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


Channel log from 9 April 2008   (all times are UTC)

00:05cyberorg has joined #ltsp
00:36
<johnny>
warren, are you about?
00:36* johnny doubts it
01:11deavid has joined #ltsp
01:24viking-ice has quit IRC
01:50plamengr has joined #ltsp
02:00praveer_cool has joined #ltsp
02:08muh2000 has quit IRC
02:15viking-ice has joined #ltsp
02:24mccann_ has quit IRC
02:26
<dberkholz>
johnny: presume you're aware of the openrc-9999 ebuild?
02:28
<johnny>
yes
02:28
<dberkholz>
ok, cool
02:28
<johnny>
looks like we need it now
02:28
i'm updating the ebuilds now
02:28
<dberkholz>
i just spent an hour trying to get my wireless working. yay for another night's work wasted
02:28
<johnny>
:(
02:28
that's ok.. i'm still banging away awt the ebuilds
02:29
and..there's also a patch needed for ldm , and the integration of warren's idea for /etc/ltsp/Xsession
02:29
i'm working around them atm, but it would be nice not to
02:30
oh. that reminds me
02:30
ldm hardcodes ubuntulooks in the greeter gtkrc
02:31
question is.. what should it depend on..
02:34
we do have ubuntulooks in portage tho
02:34
i think i'm gonna go for that , until vagrantc comes back
02:38
<dberkholz>
just let it use that for now. that's a little tweak
02:40praveer_cool has quit IRC
02:51
<johnny>
dberkholz, do you really want me mark the client ebuilds ~arch when i'm just going to override them anyways in client's /etc/portage/package.keywords/ltsp dir ?
02:51
<dberkholz>
johnny: yes, because it's one less thing to remember once we add them to the tree
02:52
<johnny>
once we add them to the tree, sure
02:52bobby_C has joined #ltsp
02:52
<johnny>
ok.
02:53
i mean.. override them in the client's package.keywords in the quickstart profile
02:53
btw
02:53
so it will be automatic
02:53
<dberkholz>
johnny: sure. that'll be a more realistic experience for what it would be like for anyone who wanted to set up ltsp using the latest testing stuff
02:54plamengr has quit IRC
02:56
<johnny>
is there some sort of accepted cli to manage /etc/portage yet ?
02:57
i don't know what is recommended these days
02:57
<dberkholz>
johnny: autounmask is nice
02:58
<johnny>
hmm.. requires 4 perl modules
03:00
<tarzeau>
HELP
03:00
why does ltsp set 127.0.0.2 for the hostname of a client?
03:00
into /etc/hosts
03:08mikkel has joined #ltsp
03:41
<mikkel>
Why has LTSP5 been removed from the Fedora 9. Anyone here was working on it ?
03:44praveer_cool has joined #ltsp
03:48
<ogra_cmpc>
mikkel, warren, but he's likely asleep
03:49
<mikkel>
ogra_cmpc: Thanks. Will try and talk to him later.
03:49
warren: Are you awake....
03:53muh2000 has joined #ltsp
04:19Q-FUNK has joined #ltsp
04:27Q-FUNK has quit IRC
04:27exodos has joined #ltsp
05:01elisboa has quit IRC
05:01sep has quit IRC
05:02sep has joined #ltsp
06:02bobby_C has quit IRC
06:16TelnetManta has quit IRC
06:17alekibango has joined #ltsp
06:35makghosh has joined #ltsp
06:35jammcq has quit IRC
06:44muh2000 has quit IRC
06:49TelnetManta has joined #ltsp
07:02cliebow has joined #ltsp
07:02daduke has quit IRC
07:03daduke has joined #ltsp
07:08elisboa has joined #ltsp
07:14mhterres has joined #ltsp
07:17deavid has quit IRC
07:23Guaraldo has joined #ltsp
07:30slidesinger has joined #ltsp
07:52Q-FUNK has joined #ltsp
07:59indradg has quit IRC
08:02plamengr has joined #ltsp
08:12indradg has joined #ltsp
08:14pdjbarber has joined #ltsp
08:21
<warren>
mikkel: it isn't removed.
08:21
<mikkel>
warren: It is not in the FeatureList anymore ?
08:22
<warren>
mikkel: not going to mention it in the release notes because it isn't feature parity with Ubuntu yet
08:22
<mikkel>
warren: Okay, but it will be usable ?
08:22
<warren>
mikkel: it is already usable
08:22
mikkel: weeks ago
08:23
<mikkel>
warren: That is good to here. Looking forward to giving it a spin.
08:29Q-FUNK has quit IRC
08:33
<ogra_cmpc>
warren, do you add any patches to libflashsupport ?
08:33
(in fedora)
08:34
<warren>
ogra_cmpc: I use the libflashsupport from pulseaudio upstream's source repo, hasn't changed in a year
08:34
<ogra_cmpc>
seems we're bitten by http://www.pulseaudio.org/ticket/225 in ubuntu
08:34
(as gentoo is as well, indicated by the reporter)
08:35
warren, i know i was wondering if you ran into that bug and had a fix probably
08:36
<warren>
ogra_cmpc: is there a particular URL that demonstrates this crash consistently?
08:37
ogra_cmpc: we do have lots of random crashes but we haven't been able to isolate them
08:37
ogra_cmpc: and the same is true of flash w/o pulse
08:38
<ogra_cmpc>
our FF maintainer says he can reproduce it all the time with pulse
08:39
<warren>
with what URL?
08:40
ogra_cmpc: you folks should really run plugins in nspluginwrapper even on i386. We've been doing it since Fedora 8. Our browser almost never crashes now.
08:40
it is really good for user support because you don't get users blaming the browser for plugin problems
08:41K_O-Gnom has joined #ltsp
08:42
<ogra_cmpc>
warren, with any url on youtube or google video
08:42
just go back and forward multiple times on video pages
08:42
at some point it will crash
08:42slidesinger has quit IRC
08:46
<warren>
ogra_cmpc: is your crash with or without nspluginwrapper?
08:47
<ogra_cmpc>
without i think
08:47
its not *my* crash :)
08:48
to me it seems something in flash has changed ... since libflashsupport and pulse didnt have any big changes
08:52
<warren>
so this was fine 4 months ago but broke 3 months ago?
