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


Channel log from 12 March 2008   (all times are UTC)

00:06Egyptian[Home] has joined #ltsp
00:27sepski has joined #ltsp
00:28
<johnny>
ogra, ogra_cmpc you about yet?
00:31
or any ltsp maintainer for that matter?
00:42rjune__ has joined #ltsp
00:43rjune has quit IRC
00:44cyberorg has quit IRC
01:07cyberorg has joined #ltsp
01:08subir has quit IRC
01:09subir has joined #ltsp
01:12subsume has quit IRC
01:17rjune__ has quit IRC
01:18rjune__ has joined #ltsp
01:45sepski has quit IRC
02:08basanta has joined #ltsp
02:11deavid has joined #ltsp
02:19basanta has quit IRC
02:35Shingoshi has quit IRC
02:35Shingoshi has joined #ltsp
02:45mikkel has joined #ltsp
03:17Nubae has joined #ltsp
03:25ogra_cmpc_ has joined #ltsp
03:25ogra_cmpc has quit IRC
03:35bjorn has quit IRC
03:35bjornh` has joined #ltsp
03:43Faithful has joined #ltsp
04:05bobby_C has joined #ltsp
04:32basanta has joined #ltsp
05:12ogra_cmpc_ is now known as ogra_cmpc
05:42Nubae has left #ltsp
05:52spectra has quit IRC
06:01Q-FUNK has joined #ltsp
06:24basanta has quit IRC
06:43Q-FUNK has quit IRC
06:54exodos has joined #ltsp
06:58mhterres has joined #ltsp
07:01Blinny has joined #ltsp
07:07cliebow_ has joined #ltsp
07:08edistar_ has joined #ltsp
07:11jammcq has quit IRC
07:20edistar has quit IRC
07:21steph_ has joined #ltsp
07:34otavio has joined #ltsp
07:36Guaraldo has joined #ltsp
07:37ogra_cmpc_ has joined #ltsp
07:37ogra_cmpc has quit IRC
07:41elisboa has quit IRC
07:52krishna_ has joined #ltsp
08:21mikkel has quit IRC
08:24mikkel has joined #ltsp
08:25Gadi has joined #ltsp
08:26steph_ has quit IRC
08:29krishna_ has quit IRC
08:39nicoAMG has joined #ltsp
08:40jammcq has joined #ltsp
09:05steph_ has joined #ltsp
09:13ogra_cmpc_ has quit IRC
09:16ogra_cmpc has joined #ltsp
09:34ogra_cmpc__ has joined #ltsp
09:35ogra_cmpc has quit IRC
09:37ogra_cmpc__ is now known as ogra_cmpc
09:44vagrantc has joined #ltsp
09:44vlt_ is now known as vlt
09:46steph_ has quit IRC
09:49steph_ has joined #ltsp
09:52Q-FUNK has joined #ltsp
09:52warren_2 has joined #ltsp
09:53
<warren_2>
my laptop is dead
09:53
I guess I'll work on mkdst for now.
09:53
don't need laptop
09:53
<cyberorg>
warren_2, hi, why are there two of you :)
09:54
<warren_2>
cyberorg, clones
09:54
cyberorg, I mean, fedora is that advanced.
09:55
<vagrantc>
warren_2: so, have you basically done most of the work you're going to do on the *-trunk branches for this release cycle?
09:55
<cyberorg>
warren, i've got FC8 packages all building along with suse all from the same spec and taarballs
09:57
<ogra_cmpc>
warren, get a classmate pc :P
09:57
that educates you and makes an awesome second laptop
09:58
<Q-FUNK>
warren: what was the fuss about March 27?
09:58
<warren_2>
vagrantc, 100% of it into trunk
09:58
Q-FUNK, it was jordan's idea
09:59tux_440volt has joined #ltsp
09:59subhodip_ has joined #ltsp
09:59
<vagrantc>
warren_2: alright. i'll be working on a new debian upload this week, and just wanted to check if the flurry of activity would quiet down... though it seemed like most of the changes were specific to fedora
09:59
<warren_2>
Q-FUNK, my idea was there are no known regressions currently so why not release, Jordan wanted to fix the known problems (which are different from regressions) first. We agreed upon a compromise of a schedule based release.
10:00subhodip_ has left #ltsp
10:00
<warren_2>
vagrantc, I'm going to do heavy work on mkdst now, I'll let you know if I plan on doing more on ltsp-trunk
10:00
vagrantc, the plans for mkdst are on the list
10:00
<vagrantc>
warren_2: yeah... i kept meaning to reply to some of that ...
10:01
best do it sooner than later :)
10:01
<warren_2>
vagrantc, well, first I'm doing the proposed changes without the plugin implementation
10:01
because I'm not really sure how to do the plugin-based commands yet
10:01
and I rather get the proposed functionality working first
10:02mccann has joined #ltsp
10:03
<ogra_cmpc>
"<warren_2> cyberorg, I mean, fedora is that advanced"
10:04
its even *that* advanced ...
