updated?
18:44
<Ryan52>
make update-po, I think..
18:44* Ryan52 was looking at that, trying to figure out how to make his shell script i18n friendly.
18:45
<warren>
Ryan52: did you use exactly the same string from gdm or kdm for asking the user, so we can simply copy all of their translations?
18:45
<Ryan52>
warren, no, because I can't figure out how to get the 'make update-po' to work right.
18:45
if that's even what update-po does.
18:45* Ryan52 is confuzzled by this entire thing
18:46
<vagrantc>
looks like it's borked
18:46
<warren>
what is borked?
18:46
<vagrantc>
make ldm.pot
18:46
make: *** No rule to make target `../gtkgreet/ldminfo.c', needed by `ldm.pot'. Stop.
18:46
<Ryan52>
fix POFILES.in, then autogen.sh again.
18:47
or, no.
18:47
run configure again.
18:47
<warren>
do you have gettext installed?
18:47
<vagrantc>
gtkgreet/ldminfo.c was moved to src/ldminfo.c
18:47
<warren>
NOTE: Before we ask people to update the .po files, we should go through the strings and eliminate duplicates, or try to eliminate strings that are very similar.
18:48
<vagrantc>
indeed.
18:48
the simplest way to do that is to generate the current set of strings and look for duplicates
18:48
or near dupes
18:48
<Ryan52>
and if it's just a rewording in english we should unfuzzy.
18:48
<warren>
the .pot file is the set of strings
18:48
<vagrantc>
right
18:49
<Ryan52>
that sounds like a lot of work, in stuff that I have no clue about, so...:)
18:49staffencasa has quit IRC
18:50
<vagrantc>
fixed the make error
18:50
just had to fix it in POTFILES.in
18:51
<lbruno>
so, to workaround these migrant interfaces: should I hack a loop that gets all interfaces from /proc and tries ipconfig on them?
18:52
<vagrantc>
eeyk.
18:53
there's a fair number of strings that just plain shouldn't be translated
18:53
and a lot that could be consolidated
18:54
<Ryan52>
what shouldn't be translated?
18:54
<vagrantc>
"x_session: %s"
18:54
neither part should be translated
18:54
<Ryan52>
if we take out the gettext call on it will update-po stop noticing it?
18:55
<vagrantc>
and there's a bunch of ERROR: get_userid from greeter failed" that should simply have a %s in them
18:55
Ryan52: should.
18:56
<Ryan52>
does gettext eliminate exact duplicates?
18:56
or whatever this magic program is called :)
18:56* vagrantc wonders what the difference with "No reponse from server, restarting..." and "No reponse, restarting" is
18:57
<vagrantc>
Ryan52: yes, it will consolidate identical strings
18:57
that's why it's good to use consistant strings
18:58hanthana has quit IRC
18:58oh207_ has joined #ltsp
18:59
<vagrantc>
overall, when we did all the changes to ldm at the PDX hackfest, i think we happened to make big improvements on the localizable strings...
18:59
it doesn't look too bad after a quick glance
19:01
i'll give it a lookover the next weeks and make a call for updated translations
19:01
probably should do the same for ltsp, actually
19:02* vagrantc heads off to aikido
19:02
<Ryan52>
vagrantc: how do I make my shell script translatable?
19:02
vagrantc: I have the code written, but how do I make the .pot files know about the strings?
19:03
<vagrantc>
Ryan52: look at the files in ltsp-trunk/po/ ...
19:04
<Ryan52>
oh, ok, thanks :p
19:07Lns has quit IRC
19:11petre has joined #ltsp
19:12X0d_of_N0d_ has joined #ltsp
19:13X0d_of_N0d has quit IRC
19:14X0d_of_N0d_ is now known as X0d_of_N0d
19:14lbruno has left #ltsp
19:16
<warren>
Ryan52: where is your code that does the dmrc stuff? is it in both ldm and ltsp or only one?
19:17
Ryan52: I can test it tonight, but before pushing it, let me test the current trees.
19:18
<Ryan52>
It's broken right now...it's only in ldm.
19:18
<warren>
ok
19:23vagrantc has quit IRC
19:24
<Ryan52>
that's really stupid. kde 4 (at least on my system) completely ignores LANG variable when it's displaying the menus, and it made me think I had a bug -_-
19:25ogra has quit IRC
19:25
<warren>
how does it decide which language to display in the menus?
19:26
<Ryan52>
no clue...I just know that gnome is in french, and kde is in english :)
19:26* Ryan52 will look at that later
19:28
<Ryan52>
warren: bzr branch http://bzr.ryan52.info/bzr/ltsp/ldm-dmrc/
19:28
(if you want to test it)
19:29
the things that are not tested are sound, localdevs, and the system wide dmrc (set with the LDM_SYSTEMWIDE_DMRC variable).
19:29
<stgraber>
Ryan52: too bad I'm not at the office until next week ...
19:30
Ryan52: I'll have a look at the code and see if I can remotely test it with some computers at the office
19:30randra has joined #ltsp
19:31
<warren>
Ryan52: did you take into account that the name in dmrc does not match the session launcher name?