08:53
<ogra_cmpc>
http://crash-stats.mozilla.com/report/index/49e43885-e464-11dc-8f26-001a4bd43e5c
08:53
ah
08:57
<warren>
ogra_cmpc: can your FF maintainer come here?
09:02
<ogra_cmpc>
i asked him
09:02
he's pretty busy (as you can imagine bets atm i guess :) )
09:03
*best
09:04asac has joined #ltsp
09:04
<asac>
ole
09:04
warren: ?
09:04
<ogra_cmpc>
hey
09:04
warren, asac is our firefox god in ubuntu ...
09:05
<asac>
s/god/dog/ :)
09:05* ogra_cmpc pokes warren to see if there is still life in there
09:05
<asac>
ill stay for a while.
09:06
<ogra_cmpc>
i guess he fell over ... killed by pre release stress :)
09:06
<asac>
i guess he runs IRC in flash and enabled sound ;)
09:06
<ogra_cmpc>
heh
09:11Dalran has quit IRC
09:17jammcq has joined #ltsp
09:17
<jammcq>
g'morning friends
09:17
<cliebow>
jammcq!!!!
09:18
<jammcq>
hey chuck
09:18
<cliebow>
good to see you..
09:18
busy??..of course
09:18
<jammcq>
yeah, but it's good to be busy
09:19mikkel has quit IRC
09:19
<cliebow>
yeah..tthings are kind of slack here..
09:23cliebow has quit IRC
09:25
<warren>
asac: do you have any procedure that can reproduce the flash crash 100% of the time?
09:27
asac: we have flash crashes, but they are not reproducible consistently, and are not limited to pulse only
09:36
<asac>
warren: yes
09:36
you go to youtube, watch a video
09:36
hit back
09:36
<laga>
"back"? in the browser or in the flash player?
09:36
<asac>
hit forward, wait till sound and repeat these steps until it crashes
09:36
in the browser
09:37
<laga>
ah. yeah, i noticed that as well..
09:37
(i listen to a lot of music on youtube ;)
09:39
<asac>
it doesn't take many attempts here
09:39
sometimes it crashes when going back and forward two times
09:39vagrantc has joined #ltsp
09:39
<warren>
asac: with or without nspluginwrapper?
09:39
<ogra_cmpc>
wb vagrantc
09:39
<asac>
warren: without
09:39
<warren>
asac: where is your libflashsupport source?
09:40
asac: what version of pulse?
09:40
<asac>
warren: its in hardy. but upstream people also did build from source and saw that
09:40
<warren>
I have no idea what hardy is.
09:40
<asac>
let e check
09:40
warren: http://www.pulseaudio.org/ticket/225#comment:4
09:40
that ticket looks really similar
09:41
warren: 0.9.10-1ubuntu1 (pulseaudio)
09:41
<vagrantc>
warren: hardy is the upcoming ubuntu release
09:41
ogra_cmpc: greetings!
09:41
<asac>
warren: please try to just reproduce. chances are hight that youll see the crash
09:41
<warren>
weird, now I have flash sound but no video
09:41
<asac>
hehe
09:42
<ogra_cmpc>
warren, we use the upstream tarball together with https://code.launchpad.net/~crimsun/libflashsupport-pulse/devel
09:42
<warren>
hmm, that does indeed crash it
09:42
<asac>
warren: the libflashsupport source is: https://edge.launchpad.net/ubuntu/hardy/+source/libflashsupport/1.9-0ubuntu1/+files/libflashsupport_1.9.orig.tar.gz
09:42
<warren>
asac: it behaves a little different in nspluginwrapper but it does crash.
09:42cliebow has joined #ltsp
09:43
<asac>
warren: we have one patch on top of that in the package but disabling that doesn't change a thing
09:43
<warren>
you suspect it is a problem in flash itself or libflashsupport?
09:43
<ogra_cmpc>
not as long as you dont use ssl encryption i guess :)
09:43
<asac>
warren: yeah ... i am aware that it crashes both archs. at least someone using amd64 pointed this out initially
09:43
<warren>
asac: since Fedora 8 we use nspluginwrapper even on i386 since it protects the browser from these frequent crashes
09:44
<asac>
warren: i think the cause is libflashsupport ... it might be that it just reveals a bug in flash/firefo
09:44
<warren>
asac: ever since our browser almost never crashes and users never blame the browser anymore.
09:44
<asac>
without libflashsupport we don't get those crashes
09:44
<warren>
we do have crashes without libflashsupport + pulse
09:44
<asac>
warren: good idea. ill consider putting pluginwrapper in between
09:45
warren: well, of course you have crashes. flash crashes in all occassions. but this particular crash only happens with flashsupport
09:45
and its really a big blocker
09:45
as it crashes on the most common use case: youtube
09:45
<warren>
https://code.launchpad.net/~crimsun/libflashsupport-pulse/devel
09:45
you build libflashsupport with gnutls? isn't that a direct license violation?
09:46
<ogra_cmpc>
of which license ?
09:46
<warren>
GPL?
09:46
<ogra_cmpc>
of which code
09:46
<warren>
oh right, you folks have a different set of morals
09:46
<ogra_cmpc>
???
09:47* ogra_cmpc doesnt understand which license of which code warren is referring to
09:47
<warren>
https://code.launchpad.net/~crimsun/libflashsupport-pulse/devel
09:47
<ogra_cmpc>
yes, i see that patch ... what exactly does it violate
09:48
<warren>
you're linking to gnutls to provide a conduit to proprietary software?
09:48
this at least isn't a license violation for openssl
09:48
<asac>
warren: openssl is not compatible with GPL, yes.
09:49
i didn't add that patch and i haven't investigated licensing ;)
09:52
<warren>
I don't see where to download or bzr get the source of this
09:52
<asac>
warren: sorry, what exactly do you want?