10:04
http://img.worsethanfailure.com/images/200802/error'd/okquestion.jpg
10:04
:)
10:05
<warren_2>
ogra_cmpc, lol
10:06
<ogra_cmpc>
:)
10:06
i knew i could entertain you with that :0
10:06
<warren_2>
ogra_cmpc, we had a better one somewhere
10:06
<cyberorg>
ogra_cmpc, that is way advanced, OK in ncurses :))
10:06
<warren_2>
ogra_cmpc, you know the openoffice save dialog when you try to save in microsoft DOC
10:07
<ogra_cmpc>
i rarely use oo.o
10:07
<warren_2>
ogra_cmpc, the buttons say "Cancel" "OK" normally. But the buttons at one point during rawhide said, "Do what I want" "Don't do what I want"
10:07
ogra_cmpc, to mock the fact that that dialog is confusing
10:08
<ogra_cmpc>
lol
10:09daduke_ has joined #ltsp
10:09
<Gadi>
ogra_cmpc: have you been keeping ur list of lts vars up to date?
10:10
<ogra_cmpc>
Gadi, up to gutsy, yes
10:10
but not recently
10:10
<Gadi>
is it still the one on launchpad?
10:10
actually, do you have a url?
10:10
<ogra_cmpc>
the code?
10:10
<Gadi>
no, the list
10:10
<cyberorg>
guys, we've got almost everything working as expected, will be doing a development release soon, anything major going into trunk?
10:10
<vagrantc>
ogra_cmpc: speaking of up to date ... where's your debian dirs? :P
10:11
<ogra_cmpc>
https://launchpad.net/~ltsp-upstream
10:11
vagrantc, ergh still not uploaded ....
10:11
<warren_2>
cyberorg, with the new mkdst will come the distro specific specfile support
10:11
<laga>
ogra_cmpc: talking about uploads... ;)
10:11
<warren_2>
cyberorg, so spec files will fork
10:11
<ogra_cmpc>
i'm knee deep in classmate stuff until the end of the week
10:11
<Gadi>
ogra_cmpc: the list ...
10:11
:)
10:11
<ogra_cmpc>
Gadi, its in the code
10:11
<Gadi>
ah, ok
10:11
<ogra_cmpc>
in the client dir
10:11
<Gadi>
in the docs
10:12
ok
10:12
<cyberorg>
warren, for now i won't be using mkdst, spec files on the build service is sufficient for our purpose
10:12
<ogra_cmpc>
if nobody dropped it
10:13
<Gadi>
got it
10:13
danka
10:14
<vagrantc>
warren_2: i know you're not thrilled at the idea ... but something you said about having the debian dirs in upstream sources is starting to happen with the .spec files ... i see a bit of information overload with many commits on the .spec files irrelevent to debian ...
10:14
<warren_2>
vagrantc, I was never against a debian dir in upstream sources
10:15* ogra_cmpc actually remebers the same
10:15
<vagrantc>
warren_2: i seem to recall you complaining about all sorts of irrelevent commits
10:15
<warren_2>
vagrantc, I was against debian dir ending up in the tarball
10:15
<cyberorg>
vagrantc, yeah, i think all distro specific things should not be in upstream
10:15
<vagrantc>
ah.
10:15
<warren_2>
vagrantc, ok, I can redesign mkdst so the spec file is elsewhere.
10:15
<vagrantc>
well, some things belong in upstream ... but i kind of feel like packaging should be kept separate
10:16
<ogra_cmpc>
warren_2, as long as we only have one rpm based distro all is fine ...
10:16
<warren_2>
vagrantc, but there's still a ton of fedora specific scripts and stuff in ltsp-trunk
10:16
<vagrantc>
warren_2: --spec-file /path/to/foo.spec
10:16
<ogra_cmpc>
as soon as the second one comes you got probs
10:16
<vagrantc>
warren_2: yeah, that's fine ...
10:16
<warren_2>
OK, I'll think about a redesign.
10:16
vagrantc, something more automatic than that.
10:17
<vagrantc>
warren_2: ok. thanks for considering it :)
10:17
<cyberorg>
i had a huge problems in finding exactly how much was relevant for me from whole bunch of scripts there
10:17
<warren_2>
vagrantc, it was a great aid to get it whipped into shape rapidly before it was included in fedora, but now it will be a pain to maintain the spec in two places anyway
10:17
cyberorg, which is why I still want to do a directory reorg within ltsp-trunk
10:18
<vagrantc>
warren_2: ah, so it'll be good for you as well ... great :)
10:18
<warren_2>
vagrantc, ogra_cmpc: about that, are any of the moves I proposed earlier going to happen?
10:18
<cyberorg>
warren_2, https://forgesvn1.novell.com/viewsvn/kiwi-ltsp/branches/suse-11.0/ltsp-server/
10:18
those are all that i am using
10:18
<vagrantc>
warren_2: were they posted to the list ?
10:18
<warren_2>
vagrantc, a while ago
10:18* warren_2 finds it
10:18
<vagrantc>
i just want to refresh my memory
10:19
<warren_2>
vagrantc, I want things debian/ubuntu specific to be moved into their own directories
10:19
clean up the directory structure
10:19
more hierarchal
10:19
and formalized
10:19
cyberorg,
10:19
* xinetd configurations have all disabled=yes, it should be
10:19
disabled=no, so the services become active when we install those
10:19
files.