19:31
gdm for example writes "gnome" in there
19:31
somehow
19:32
<Ryan52>
what do you mean session launcher name?
19:32
the Session setting in .dmrc is the name of the file in /usr/share/xsession
19:32
s
19:32* Ryan52 read gdm's docs
19:32
<Ryan52>
and it works with gnome and kde.
19:33
<warren>
/usr/share/xsession or xsessions?
19:33
<Ryan52>
xsessions.
19:33
<warren>
ok good
19:33
Ryan52: could you please rebase this against the current ldm-trunk?
19:33
can't test it yet, gotta make sure everything before it works
19:33
<Ryan52>
hrm. a rebase with bzr...I can try :)
19:35
$ bzr push bzr+ssh://bzr.ryan52.info/bzr/ltsp/ldm-dmrc
19:35
bzr: ERROR: These branches have diverged. Try using "merge" and then "push".
19:35
(after rebasing)
19:35
what do I do?
19:35* Ryan52 tries with --overwrite
19:36
<Ryan52>
oh, ok, that worked.
19:36
Ryan52: lrn2RTFM!!! :p
19:36
warren, rebased
19:36japerry has quit IRC
19:52jammcq has joined #ltsp
19:56
<warren>
Ryan52: do your changes require any files installed in different places or new files?
19:57
<Ryan52>
yes, there are two new files in rc.d/
19:58
X50-dmrc-processing and X99-run-x-session
19:58
both of which need to be installed wherever you put rc.d files...
19:58
<warren>
that's the only new/changed files?
19:58
<Ryan52>
ya.
19:58
<warren>
Ryan52: hmm, it is about time we stop supporting esound and nasd
19:59
<Ryan52>
oh, I didn't add those to the makefile..
19:59
I just reimplemented what was already there :)
20:00* Ryan52 fixed the Makefile.am
20:00
<warren>
when is X run?
20:00
X* scripts
20:01
<Ryan52>
X is ran wherever it currently is ran...the X* scripts are for the "x_session".
20:01
(so connecting to the remote server and running stuff)
20:06
hrm. kdm doesn't even look at the .dmrc wrt language..
20:07
oh, user error, with kde you have to install a separate package for different localizations.
20:07
<warren>
ah
20:10japerry has joined #ltsp
20:19
<warren>
Ryan52: I'm surprised that you implemented the entire thing in shell
20:19
That was a lot of work
20:19
Ryan52: I'm not sure this is the ideal way to do this, we should get more opinions on the list. i'm testing it now.
20:20
Ryan52: was ldminfod changed at all for this?
20:20
<Ryan52>
no.
20:20
<warren>
nice
20:23
Ryan52: oh boy, where did you get the zenity/xmessage/omgwtfbbq idea?
20:23* warren wrote similar shit to this 6 years ago
20:24* warren still wouldn't do anything better
20:24
<Ryan52>
hrm. what else should I have done? :)
20:24
<warren>
Ryan52: you did nothing wrong, I'm just impressed
20:25
<Ryan52>
oh, ok, good :)
20:26
<warren>
Ryan52: what happens if none of those 3 are installed?
20:27
Ryan52: implementing this in ldm itself might be a good thing, because zenity pulls in an additional 14 packages, and the others are ugly.
20:27
<Ryan52>
xorg depends on x11-utils which provides xmessage, so one of them will always be there, at least on Debian...
20:27
ya.
20:29
implementing it in C wouldn't be *that* hard, but it would be a mess, and I think that allowing people to add scripts that actually change ldm's behavior instead of just running stuff is a good thing in general...
20:29
people meaning users
20:30
and we have other stuff using zenity :)
20:31
LDM_LIMIT_ONE_SESSION and LDM_USER_ALLOW both use zenity or xmessage if zenity isn't there.
20:31
though this would be more commonly seen that those..
20:31siki has quit IRC
20:32
<warren>
in the chroot?
20:32
other stuff in the chroot using zenity?
20:32
oh
20:32
<Ryan52>
yup.
20:33
<warren>
zenity is pretty heavy in reqiurements
20:33
xmessage doesn't even print out the entire message
20:34
Your default window session is GNOME, but you choose
20:34
Ryan52: where did you copy this text from?
20:34
<Ryan52>
from my head :p
20:34
<warren>
Ryan52: please get the equiv from gdm source so we can reuse translations
20:35
<Ryan52>
until it supports i18n there's no point to doing that...
20:35randra has quit IRC
20:38
<warren>
Ryan52: does debian's initscripts use gettext to translate strings?
20:38
fedora's initscripts do
20:38
Ryan52: could you change the default behavior, if none of those three dialog apps are installed, save as the new default without asking?
20:39
<Ryan52>
you don't have xmessage installed with xorg?
20:39
<warren>
Ryan52: nope, and xmessage is both ugly and doesn't display the entire message
20:39
<Ryan52>
b
20:39
<warren>
Ryan52: Ryan52: sometime later we should have a generic dialog app implemented as a tiny binary to include in ldm. I really don't want to pull in 14 packages just for zenity.