09:52
<ogra_cmpc>
there is no branch yet
09:52
<asac>
i posted the tarball
09:52
<warren>
asac: the source of libflashsupport that you are using
09:52
<asac>
https://edge.launchpad.net/ubuntu/hardy/+source/libflashsupport/1.9-0ubuntu1/+files/libflashsupport_1.9.orig.tar.gz
09:52
<ogra_cmpc>
the 30 lines of patch is all there is
09:52
<warren>
oh
09:52
<asac>
thats the source
09:53
<warren>
ogra_cmpc: patch on top of what though? there's 30 different forks out there
09:53
<asac>
we only have that gnutls patch, but not using that doesn't change a thing
09:53nelf has joined #ltsp
09:53
<warren>
there is no "1.9" upstream version
09:53
there is no upstream
09:53sepski has joined #ltsp
09:54
<asac>
warren: do you know where libflashsupport originated? the license looks like that adobe gave birth to it ... but now its hosted at pulseaudio or something
09:54
hmm
09:54
<warren>
asac: adobe wrote the first version then abandoned it
09:54
ever since there's been forks of it everywhere
09:54
<asac>
ok ... i guess we are using the pulseaudio release
09:54
<warren>
with no version numbers or upstream authority
09:54
<asac>
but let me check if there is more info in our copyright
09:54deavid has joined #ltsp
09:54
<asac>
ogra_cmpc: where did you get that tarball from?
09:55
<ogra_cmpc>
from crimsun
09:55
<asac>
warren: what i see is: for sound output of Flash 9 with pulseaudio
09:55
oops
09:55
http://git.0pointer.de/?p=libflashsupport.git;a=summary
09:55
thats
09:55
what is documented at least
09:56viking-ice has quit IRC
09:56Blinny has joined #ltsp
09:59
<warren>
asac: yeah, we're using this code as of September 4th
09:59
<ogra_cmpc>
asac, oh
09:59
<asac>
warren: did you read the ticket i pointed to?
09:59
<ogra_cmpc>
http://git.debian.org/?p=pkg-pulseaudio/flashplugin-nonfree-pulse.git
09:59
<warren>
asac: where did the 1.9 version number come from?
09:59
<ogra_cmpc>
seems debian edu has done some work as well
09:59
<asac>
for sound output of Flash 9 with pulseaudio
09:59
damn
10:00
sorry my paste finger is a bit nervous today
10:00
<ogra_cmpc>
but doesnt seem like they have any fixes
10:01
<warren>
where did 1.9 come from?
10:01
<ogra_cmpc>
no idea
10:04
<asac>
warren: do you see a diff of the tarball above and the version you use?
10:04
otoh, as you see the crash as well i doubt it
10:05
<warren>
the crash is less consistent here
10:05
<ogra_cmpc>
yeah i really think its flash
10:05
<warren>
but we're 100% in nspluginwrapper
10:05
it might not be an apples to apples comparison
10:05
asac: flash 9.0.124.0?
10:05
<asac>
yes. the latest version shows this behaviour and the version before did as well
10:07viking-ice has joined #ltsp
10:17nelf has quit IRC
10:21indradg has quit IRC
10:33K_O-Gnom has quit IRC
10:34Faithful has joined #ltsp
10:38mccann has joined #ltsp
10:40
<vagrantc>
well, i didn't work on the LDM_SESSION stuff on the train, but i did get the cdpinger/ltspfsd from udev working pretty good: bzr get http://llama.freegeek.org/~vagrant/bzr/ltspfs/ltspfs-udev-cdpinger
10:44
<ogra_cmpc>
slacker
10:44
you should have done both !
10:44
:)
10:48
<laga>
ogra_cmpc: let's assume my nbdrootd is runnong on port 2001. do i have to specify nbdport in /var/lib/tftpboot../pxelinux.cfg/default or is it auto-detected somehow when the client boots?
10:48
<ogra_cmpc>
ltsp-update-image should have set it up for you
10:49
<laga>
ogra_cmpc: thought as much. but it hasn't :/
10:50
oh.
10:50* laga scratches head
10:53
<laga>
ogra_cmpc: it seems to me that ltsp-update-image sets up /opt/ltsp/amd64//etc/ltsp/update-kernels.conf while ltsp-update-kernels wants to read /etc/ltsp/update-kernels.conf. i guess something is wrong there?
10:54
<ogra_cmpc>
nope
10:54
there is an in-chroot bit as well
11:03
<vagrantc>
the update-kernels hooks is a crazy zoo that needs to be set free
11:03
release the lions and elephants, and let's start over.
11:05* ogra_cmpc wants a cage first ...
11:05
<ogra_cmpc>
(to sit in until they are gone)
11:13exodos has quit IRC
11:21
<laga>
ogra_cmpc: since there's no DND sign on your cage: it looks like ltsp-update-image updates /opt/ltsp/amd64//etc/ltsp/update-kernels.conf in 095-update-image. but because there's no subsequent ltsp-update-kernels run, the pxelinux config is not updated.
11:21
ogra_cmpc: at least that's what i think ;)
11:23
ogra_cmpc: another (additional) problem is that while /opt/ltsp/amd64//etc/ltsp/update-kernels.conf is written, /etc/kernel/postinst.d/ltsp-update-kernels never runs unless you update the kernel. so even if you run ltsp-update-kernels outside the chroot, it'll still get the old config
11:24
<ogra_cmpc>
which is moot since every run of update-initramfs (i.e. on kernel package instalation) will trigger it
11:26
its used by kernel packages and update-initramfs ... ltsp-update-kernels should only be used to get the new kernel in place
11:26Faithful has quit IRC
11:26staffencasa has joined #ltsp
11:29
<vagrantc>
update-initramfs doesn't trigger it on debian
11:30
though maybe in the future
11:30
<laga>
ogra_cmpc: /etc/kernel/postinst.d/ltsp-update-kernels needs to run at least once after ltsp-update-image in 095-update-image is run, otherwise $CHROOT/boot/pxelinux.cfg/default is not updated.
11:31
<ogra_cmpc>
??
11:31
<laga>
ogra_cmpc: okay. let me rephrase.
11:31
<ogra_cmpc>
i wonder why it works on gutsy then :)
11:32
<laga>
ogra_cmpc: it doesnt work on hardy
11:32
for me, with my mythbuntu plugin. which shouldn't change that part of LTSP
11:32
<ogra_cmpc>
vagrantc, /etc/kernel/postinst.d/ isnt parsed on debian ???