10:19
cyberorg, can't do this, against Fedora policies.
10:20
cyberorg, we can either make distro specific xinetd files, or RPM can frob it during rpmbuild
10:20
frob is a verb
10:20
<vagrantc>
warren_2: well, i think client/initramfs -> client/initramfs-tools ... it's not debian-specific, it's software specific
10:20
<warren_2>
frob (v) to frob
10:20
vagrantc, please suggest better names or structure ideas?
10:20
<vagrantc>
warren_2: i also wanted to get the "boolean_is_true" function available in both client and server ...
10:21
<warren_2>
vagrantc, yeah, that'd be good.
10:21
<vagrantc>
warren_2: i haven't found the posting ... you do have a subject i can look for?
10:21
<warren_2>
Subject: You are all fools
10:21
oh wait, wrong list.
10:21
<vagrantc>
ah, now my memory is refreshed
10:22
"A few notes..."
10:22
<warren_2>
yeah
10:22
<Blinny>
I haven't heard 'frob' since AberMUD days
10:22
<warren_2>
Blinny, standard vocabulary here
10:22
<cyberorg>
warren_2, see ltsp-server.spec i am using, it just has the following for suse %if 0%{?suse_version}
10:22
# %__mv ltsp-server/* %{buildroot}/
10:23
<warren_2>
Blinny, our Intranet interface is a MUD
10:23
cast spell vpnclient
10:23
<cyberorg>
everything goes where it is supposed to go, instead of countless %__install -m
10:23
<warren_2>
cyberorg, ok well, I'm in the process of removing spec files from the upstream trees entirely
10:24
cyberorg, except perhaps the really simple packages *IF* all distros are happy with the common spec file
10:24
<Blinny>
warren_2: That's awesome. Do you have PK enabled? (;
10:24
<cyberorg>
here is the spec which has both fedora as well as suse, want me to pastebin it? https://build.opensuse.org/package/view_file?file=ltsp-server.spec&package=ltsp-server&project=home%3Acyberorg%3Altsp
10:24
<warren_2>
cyberorg, please point me to spec diffs for mkdst, ldm, ltspfs
10:24
Blinny, yes.
10:24
cyberorg, doh, can't connect to your URL
10:24
<vagrantc>
warren_2: so, is "A few notes..." the only thread?
10:25
<warren_2>
cyberorg, firefox3 is really anal about http certs
10:25
vagrantc, that's very old now, we should make a new list of ideas
10:25
<cyberorg>
warren_2, ok, pastebining it
10:25
<vagrantc>
warren_2: yeah, that sounds like a good idea ...
10:27steph_ has quit IRC
10:28
<vagrantc>
warren_2: seems like the issues raised were: some stuff in ltsp-trunk/server/* should maybe be moved to sub-dirs ...
10:28
<Gadi>
ogra_cmpc: I notice all the serial PRINTER_ vars are still there, but since the move to jetpipe we dont support serial printers anymore, right?
10:29
<ogra_cmpc>
Gadi, scott had plans thats why he left it in (we discussed that)
10:29
<Gadi>
ah, ok
10:29
<ogra_cmpc>
but since he dropped off the face of the earth we should probably pull it out for nbow
10:30
<cyberorg>
warren_2, wget http://cyberorg.kicks-ass.org/ltsp/ltsp-specfiles.tar.bz2
10:30
<Gadi>
sorry if I forgot - too little brain memory - need upgrade
10:30
<ogra_cmpc>
if he comes back we'll have more pressing issues than serial printers
10:30
<warren_2>
cyberorg, could you please send that to the list, I'm on a temporary computer and I have to reboot now
10:30
<vagrantc>
warren_2: one of the things that i'm hesitant about setting up distro-specific dirs for everything ... is that i want to move *as much as possible* to be common across distros ... while still maintaining autonomy ... so, while i recognize we'll need to fork on numerous issues ... i guess i just want to try and keep the sentiment of aiming towards common code as much as possible
10:31
<warren_2>
vagrantc, let me put it a different way
10:31
vagrantc, we need distro specific things now
10:31
vagrantc, and the only way we're going to get common is if we figure out what is common
10:31
vagrantc, every step of the way I've been trying to use your stuff, failing, then figuring out where i need my own thing.
10:31
<vagrantc>
warren_2: agreed. never disagreed.
10:31
<warren_2>
vagrantc, when I do write my thing, I try to split out everything that *could* be common
10:32
<cyberorg>
is tarball allowed on ltsp-devel list?
10:32
<vagrantc>
warren_2: i'm not critiquing how it has gone, i just think there is a danger that we need to be aware of.
10:32
<warren_2>
vagrantc, so I'd like to organically grow into distro-specific dirs with common split out, to eliminating duplication and getting rid of distro specific
10:32
cyberorg, just post the URL
10:32
cyberorg, with description
10:32
<vagrantc>
warren_2: that sounds good.
10:33
<warren_2>
vagrantc, i've been very careful and probably too slow in this methodology of 100% in trunk
10:33
vagrantc, but it might be best off for everyone if everyone puts their cards in the middle
10:34
vagrantc, at first overlapping cards, then figure out the overlapping cards and eliminate, each checkin incremental and trying not to break anybody.