20:40
Ryan52: xmessage is at least one package, but uses an additional 554KB of space in the chroot
20:40
in our case
20:41ltsppbot has quit IRC
20:42
<Ryan52>
okay, changed.
20:46ltsppbot has joined #ltsp
20:48* Ryan52 isn't feeling well and is gonna go watch TV or sleep or something
20:49
<Ryan52>
warren, can you email teh list for me wrt "we should get more opinions on the list"? :)
21:00alekibango has quit IRC
21:06
<warren>
Ryan52: please don't feel afraid to post to the list, you do awesome work and should take visible "ownership" of it.
21:07
Ryan52: sound and local apps both work
21:07
Ryan52: what else do you need tested?
21:09
<Ryan52>
warren, LDM_SYSTEMWIDE_DMRC
21:09
<warren>
oh right
21:10alekibango has joined #ltsp
21:12
<warren>
Ryan52: hmm, I set it to =kde and it is going into gnome on a new user
21:13
<Ryan52>
you set LDM_SYSTEMWIDE_DMRC to a path to a file on the server that contains "Session=kde"?
21:14
<warren>
oh, you need the full path?
21:14
like /usr/share/xsessions/kde.desktop you mean?
21:14
<Ryan52>
no...
21:14
I mean, it's a path to a file like ~/.dmrc
21:14
<warren>
ooh
21:14
<Ryan52>
(that's what I thought you were asking for..)
21:15
<warren>
yeah, you're right, that is what I was asking for.
21:15
<Ryan52>
if not, then we can put another rc.d script in rc.d :)
21:15
<warren>
My brain is just stupid
21:15
Ryan52: hmm... might be better to do it without a .desktop file
21:15
Ryan52: so the user can configure it all in one place instead of two places.
21:16
Ryan52: like two lts.conf options instead of a lts.conf option and desktop file somewhere
21:16
<Ryan52>
but for different app servers you might want different default session..
21:17
<warren>
hmm
21:17
Ryan52: OK, but I'm going to add a /etc/somewhere/ldm-dmrc somewhere.
21:18
packager can decide if they want to ship it
21:18chupacabra has joined #ltsp
21:19
<warren>
hmm, maybe it isn't working after all
21:19
<chupacabra>
Y'all seen this? http://www.eweek.com/c/a/Virtualization/HP-Offering-Revised-Virtual-Desktop-Infrastructure-Software-Suite/
21:20
<Ryan52>
warren, okay, I'll look at it tomorrow then...
21:20hanthana has joined #ltsp
21:26
<warren>
Ryan52: IMHO, push this to trunk now, it will need more cleanup but it works good
21:26
Ryan52: I'll do more on it tonight and tomorrow
21:27
<Ryan52>
uhhh, what happened to discussing on teh list? :)
21:27
<warren>
Ryan52: forget that
21:28
<Ryan52>
okay :)
21:28* Ryan52 pushes
21:29
<warren>
Ryan52: ok, systemwide works
21:29
<Ryan52>
okay
21:30
oh, crap, bzr sends out an email per commit...that's like 20 emails :p
21:30
<warren>
just do it
21:31
<chupacabra>
some shy away from mail lists.
21:31* Ryan52 remembered that after he saw them in his inbox :p
21:51hanthana has quit IRC
21:55
<Ryan52>
warren, oh, you were supposed to test local devices not local apps...they look too similar.
21:55
<warren>
Ryan52: oh
21:56
testing now
21:57
oops, need real hardware for this
21:58petre has quit IRC
22:00
<Ryan52>
really? why? 0.o
22:00
<warren>
Ryan52: because I haven't bothered to setup USB port isolation
22:01
<Ryan52>
does it really need an actually physical device?
22:01
can't you just use "-hdb blah.img"?
22:01CaScAdE^1arAway has joined #ltsp
22:02
<warren>
I dunno, haven't tried htat
22:03CaScAdE^FarAway has quit IRC
22:03
<Ryan52>
ltspfsmounter runs on the server or the client?
22:04
if it runs on the server like I expect, then my code _should_ work fine. otherwise, I need to change it.
22:04
no testing needed :)
22:04
<warren>
Ryan52: works
22:05
<Ryan52>
and the cleanup thing is ran correctly?
22:05
<warren>
cleaning in what event?
22:05* warren boots it again
22:05
<Ryan52>
if localdevs is on, then it adds "; /usr/sbin/ltspfsmounter all cleanup" to the ssh command..
22:06
should work, if that's the right context for it to be used in :)
22:06
(I have no clue what that actually does)
22:06* Ryan52 should learn more about how the non ldm pieces of ltsp work..
22:07
<warren>
don't know what it does
22:07
but it seemed to have cleaned up ok
22:07
<Ryan52>
okay.
22:08
<warren>
the entire pile seems to work
22:45japerry_cat has joined #ltsp
22:48alkisg has joined #ltsp
23:02japerry has quit IRC
23:27alkisg has quit IRC