11:35
<warren>
hmm, /etc/kernel/postinst.d/ is new in fedora 9
11:35
I'm not sure how it works yet
11:35
<vagrantc>
ogra_cmpc: only on kernel postinst, not on update-initramfs
11:35
<warren>
only package using it is dell's add-ons
11:36
vagrantc: you use /etc/kernel/postinst.d/ to trigger update-kernels?
11:36
<ogra_cmpc>
vagrantc, hmm, i thought that was integrated into update-initramfs
11:37
<warren>
/etc/kernel/postinst.d/ is part of what exactly?
11:37
what do you use it for?
11:37
when does it trigger?
11:38
<ogra_cmpc>
its parsed and executed by trhe kernel package on install
11:38
<laga>
ogra_cmpc: a quick & dirty solution would be calling $CHROOT/etc/kernel/postinst.d/ltsp-update-kernels to add BOOTPROMPT_OPTS from update-kernels.conf to $CHROOT/boot/pxelinux.cfg/default. afterwards, running ltsp-update-kernels outside the chroot will copy the kernels into place
11:38
ogra_cmpc: could be added to another plugin maybe, dunno
11:38
<ogra_cmpc>
laga, in ltsp-update-image ?
11:38
<vagrantc>
warren: on debian /etc/kernel/post*.d/ltsp-update-kernels is a symlink to /usr/share/ltsp/update-kernels
11:38
<ogra_cmpc>
or where exactly
11:38
<vagrantc>
warren: in the kernel postinst or postrm scripts, it gets executed.
11:38
<warren>
vagrantc: oh, cool
11:39Guaraldo has quit IRC
11:39
<vagrantc>
i basically got it to the point that you could actually configure it such that ltsp-update-kernels server-side wasn't needed.
11:40
by somehow pointing your tftp to the chroot's /boot directory
11:40
but it's an ugly mess, and i'd like to start over at some point.
11:41
ltsp-update-kernels used to call $CHROOT/usr/share/ltsp/update-kernels ... but that was removed by sbalneav or ogra at some point.
11:44staffencasa has quit IRC
11:44
<warren>
I guess I can soon use /etc/kernel/post*.d/ltsp-update-kernels too
11:44
but I want to rewrite update-kernels first
11:44
I really don't like set -x
11:44
I mean set -e
11:46staffencasa has joined #ltsp
11:47
<ogra_cmpc>
warren, thats a standard in debian scripts
11:47
<warren>
why?
11:47
is it wrong to check return values and print error messages?
11:50
<vagrantc>
is it wrong to do both?
11:50
<warren>
uh
11:51
if you have set -e, wont a non-zero return code make it fail before it checks it?
11:51
in the next line
11:51
<vagrantc>
depends on how you check.
11:51
<ogra_cmpc>
Shell scripts (sh and bash) should almost certainly start with set -e so that errors are detected. Every script should use set -e or check the exit status of every command.
11:51
from debian policy
11:52
http://www.debian.org/doc/debian-policy/ch-files.html
11:52
<vagrantc>
at any rate, it's better to not blindly proceed forward if an error is encountered. how you actually handle that is the issue.
11:52
<warren>
hmm
11:54mccann has quit IRC
11:55
<ogra_cmpc>
why oh why doesnt my virtualbox like me today
11:55
grmbl
11:56* vagrantc is dealing with new insane recommends for devscripts: http://bugs.debian.org/474727
11:57
<ogra_cmpc>
vagrantc, tell him to just not use aptitude :)
11:58
<vagrantc>
same problem with apt-get
11:58
as apt-get installs recommends by default
11:58
<ogra_cmpc>
ugh, since when ?
11:58
<vagrantc>
the got it down from installing 199 extra packages to only 53.
11:58
ogra_cmpc: probably 6 months or so
11:58
<ogra_cmpc>
we made that change only for metapackages in ubuntu
11:58
<vagrantc>
maybe only 4
11:59
i think it is correct behavior to have recommends installed by default.
11:59
the only difference between recommends and dependencies is it should be possible to install without satisfying all recommends.
12:00
<ogra_cmpc>
more important is that you can remove unwanted deps with it
12:00
and only keep the essential bits for minor operation
12:01
<vagrantc>
but because recommends *haven't* been handled properly for so long, there's lots of packages that recommend on crufy needlessly
12:01
s,cufy,cruft,
12:01
oh hell, you know what i mean. :)
12:01
<ogra_cmpc>
yeah
12:02
we had that fun at the beginning of hardy
12:02
<vagrantc>
problem is devscripts promoted all of their suggests to recommends recently. which is pulling in esound and X and other crufy.
12:02
cruft.
12:02
<ogra_cmpc>
but luckily only for the metapackages
12:04Guaraldo has joined #ltsp
12:06
<laga>
ogra_cmpc: well, i guess the problem needs to be fixed in ltsp-update-image, so modifiying an existing plugin or adding another one would be the most sensible approach
12:11
<ogra_cmpc>
laga, somewhere i dont get the logic of that sentence ...
12:11
s/somewhere/somehow
12:12
<laga>
ogra_cmpc: that's ok, i've been incoherent all day long.
12:12
<ogra_cmpc>
it should be fixed in ltsp-update-image so we add a new plugin ?
12:12
<laga>
ogra_cmpc: add a new plugin OR modify an existing plugin, eg 095-update-image
12:13
<ogra_cmpc>
well, why not just call it directly from ltsp-update-image ?
12:14
juswt adding a "chroot $CHROOT /etc/kernel/postinst.d/ltsp-update-kernels" should suffice, right ?
12:14
th eplugin only gets executed at build time
12:14
<laga>
ogra_cmpc: chroot $CHROOT /etc/kernel/postinst.d/ltsp-update-kernels && ltsp-update-kernels
12:14deavid has quit IRC
12:14
<ogra_cmpc>
err, wait
12:14
<laga>
ogra_cmpc: the problem only occurs at build time.