10:34
<vagrantc>
warren_2: sounds great.
10:34
<warren_2>
vagrantc, you don't suddenly rewrite/reorg things in the kernel, you prove incremental steps. =)
10:34abadger1999 has joined #ltsp
10:35staffencasa has joined #ltsp
10:35
<vagrantc>
warren_2: so definitely, the 3 things in ltsp-trunk/server/scripts are all pretty debian/ubuntu specific
10:35
<ogra_cmpc>
overlapping cards ? ....
10:35
do we do the next ltsp sprint in vegas ?
10:35
<vagrantc>
heh
10:35
<cliebow_>
no
10:35tux_440volt has quit IRC
10:35* vagrantc has no desire to ever see vegas
10:35elisboa has joined #ltsp
10:36
<cliebow_>
nor do i
10:36
<warren_2>
ogra_cmpc, my company would never pay for that
10:36
<vagrantc>
are folks going to be coming to oscon this year?
10:36
<warren_2>
ogra_cmpc, and you can't call that "by the sea"
10:36
vagrantc, where?
10:36
<cliebow_>
by=the sand
10:36steph_ has joined #ltsp
10:36
<ogra_cmpc>
vagrantc, i wasnt asked to speak this year :(
10:36
<vagrantc>
warren_2: i think it's usually in portland, or
10:36
<warren_2>
hm
10:36
I have no budget =(
10:36
maybe Eric's foreign donors can get me there
10:37* ogra_cmpc has the prob that half of the company staff lives in portland
10:37
<warren_2>
ogra_cmpc, you rather not meet them? =)
10:37
<ogra_cmpc>
so we have a resident overhead
10:37
<vagrantc>
http://en.oreilly.com/oscon2008/public/content/home
10:38
<warren_2>
ogra_cmpc, that means you could crash on someone's couch, no hotel
10:38
<vagrantc>
july 21st-25th
10:38
<ogra_cmpc>
warren_2, trams are free in portland ... planes are not
10:38
<warren_2>
hm
10:38
If that's an official LTSP hackfest I might be able to go
10:38
otherwise no
10:39
<vagrantc>
warren_2: well, i'm kind of thinking to try and propose an LTSP hackfest around oscon
10:39
to possibly get some folks who already might be there, or have other excuses to be there
10:39
<ogra_cmpc>
warren_2, well, hackfests are something ltsp-upstream has to decide on
10:39
warren_2, just put that other hat on and approve one ;)
10:40* warren_2 doesn't even like hats.
10:40
<ogra_cmpc>
heh
10:40
<vagrantc>
most ironic
10:40
<warren_2>
I did name Fedora.
10:40
and all problems after that are my fault
10:40
<ogra_cmpc>
indeed
10:41
but we all know that, nothing to whine about ....
10:41* ogra_cmpc hides
10:41
<vagrantc>
so, if we proposed an LTSP hackfest ... either the weekend before fri 18th-20th, or the weekend after fri 25th-28th
10:42* vagrantc counts mondays as weekends
10:42
<vagrantc>
albeit, ineffectually
10:42
<warren_2>
I'll have to check if I can crash at Eric's house
10:42
<ogra_cmpc>
vagrantc, well, if i come and have a good reason, canoonical will ask me to participate in ubuntulive
10:42
<warren_2>
company would prefer I stay in a cardboard box
10:42
<vagrantc>
i can likely get space at freegeek saturday evenings till monday evenings.
10:42
<ogra_cmpc>
which means i'll be busy the week before oscon
10:43
(but in portland)
10:43
<vagrantc>
ogra_cmpc: ubuntu live is scheduled 21st-22nd ... ?
10:43
<ogra_cmpc>
not sure
10:43* vagrantc is looking at the web page
10:43
<ogra_cmpc>
since i wasnt supposed to go i didnt check
10:43
<warren_2>
ogra_cmpc, when is UDS?
10:44
<ogra_cmpc>
vagrantc, right, it says 21-22
10:44
<vagrantc>
i can probably get a space at freegeek and arrange for couches to crash on either weekend or some other weekend with advance notice
10:44
<ogra_cmpc>
warren_2, 19th - 23rd May 2008
10:45
<vagrantc>
ogra_cmpc: where at?
10:45
<ogra_cmpc>
http://towers.corinthia.cz/hotel/en/
10:45
<warren_2>
ogra_cmpc, oh damn, can't do that, I'm in Asia
10:45
<ogra_cmpc>
prague :D
10:45
<warren_2>
ogra_cmpc, unless I go from Asia to CZ directly...
10:45
which might make sense
10:46* ogra_cmpc loves prague ... the most beautiful european city i know
10:46
<warren_2>
prostitution capital of europe
10:46
<cliebow_>
8~)
10:46
hmmm
10:46
<warren_2>
according to a documentary
10:46
<ogra_cmpc>
sadly
10:46
it wasnt always
10:46* ogra_cmpc was in prague the first time 1986
10:47
<ogra_cmpc>
it was quite different the first several times i was there ... and was prtty hefty to see it changing
10:47
<warren_2>
according to the documentary it used to be Amsterdam, now they are trying to clean up their image, so the "capital" is moving east
10:48
<ogra_cmpc>
yeah
10:48
i once drove there by car
10:49
right after the border the sides of the sreets are plastered with little wooden shags that have a glass wall to the streetside
10:49
and crowds of half naked girls
10:49
<warren_2>
ogra_cmpc, in the documentary, the government began buying out entire buildings in order to close down those displays.
10:49
ogra_cmpc, because the government can't ban it
10:51
<ogra_cmpc>
well prague has a lot of empty housing spaxce
10:51* warren_2 thinks it is really sad that this is "economic development" for eastern europe.
10:51
<ogra_cmpc>
yeah
10:51
we had that pretty detailed here in germany the last 20 years
10:52
<warren_2>
detailed?
10:53
<ogra_cmpc>
people from the west opening companies in the east ... cashing in lots of money from that state and turning bankrupt (indeed not for real) within two years
10:53
<warren_2>
wow, that's even worse.
10:53
<ogra_cmpc>
eastern germany has the most recent and coolest infrastructure ...
10:53
<warren_2>
ogra_cmpc, after paying the owner all profits first i'm guessing
10:53
<ogra_cmpc>
new streets everywhere, fiber toyour house etc
10:53
but no jobs
10:53
<warren_2>
ogra_cmpc, is eastern Germany considered part of eastern europe?
10:53
<ogra_cmpc>
it was
10:54
the border was goug through the middle of germany
10:54
*going
10:54
<warren_2>
ogra_cmpc, the phone company here has fiber everywhere except to the capital city in Massachusetts. They are holding the capital hostage, refusing to deploy until they are given special treatment to smash their competitors.
10:54
<ogra_cmpc>
it was about 50km of where i sit atm
10:54
<warren_2>
ogra_cmpc, I know
10:54
ogra_cmpc, I'm just asking about how people consider it today.
10:55
<ogra_cmpc>
well, there is strill a west east difference
10:55
and i guess that will take another 20 years to overcome completely
10:56
<warren_2>
ogra_cmpc, another documentary showed a whole ton of solar power in germany
10:56
<ogra_cmpc>
people in the east get about 20% lower salary for the same work
10:56
<warren_2>
ogra_cmpc, do you see that?
10:56
<ogra_cmpc>
http://people.ubuntu.com/~ogra/CIMG0923-1.JPG
10:56
every time i look at my roof :)
10:57
the open door is my office btw
10:57* vagrantc imagines ogra_cmpc staring at the roof for hours
10:57
<ogra_cmpc>
lol
11:00* Gadi just solved a longtime ubuntu mystery :) so rewarding...
11:00
<ogra_cmpc>
what was it ?
11:00
<Gadi>
I couldnt understand why I couldnt make upstart do the kbdrequest stuff
11:01
turns out the console keymap didnt have alt-Up defined
11:01
<ogra_cmpc>
heh
11:01
<Gadi>
whats worse is that the one in dapper did
11:01
but the one in gutsy did not
11:01
<vagrantc>
warren_2: pushed the ltsp-trunk/server/scripts/* -> server/scripts/debian/
11:01
<Gadi>
in gutsy, there is /etc/console and /etc/console-setup
11:01
<ogra_cmpc>
Gadi, the systems have changed
11:01
right
11:01
<Gadi>
and they both get called by different init scripts
11:02
the keymap that wins in gutsy is the latter
11:02
which does not define alt-uparrow
11:02
<ogra_cmpc>
err
11:02
wait
11:02
where is console-tools
11:03
<Gadi>
well, my system may also not be the cleanest
11:03
<ogra_cmpc>
thats the actual configuation tool (i'm not sure it was in gutsy though)
11:03
<Gadi>
but, I have an /etc/init.d/keymap.sh and /etc/init.d/console-setup
11:03
<ogra_cmpc>
/etc/default/console-setup is what you want to set it up manually
11:04
<Gadi>
right, but the default keymap there does not define alt-up
11:04
so, I would need to specify the right one, which is the old one
11:04
dunno
11:04
so confusing
11:04
does this change again in hardy ?
11:04
<ogra_cmpc>
sudo dpkg-reconfigure console-setup
11:05
<Gadi>
yeah, ive done that a million times, I think
11:05
doesnt help unless you know which keymap is the one you need
11:05
thats the part that takes figuring out
11:05
why there are a million different keymap systems is beyond me :P
11:06
<ogra_cmpc>
there are only two
11:06
<Gadi>
oh, really?
11:06
X has one
11:06
<ogra_cmpc>
one of the just changes all the time
11:06
<Gadi>
console* have there own
11:06
<ogra_cmpc>
*them
11:06
right
11:06
<Gadi>
rdesktop has *its* own!
11:06
<ogra_cmpc>
oh
11:06
<Gadi>
its ridiculous
11:06* ogra_cmpc didnt know
11:07
<Gadi>
they are the same friggin keys!
11:07
:)
11:07
<vagrantc>
well, that probably has to do with windows doing some crazy-fool thing.
11:07
<ogra_cmpc>
they will all be gone soon anyway
11:07
<Gadi>
but, I understand - the keyboard being a relatively new invention
11:07
:P
11:07
<vagrantc>
at least, i can maintain my high horse.