12:16
<ogra_cmpc>
hmm
12:16
<laga>
it's "apply changes, modify config file" instead of the other way around
12:19mccann has joined #ltsp
12:20
<ogra_cmpc>
ok, added to 095-update-image right before /proc is unmounted
12:21
<laga>
ogra_cmpc: feel hugged. :)
12:22
ogra_cmpc: i'll get that debdiff to you later tonight, still have to debug something here
12:22
<ogra_cmpc>
damned, do you remember th ebug number from the update-rc.d breakage ?
12:22
laga, well, i can move the fi five lines down ...
12:23
<laga>
ogra_cmpc: i'd like to fix the update-rc.d calls so updates wont break it
12:23
<ogra_cmpc>
indeed if you want to clean up the update-rc.d stuff that needs a debdiff
12:23
but i.d like to haqve installable CDs tomorrow :)
12:23
<laga>
i've actually spent too much time on it already so it has to go in ;)
12:24
<ogra_cmpc>
freeze is tomorrow ...
12:24
<laga>
ogra_cmpc: do you install l-r-m in the chroot?
12:24
<ogra_cmpc>
nope
12:24
<laga>
ogra_cmpc: yes, give me another hour or so.
12:25
<ogra_cmpc>
deliberately not
12:25
it eats about 30M or so
12:25
<laga>
ogra_cmpc: okay. because i do and it mounts /opt/ltsp/amd64/lib/modules/2.6.24-15-generic/volatile which is rather annoying ;)
12:25
ah, right.
12:25
<ogra_cmpc>
right, thats what occupies your ram
12:26
<laga>
i guess there's a reason why it's on a ram disk
12:26
<ogra_cmpc>
licensing afaik
12:26
<laga>
stupid licenses taking away my RAM.
12:26
<ogra_cmpc>
heh
12:27
<laga>
well, giving up 30M is better than having people complaining that their tv-out doesnt work
12:27
<ogra_cmpc>
#ubunt-kernel could tell you more :)
12:27
*ubuntu
12:28
<laga>
no, i'm sure there's a good reason and i'd rather have people work on $stuff instead of answering my silly questions one day before rc freeze :)
12:31
<ogra_cmpc>
well, kernel might have a different deadline
12:31
not sure they fall under RC freeze with kernel fixes
12:32
they might work towards the RC CD though
12:32
<laga>
oh, about the mount. i thought you suggested i complain about the ram disk ;)
12:45indradg has joined #ltsp
12:48
<laga>
ogra_cmpc: just found my problem. forgot a "chroot $CHROOT" somewhere
12:48
which means i get to clean /etc/rc*.d manually on this box
12:50K_O-Gnom has joined #ltsp
12:57Skarmeth has joined #ltsp
13:01rjune has joined #ltsp
13:04muh2000 has joined #ltsp
13:08Topslakr has quit IRC
13:11milesd has joined #ltsp
13:13
<warren>
vagrantc: doesn't your /etc/kernel/postinst.d/ give parameters as $1 and $2 to the scripts?
13:13
vagrantc: (why does update-kernels do it for all kernels?)
13:15deavid has joined #ltsp
13:16topslakr has joined #ltsp
13:23
<vagrantc>
warren: yes, it passes some arguments indicating which version is being installed and such. exactly what $1 or $2 does, i don't recall off the top of my head.
13:24
<warren>
vagrantc: if that's the case, then why does update-kernels loop through all?
13:24
<vagrantc>
warren: i don't remember, but there was a reason.
13:24
<warren>
mind if I change it to it only to the one specified by args if args are supplied, otherwise loop through all?
13:25joebaker has joined #ltsp
13:25
<vagrantc>
well, assuming that the arguments are the passing the same information ... i guess that would be ok. but i'm thinking it's still missing something.
13:26
<warren>
well I'll experiment with it...
13:26
<vagrantc>
overall, i would rather see a complete re-write than breaking what's there now.
13:26
<elisboa>
sono...
13:26
<warren>
OK, I'll do that then.
13:26
I'll use the rewrite, then we can bring it back into common again
13:27
<vagrantc>
sounds perfect.
13:30
<warren>
vagrantc: apparently the documentation doesn't guarnatee that DHCP root-path option is NFS, and apparently iscsi spec uses root-path as well, although I don't see an implementation of it yet.
13:31
vagrantc: http://www.ietf.org/mail-archive/web/ips/current/msg02406.html
13:31
<vagrantc>
yeah, i'm not so against using root-path for something other than NFS ... if you've got multiple things on the same network, you probably need to send them different information via DHCP anyways.
13:32
<warren>
yeah
13:32
apparently I was misled by different opinions back then
13:33
vagrantc: so apparently we should still write an IETF RFC, but it will just specify the standard format for root-path nbd:
13:33
<vagrantc>
jammcq: just re-sent my email regarding LTSP hackfest in july pre/post oscon
13:33
warren: that would be nice!
13:35topslakr has quit IRC
13:37viking-ice has quit IRC
13:37topslakr has joined #ltsp
13:42
<vagrantc>
speaking of nbd, the last upload to debian fixed 4 of my wishlist bugs :) http://packages.debian.org/changelogs/pool/main/n/nbd/current/changelog
13:43
<warren>
"New '-c' option to check whether a device is active" yes!!!
13:43
<vagrantc>
:)
13:43
requires linux 2.6.24+, i think
13:44
<warren>
not a problem here =)
13:45
<vagrantc>
figured i'd at least mention it, though
13:45
warren: so, i'm wondering how to coordinate between our various translation teams ...
13:45
<warren>
vagrantc: doesn't appear that we have an easy way
13:45
vagrantc: given that every distro has their own distro-centric translation system
13:45
<vagrantc>
warren: i'm sort of thinking that we just point them all at each other and have them figure out how they want to handle it.
13:46Gadi has joined #ltsp
13:46
<vagrantc>
oh, i guess i'm still thinking that there's teams of people working on it, but that isn't really always the case.
13:49
warren: also, when klausade was working on the norwegian bokmal translation, he noticed that some strings were basically identical that really should probably be exactly identical ... i.e. the user interface reports an error one way, and then ldm logs the error a different way ...
13:49
<warren>
yeah
13:49
that needs to be cleaned up upstream first
13:49
<vagrantc>
that could probably cut the strings down by 1/3rd, and even possibly make the code a little simpler.
13:52viking-ice has joined #ltsp
13:52* klausade misses hacking at lunch with red wine ...