11:07
<ogra_cmpc>
hal is about to take over for at least both of the linux worlds
11:07
<rjune_>
I'm afraid I cannot let you do that ogra
11:08
<Gadi>
great
11:08
<ogra_cmpc>
rjune_, not my decision
11:08
<rjune_>
!g
11:08
<ltspbot>
rjune_: "g" is Gadi!!!!!!!!!!!!!!!!!!!!!!!!
11:08
<Gadi>
then there will be yet another system
11:08
<rjune_>
ogra_cmpc: it's a 2001 reference
11:08
the thinking is there will be one system.
11:08
everything talks to hal
11:08
<ogra_cmpc>
rjune_, crap, i didnt get that ...
11:08* ogra_cmpc is definately to tired
11:09
<rjune_>
when hal says to dave bowman, I'm afraid I cannot let you do that dave or some such
11:09
<ogra_cmpc>
yeah
11:09
<Gadi>
btw, ogra - default CONSOLE_KEYMAP is not "en"
11:09
<rjune_>
it's C
11:09
<Gadi>
there is no "en"
11:09
<rjune_>
LANG=C
11:09
<ogra_cmpc>
i remember the scene
11:09
<Gadi>
it is "us"
11:09
(in lts-parameters.txt)
11:09
:P
11:09
<rjune_>
I thought your keymay was he
11:09* Gadi hates keymaps!!!!!!
11:09
<rjune_>
keymap
11:10
<Gadi>
il
11:10
;)
11:10
<rjune_>
man, somebody riled gadi up.
11:10
<ogra_cmpc>
uk english keymaps are not en ?
11:10
<Gadi>
nope
11:10
<ogra_cmpc>
oh
11:10
<rjune_>
bg
11:10
gb
11:10
or uk?
11:10
<Gadi>
but, its ok
11:10
we dont actually set a default
11:10
<ogra_cmpc>
hmm, where did i get that from then
11:10
<rjune_>
probably the sane part of your brain.
11:10
<ogra_cmpc>
i'm sure i parsed the available keymaps for the list
11:11
<warren_2>
vagrantc, ok, let's think about further moves if they are appropriate
11:11
<ogra_cmpc>
with the same script i used for the xservers
11:11* Gadi gets coffee and puts down revolver
11:11
<vagrantc>
warren_2: i'm comparing your ltsp.spec file against what's in the directories...
11:12
<rjune_>
You live in NYC, you don't own a revolver
11:13
<vagrantc>
warren_2: by the way .. i both love and am terrified by your ltsp-dhcpd.init script :)
11:13
<ogra_cmpc>
heh
11:14* ogra_cmpc rather tends to the latter
11:14
<warren_2>
vagrantc, then don't use it
11:14
=)
11:14
<vagrantc>
warren_2: but then where's the love?
11:14
warren_2: it's very similar to a proposal of how to export the NFS directories ... although that's a little less evil.
11:14
<Q-FUNK>
warren_2: well, Jordan is not entirely happy about the libpciaccess changes from airlied or the wxga patch, and he's worried about recent panel bugs that pop out of nowhere. so, I'm no so sure about the no regresion part.
11:15
<warren_2>
vagrantc, what is?
11:15
Q-FUNK, the recent panel bugs are not a regression from 2.7.7.6 he says?
11:15
<vagrantc>
warren_2: the ltsp-dhcpd stuff ... i.e. starting a custom dhcpd with our own configuration ...
11:15
<ogra_cmpc>
own ports ?
11:16
<Q-FUNK>
warren_2: actually, we don't even know if it's a regression or a bug that simply wasn't noticed before because almost nobody uses the TFT output
11:16
<warren_2>
vagrantc, oh that.
11:16
<vagrantc>
warren_2: there was a proposal to call the appropriate exportfs stuff to ensure that ltsp dirs were added to the NFS server without having to edit /etc/exports
11:16
<warren_2>
vagrantc, that's part of our more radical idea of doing everything through a bridge by default.
11:16
<ogra_cmpc>
nfs is sooo last century ...
11:16
<warren_2>
vagrantc, wow, that's possible?
11:16
<vagrantc>
ogra_cmpc: and forks and spoons are like, ancient.
11:17
warren_2: i haven't actually tried it, but i'd like to.
11:17
where the heck did those patches go ...
11:17
<ogra_cmpc>
werent they from stratus ?
11:17
<vagrantc>
yeah
11:17
google stole stratus from us
11:17* ogra_cmpc remembers the patch but not the url
11:17
<warren_2>
vagrantc, the ltsp-dhcpd is only a separate launcher for the regular dhcpd. I plan on locking it to ltspbr0 by default.
11:18
<cyberorg>
it is easily possible to export without adding it to /etc/exports, maybe it can be triggered from xinetd also?
11:18
<vagrantc>
at least, i think that's what happened
11:18
<ogra_cmpc>
vagrantc, thats lame ... google forces every employee to work about 20% on his own projects ...
11:18
he should have choosen us for that
11:18
<vagrantc>
warren_2: ubuntu does this crazy thing where if /etc/ltsp/dhcpd.conf is present, it uses that *instead* of /etc/dhcp3/dhcpd.conf
11:19
<warren_2>
vagrantc, http://wtogami.livejournal.com/ see the part about "Network Setup", very different from any existing LTSP default, but people seem to like the idea.