13:54indradg_ has joined #ltsp
13:54* warren can't remember why he built a 32bit nbd-client in the 64bit package...
13:55indradg__ has joined #ltsp
13:57indradg has quit IRC
13:57indradg_ has quit IRC
13:57indradg__ is now known as indradg
14:07
<warren>
vagrantc: huh? http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=471591
14:07
vagrantc: you already use dhcp root-path?
14:07
<vagrantc>
klausade: i just can't forget one of the happier moments of my life: walking out of the fancy hotel restaurant barefoot, with a glass of wine in one hand, and a full bottle in the other ...
14:08
warren: that just gets next-server
14:08
<warren>
oh
14:08
next-server is only IP address?
14:08Blinny has quit IRC
14:09
<vagrantc>
warren: right
14:09
<warren>
vagrantc: is the usual purpose of next-server?
14:09
<vagrantc>
warren: well, next-server is when you have DHCP and tftp hosted on different servers.
14:10
warren: unless you're doing weird stuff, tftp and NFS/NBD will be on the same machine.
14:10elisboa has quit IRC
14:11
<vagrantc>
warren: the dhcp client used, klibc's ipmasq, goes so far as to call it $ROOTSERVER ...
14:15
warren: it wouldn't be hard to patch initramfs-tools to get the nbd port from root-path either ...
14:17
<warren>
vagrantc: let's define a standard for root-path syntax first
14:17
<vagrantc>
well sure :)
14:22plamengr has quit IRC
14:24
<laga>
ogra_cmpc: https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/214481
14:24
ogra_cmpc: hope this patch will cause less yelling ;)
14:25* vagrantc watches ltsp hit #7 in the buildd queue
14:25
<vagrantc>
will actually have ltsp, ldm and ltspfs all in sync between debian testing and unstable ... for a few minutes :)
14:26milesd has quit IRC
14:39
<Gadi>
warren: next-server is also needed if you use IP-helper subnets
14:40
to identify to the tftp client where to grab stuff
14:40
<warren>
IP-helper subnets?
14:41
what does this mean?
14:41
<Gadi>
ip-helper is a function on cisco routers to route broadcast dhcp requests across subnets
14:41
well, broadcast requests in general
14:41
and is not limited to cisco
14:42
:)
14:42
just, afaik, all cisco routers have the capability
14:42
<warren>
ok
14:42
<Gadi>
its for larger deployments
14:44Gadi has left #ltsp
14:59Q-FUNK has joined #ltsp
15:11* johnny builds a chroot from scratch
15:13
<cliebow>
go johnny go....go
15:14* johnny doubts it will be finish up completely the 2nd time either
15:14
<johnny>
first time.. an upstreams snapshot got removed, so i had to bump the version of our kernel generator
15:15
<cliebow>
;-]
15:15
<johnny>
so.. now the kernel is building
15:16joebaker has quit IRC
15:18* vagrantc releases a few versions while johnny builds a kernel
15:18
<johnny>
vagrantc, i have some issues with ldm from scratch
15:18
LDM_THEME is not set ?
15:19
<vagrantc>
should default to /usr/share/ldm/themes/default
15:19
<johnny>
but ldm only distributes ltsp
15:19
<vagrantc>
well, on debian/ubuntu that symlink is a configurable symlink to whatever your selected theme is
15:20K_O-Gnom has quit IRC
15:20
<johnny>
but it should be able to work out of the box imo
15:20
<vagrantc>
maybe it should fall back to the ltsp theme in absence of any other theme.
15:20
<johnny>
yes :)
15:20
<vagrantc>
so fix it :P
15:21
<johnny>
my C fu is weak.. i'll see what i can do
15:21
i like to talk about things before jumping right in anyways :)
15:21
<vagrantc>
mine is too, and i've managed to do all sorts of evil with ldm and ltspfs :)
15:21
<johnny>
there goes the kernel
15:22
building a kernel takes no more than 15 on this default config.. i haven't tweaked it at all yet
15:22
logger.
15:23
<vagrantc>
speaking of loggers ... sounds like debian may switch to rsyslog by default
15:23
<johnny>
yeah.. i'll look into that at some point
15:23
i'm just going with what i have working now
15:23
which is syslog-ng
15:25
vagrantc, did you notice that i have ldm working now?
15:25
i don't know if you were here
15:25makghosh has quit IRC
15:25Q-FUNK has quit IRC
15:25makghosh has joined #ltsp
15:26
<vagrantc>
johnny: yay!
15:26
<johnny>
so.. now i'm testing a from scratch chroot build
15:27
altho it doesn't handle the /var/lib/tftpboot stuff yet
15:27
nor nbd
15:27
<vagrantc>
nbd is over-rated.
15:27
tftp is a necessary evil.
15:27
<johnny>
there must be a reason it is used right?
15:28
<vagrantc>
it's a bit faster than NFS ...
15:28
<johnny>
logger done, now updating the system target
15:28
X is gonna be the annoying part..
15:28
<vagrantc>
but because you have to regenerate the NBD image for nearly every change, i wouldn't want to use it during development.
15:29
<johnny>
yes.. truly now
15:29
err not
15:29
<vagrantc>
how'd you get LDM without X ?
15:29
<johnny>
i did have X
15:29praveer_cool has quit IRC
15:29
<johnny>
i built a basic chroot.. and just kept updating my installer script forall the modifcations i made locally to get it to work
15:29
and now i'm installing it from scratch
15:30
<stgraber>
vagrantc: nbd just exports a block device, you can use an ext3 image if you like :)
15:30
<johnny>
stgraber, he uses ext2..
15:30
ok.. 30 packages to upgrade..
15:31
<vagrantc>
stgraber: yes, i have a plugin to generate an ext2 image... but modifying the NBD image directly while other thing shave it mounted is ... evil.