11:19
ogra_cmpc, oh yeah? Red Hat forces us to work 100% on our own projects, on top of our RH work. =)
11:19
<ogra_cmpc>
hah
11:20
<vagrantc>
warren_2: i *love* it!
11:20
<ogra_cmpc>
google actually sets you free on paid time ... seems they found the creative outcome is good :)
11:22
<vagrantc>
warren_2: the ltspbr0 thingy, that is.
11:22
<ogra_cmpc>
"It would be nice if everyone came to agreement on a new standard location, but nobody wants to fight that battle right now."
11:22
there is a battle ahead ?
11:22* ogra_cmpc doesnt disagree on /srv
11:22* vagrantc has been very consistantly going to aikido ... minus march
11:23
<warren_2>
ogra_cmpc, i hate /srv
11:23
<ogra_cmpc>
i would propose to compare that to userspace models (tun/tap vs bridges) but overall thats a sexy idea :)
11:24
<warren_2>
ogra_cmpc, oh
11:24
ogra_cmpc, well
11:24
ogra_cmpc, bridges work better for the actual IP in my testing
11:24
ogra_cmpc, because any interface (eth, wlan, tun/tap) can be trivially added to it
11:24
ogra_cmpc, the other way is messy
11:25
<vagrantc>
i have had troubles with bridges and firewalling not picking up changes
11:25
i.e. adding an interface to the bridge and it being rejected ...
11:25
<warren_2>
vagrantc, oh, there might be iptables syntax needed
11:25* warren_2 looks
11:26
<ogra_cmpc>
i have nevr had software probs with my bridges ...
11:26
<vagrantc>
yeah. and i use a wonky firewalling software
11:26
<ogra_cmpc>
they can be pretty CPU intense though
11:26
<warren_2>
it is true that firewalls will be a challenge for this, but it is always a challenge
11:26
ogra_cmpc, eh?
11:26
ogra_cmpc, our kernel guys said not much
11:26
anyway I'm still only testing this, if this turns out to be a bad idea it wont be default in fedora
11:26
<ogra_cmpc>
warren_2, the last bridge i built was around 1999
11:26
<warren_2>
people seem to like it though
11:27
this isn't a bridge between two real eth's
11:27
<ogra_cmpc>
not very recent :)
11:27
<warren_2>
This is a bridge to nowhere. =)
11:27
<ogra_cmpc>
well, its a loopback through your kernel
11:27
<warren_2>
which should be zero copy these days
11:28
<johnny>
hi
11:28
<vagrantc>
the thing i love about using an ltsp-specific interface that we can dynamically add interfaces to ... it's really easy to set up out-of-the-box without any additional configuration, and not fairly easy to switch to some more conventional dhcpd
11:28
<ogra_cmpc>
i didnt say bridges are bad :) ... but i want to look at the different angles before agreeing to one system ... the design idea is awesome
11:29
<vagrantc>
s,not fairly, fairly,
11:30bobby_C has quit IRC
11:30
<vagrantc>
warren_2: what's server/scripts/k12linux/hosts-update ?
11:31
<ogra_cmpc>
sounds like a 4.2 leftover
11:31
<warren_2>
vagrantc, something I copied from eharrison's ltsp-initialize
11:31
vagrantc, I didn't test ltsp-server-initialize yet
11:31
<johnny>
what does one have to do get committ access if necessary?
11:32
<warren_2>
johnny, what distro do you work on?
11:32abadger1999 has quit IRC
11:33
<johnny>
gentoo
11:33
<warren_2>
who is in charge of those acl's?
11:33
<johnny>
i deploy ltsp on ubuntu, but.. i use gnetoo at home, figured it'd be good to make it work there :)
11:33
<vagrantc>
warren_2: hmmm... looking at your ltsp-build-client.conf ... seems like you're using the commandline processing variables directly ... i guess that works ... seems a bit long variable names ...
11:34
johnny: i'd talk to dberkholz about getting access, then.
11:34
<warren_2>
vagrantc, I couldn't figure out another way, how else are you supposed to do it?
11:34
<johnny>
can he add me?
11:35
i've been talking to dberkholz
11:35
<vagrantc>
warren_2: well, in the "configure" phase, we set variables based on the result of the $option_FOO_value parts in each plugin.
11:35
<johnny>
i just didn't know who can do what where
11:35
<vagrantc>
johnny: if dberkholz can't add you now, we should give dberkholz access
11:35steph_ has quit IRC
11:36
<warren_2>
vagrantc, oh, so the plugin is supposed to read its own syntax from config files if you provide one?
11:36
<johnny>
my ltsp-build-client style is more like warren's
11:36
is that a problem?
11:36
<warren_2>
style?
11:36
johnny, in what way?
11:36
<johnny>
making kickstart do everything
11:36
<warren_2>
I make kickstart do much as possible
11:36
<johnny>
yes
11:36
<warren_2>
johnny, gentoo has kickstart, or something similar?