15:31joebaker has joined #ltsp
15:31
<vagrantc>
http://sourceforge.net/mailarchive/message.php?msg_name=20080409183034.GH26187%40ryukin.fglan
15:34cliebow has quit IRC
15:34
<johnny>
c'mon gnu MIRROR
15:36praveer_cool has joined #ltsp
15:37
<johnny>
lame.. the mirror script didn't check any other mirror.. just the same one :(
15:37
i manually downloaded it and injected it forn ow
15:46TelnetManta has quit IRC
15:47muh2000 has quit IRC
15:52joebaker has quit IRC
15:58deavid has quit IRC
16:06Egyptian[Home1 has quit IRC
16:08Egyptian[Home] has joined #ltsp
16:11Skarmeth has quit IRC
16:17
<johnny>
fail.. let's try again..
16:17
missed a few things
16:29sepski has quit IRC
16:30mhterres has quit IRC
16:35bobby_C has joined #ltsp
16:37J45p3r has joined #ltsp
16:55J45p3r_ has joined #ltsp
17:07cpunches2 has joined #ltsp
17:12J45p3r has quit IRC
17:17
<johnny>
ok.. xorg and friends are building :)
17:18Guaraldo has left #ltsp
17:27vagrantc has quit IRC
18:01bobby_C has quit IRC
18:06nickel2 has joined #ltsp
18:13joebaker has joined #ltsp
18:35jammcq has quit IRC
18:43J45p3r_ has quit IRC
18:43J45p3r has joined #ltsp
18:48milesd has joined #ltsp
18:56cesar_ has joined #ltsp
18:56
<cesar_>
hi
18:56
people
18:56staffencasa has quit IRC
18:56
<cesar_>
i have problems with local devices in WS
18:56nickel2 has left #ltsp
18:57
<cesar_>
i can check in a troubleshoting
18:58
i pass nine points of this troubleshooting
18:59
i can pass step ten
18:59
this is the point to pass
18:59
http://wiki.ltsp.org/twiki/bin/view/Ltsp/LTSP-42-LocalDev#Step_10_Does_your_DISPLAY_variab
19:00
my distro in the TS is this
19:01
Linux ltsp 2.6.18-5-amd64
19:03
and the version of ltsp package in my debian is ltsp-4.2
19:04
when i execute this "echo $DISPLAY"
19:04
i don't see anything in the screen
19:07
i need see "ws002:0.0" when i execute "echo $DISPLAY" on the WS terminal but this not appear thee
19:07
there
19:15milesd has quit IRC
19:16TelnetManta has joined #ltsp
19:29
<johnny>
i don't know anything about ltsp4
19:29
UGGH.. portage mirrors are really bad today
19:35
<cesar_>
ok..
19:35vagrantc has joined #ltsp
19:36
<cesar_>
i could pass nine steps but in step ten
19:37
i can't get information with "echo $DISPLAY" in the WS
19:38
the information have say me something that "ws002.ltsp:0.0" but nothing is showed on the screen
19:41
and.. i can see in the desktop of the workstation client a icon what apparently have mounted the floppy drive .. but when i try see the content of this drive i do double click over this and konqueror is opened but ... anything is into the floppy drive.. the floppy is empty
19:43
i don't know how to fix that.. i was finding in google and forums for a answer what help me but.. i didn't find anything
19:44
although other devices are working.. for example i can configure a laser printer and i could print without problems
19:49
ok... i am thinking what doing now.. what can i trying to fix that
19:50
but.. i don't find the response
19:50
<vagrantc>
be patient
19:50
<cesar_>
:-)
19:50
yes.. but.. i was finding a lot on google
19:51
and i didn't find anything.. nothing is clear
19:51
is a big mistery... :-)
19:51* vagrantc missed the beginning
19:52
<vagrantc>
cesar_: basically, local devices aren't working for you?
19:52
<cesar_>
yes
19:52
<vagrantc>
cesar_: what linux distro and release?
19:52
<cesar_>
on the server?
19:52
<vagrantc>
well, we'll start with that, yes.
19:53
<cesar_>
Linux ltsp 2.6.18-5-amd64 #1 SMP Sat Dec 22 20:43:59 UTC 2007 x86_64 GNU/Linux
19:53
<vagrantc>
what linux distro?
19:53
<cesar_>
debian
19:53
<vagrantc>
what release?
19:53
etch?
19:54
!vers
19:54
<ltspbot>
vagrantc: Error: "vers" is not a valid command.
19:54
<vagrantc>
ltspbot: factoids search ver
19:54
<ltspbot>
vagrantc: 'serversize' and 'ver'
19:54
<vagrantc>
!ver
19:54
<ltspbot>
vagrantc: "ver" is to get version info on debian/ubuntu (please use the pastebot): COLUMNS=200 dpkg -l 'ltsp*' | awk '/^ii/{print $2 $3}' ; COLUMNS=200 dpkg --root=/opt/ltsp/i386 -l 'ltsp*' ldm | awk '/^ii/{print $2 $3}'
19:54
<cesar_>
yes
19:55
<vagrantc>
cesar_: please try that command and paste the output to the pastebot:
19:55
!pastebot
19:55
<ltspbot>
vagrantc: "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.
19:56
<cesar_>
vagrantc: what command ?
19:56
<vagrantc>
!ver
19:56
<ltspbot>
vagrantc: "ver" is to get version info on debian/ubuntu (please use the pastebot): COLUMNS=200 dpkg -l 'ltsp*' | awk '/^ii/{print $2 $3}' ; COLUMNS=200 dpkg --root=/opt/ltsp/i386 -l 'ltsp*' ldm | awk '/^ii/{print $2 $3}'
19:58
<ltsppbot>
"cesar_" pasted ""ver"" (3 lines) at http://pastebot.ltsp.org/501
19:58
<cesar_>
http://pastebot.ltsp.org/501
19:58
<vagrantc>
cesar_: the ltsppbot will tell us :)
19:59
cesar_: looks like you've installed ltspfs from lenny or sid
19:59
<cesar_>
yes.. i've just see that.. great.. i didn't knew that
19:59
<vagrantc>
cesar_: also, i have no idea how ltsp 4.x (which is what ltsp-utils creates) works with that version of ltspfs, which was made for working with ltsp 5
19:59
<cesar_>
sorry.. my distro is lenny
20:00
<vagrantc>
ltsp-utils isn't in lenny, so you're using some of both.
20:00
<cesar_>
ohhh .. i installed this package because
20:00
is a package .deb
20:01
<vagrantc>
!debian
20:01
<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
20:02
<cesar_>
yes..... well.. so i need see this page now
20:02
<vagrantc>
ltsp-server can be used to create an LTSP installation out of debian packages.
20:02
which is the ltsp 5 style of doing things.
20:03
cesar_: you'll probably want to look at the Howto mentioned there.
20:03
<cesar_>
well.. i will try fix this installation
20:03
<vagrantc>
if you must...
20:03
it's basically unsupported
20:04
<cesar_>
:-(
20:04
so...... this is a big problem
20:04
<vagrantc>
you're stuck in 2006, i think.
20:05
i think that's the last release of ltsp 4.x
20:05
<cesar_>
:-/
20:05
<vagrantc>
there's active work on ltsp5
20:06
in fact, a new version of ltsp-server should hit lenny within the next couple days :)
20:06
<cesar_>
of course.. all say wath i need change my version of ltsp
20:11
vagrantc: thank you for your advice
20:12
i will think what do with that
20:12
bye
20:12
<vagrantc>
some of the ltsp5 developers for debian are also brazillian
20:12
although they're not as active anymore
20:13
<cesar_>
but brazilians ...
20:13
i don't know yours works
20:13
<vagrantc>
oh, your error message was in spanish :)
20:13
<cesar_>
and i don't know portugeis language
20:14
i prefer read information in spanish or english
20:14
<vagrantc>
halbo un poco
20:14
<cesar_>
but not in portugeis
20:15
<vagrantc>
yes, i understand now.
20:15
<cesar_>
yes.. you speak some words of spanish
20:15
<vagrantc>
entiendo.
20:16
<cesar_>
the developers of Brasil write man and information in portugeis
20:16
90 %
20:16
<vagrantc>
si.
20:16
<cesar_>
and 10% remember write for the rest of the world
20:16
:-D
20:16
<vagrantc>
cesar_: i mis-read your error message and thought it was portuguese. lo siento.
20:17
<cesar_>
no no.. my message error
20:17
was in spanish
20:17
<vagrantc>
si, entiendo ahora.
20:17
<cesar_>
argie-spanish
20:17
:-D
20:17
yes..
20:17
yo tambien entiendo
20:18
<vagrantc>
espero voy a argentina en augusto.
20:18
<cesar_>
vagrantc :where are you from?
20:18
in august?
20:18
for Cafeconf?
20:18
<vagrantc>
cesar_: yes, the debian conference is in mar del plata.
20:18
<cesar_>
in Cafeconf in Buenos Aires?
20:18
ohh.. yes
20:19
<vagrantc>
don't know about cafeconf
20:19
<cesar_>
debian developers metting
20:19
Cafeconf are forum about free software in Buenos Aires
20:20
and .. in two or three days in Cafeconf you can hear a lot of types of conferences
20:20
<vagrantc>
cesar_: you from buenos aires ?
20:20
<cesar_>
no.. i am from Santa Fe city in Santa Fe province
20:20
:-D
20:20
this is my blog
20:21* vagrantc has been to santa fe ... santa fe, new mexico
20:21
<cesar_>
cesarediaz.blogspot.com
20:21
no no..this is in Argentina
20:21
<vagrantc>
yes, i know. :)
20:21
<cesar_>
i think there are three cities with name 'Santa Fe' in the world
20:22
or maybe there are other
20:22
and i don't know
20:23mccann has quit IRC
20:25cesar_ has quit IRC
20:32joebaker has quit IRC
20:43supreme has joined #ltsp
20:43
<supreme>
hi all
20:46J45p3r has quit IRC
20:47joebaker has joined #ltsp
20:54viking-ice has quit IRC
20:56
<supreme>
hi
20:56
I finally discovered the problem with session restarting in a intel chipset !!!
20:56
<warren>
?
20:56
<supreme>
it's caused by e1000 module
20:57
warren, i got an 'Corrupted mac on input' in auth.log after randomly restart xsession in terminals
20:57
<warren>
supreme: oh, that could be the power saving bug from the BIOS
20:57
<supreme>
:o
20:57
<warren>
supreme: newer kernels "fixed" that by turning off one type of power management that never did work right
20:58
talk to your own distro's kernel people to see if they fixed it
20:58
<supreme>
actually i got this error with edubuntu 7.10
20:58
in an intel mobo
20:58
<warren>
I don't care =)
20:58
I'm just stating that was one possible cause of that error for e1000
20:59
<supreme>
i tell you cause you tell me that this problem could be fixed by newer kernels
20:59
<warren>
I have no idea what 7.10 is
20:59
<supreme>
edubuntu...
20:59
<vagrantc>
released in the 10th month of 2007
21:00
<warren>
yeah, I think this kernel bug was fixed late 2007
21:00
it *could* be something else
21:01
<supreme>
there's a way to fix it disabling something in bios?
21:02
what do you think warren ?
21:04
<warren>
supreme: the machines I saw, there was no way
21:05
supreme: at the time there were several different proposed patches to fix it from various places
21:05
supreme: they all involved rebuilding the kernel though
21:06
<supreme>
hummmm
21:07
that is a critical problem :S
21:08viking-ice has joined #ltsp
21:13jammcq has joined #ltsp
21:19viking-ice has quit IRC
21:21mccann has joined #ltsp
21:39praveer_cool has quit IRC
21:46vagrantc has quit IRC
21:49makghosh has quit IRC
21:49makghosh has joined #ltsp
22:10alekibango has quit IRC
22:10alekibango has joined #ltsp
22:21
<johnny>
dberkholz, i'm testing a from scratch chroot now
22:21
i keep having troubles with file downloads tho
22:22
makes me think we should do some -f first
22:22
i've found a few bugs in the process, i didn't unmask openrc correct, and i had to add the cups use flag
22:22
so updated profile is forthcoming
22:31
or rather.. negate the cups use flag :)
22:40spectra has quit IRC
22:46alekibango has quit IRC
22:46alekibango has joined #ltsp
22:48cpunches2 has quit IRC
22:51cyberorg has quit IRC
23:29supreme has quit IRC
23:54mccann_ has joined #ltsp
23:54asac_ has joined #ltsp
23:56mccann has quit IRC