11:37
<johnny>
quickstart
11:37
<warren_2>
heh
11:37
<ogra_cmpc>
johnny, you can do whatever you want in your plugin dir
11:37
<vagrantc>
warren_2: let me find a good example in the Debian dir ...
11:37
<ogra_cmpc>
you could weve just wget a prebuilt image :P
11:37
*even
11:37
<warren_2>
vagrantc, I don't mind the longer names in my .conf, it presumably needs less code to support it
11:37
<johnny>
so, what is early packages vs late packages usually used for?
11:38
<ogra_cmpc>
late packages i the kernel
11:38
<warren_2>
johnny, I don't use it personally
11:38
<ogra_cmpc>
in ubuntu/debian
11:38
<warren_2>
I now understand why you would do kernel late, because then you have the userspace tools to do the initrd.
11:38
<johnny>
ok, then i prolly won't need it either
11:38
since quickstart will handle it
11:38
<vagrantc>
warren_2: you might be right ...
11:38steph_ has joined #ltsp
11:38
<johnny>
sweet..
11:39dfullo has joined #ltsp
11:39
<johnny>
i'm excited
11:39
<warren_2>
johnny, although I had to drop config files for mkinitrd to build a LTSP capable initrd between creation of the chroot and kickstart.
11:39
johnny, see my chroot-creator script
11:39
<vagrantc>
warren_2: right about the variable name stuff, that is ...
11:39
<johnny>
where is that located?
11:39
<warren_2>
johnny, somewhere in server
11:40
<johnny>
ok
11:40
good nuff
11:40
<warren_2>
it is subclassed from our livecd-tools so wouldn't be much of help for you.
11:40
simply our latest way of doing a chroot from packages
11:40steph_ has quit IRC
11:40
<ogra_cmpc>
heh
11:41
<warren_2>
we really have to rename livecd-tools because most people don't use it for that anymore
11:41
<ogra_cmpc>
seems every distro abuses their live builder scripts everywhere :)
11:41
<johnny>
well, it'll help me understand a bit.. so word..
11:41
<ogra_cmpc>
johnny, just look at gentoos scripts for livecds ;)
11:41
<warren_2>
ogra_cmpc, yes, although it was recently reorganized to a pretty good abstracted class-based architecture
11:41
<johnny>
ogra, quickstart is a bit simpler
11:41
<warren_2>
ogra_cmpc, very cleanly build different types of chroot builders and rely on base classes
11:41
ogra_cmpc, very little code
11:42
ogra_cmpc, what is Ubuntu Live event about?
11:42
<ogra_cmpc>
ours as well, its about 300 lines shellscript and covers all *buntu variants
11:42Q-FUNK has quit IRC
11:42
<ogra_cmpc>
its for interested business ppl mostly
11:42
as well as admins and devs of such kind
11:43
<johnny>
it's where the distro comes alive and then ogra has to fight it and put it back to sleep
11:43
<ogra_cmpc>
heh
11:44
<johnny>
so what is that business about releases you guys were talking about?
11:44steph_ has joined #ltsp
11:44* ogra_cmpc releases around april 20th
11:44* vagrantc releases ... someday
11:44
<johnny>
i know you do ;)
11:44
<ogra_cmpc>
vagrant in some months
11:44
and warren very soon
11:45
<johnny>
i thought you were talking about ltsp itself
11:45
some hours ago
11:45
<ogra_cmpc>
these are the ltsp reelases
11:45
<vagrantc>
last debian release date i heard was september 2008
11:45
<ogra_cmpc>
oh
11:45
<warren_2>
we need actual tagged versions and tarballs of ltsp, ldm and ltspfs
11:45
<ogra_cmpc>
thats later than the oct i heard
11:45
<warren_2>
I want to do some reorg of the source trees first
11:45
remove spec files and redo mkdst
11:45
<vagrantc>
ogra_cmpc: it wouldn't suprise me if it were december
11:46
<ogra_cmpc>
or jan even :(
11:46krishna_ has joined #ltsp
11:46
<dfullo>
Hello, I am having an issue with using ubuntu 7.10, ltsp5 and ldap authentication
11:46
<vagrantc>
just as long as it's within 2 years of the last release and doesn't have a freeze that lasts forever, i'll be happy.
11:47
<dfullo>
whenn I reboot the server it hangs
11:47
<dberkholz>
johnny: do you have a diff of the changes you've made?
11:47
<cliebow_>
dfullo:bind policy soft?
11:48
<dfullo>
yes
11:48steph_ has quit IRC
11:48
<johnny>
dberkholz, almost have something working, give me another two days i think.. mainly due to work
11:49
<ogra_cmpc>
dfullo, tried #ubuntu-server ?
11:49
<vagrantc>
cyberorg: with your "*-trunk some observations and questions" post to ltsp-developer, it would have been useful to specify which observations applied to ltsp, ldm and ltspfs ...
11:49
<dfullo>
will do thanks
11:49
<cyberorg>
vagrantc, those were just some random things that occupied me last couple of days
11:50
<dfullo>
If the server is running and I activte ldap authentication it works
11:50
<vagrantc>
cyberorg: it's a little hard to read because it's just a list of issues and i have to actually think about which applies to which piece of code
11:52mccann has quit IRC
11:52
<vagrantc>
cyberorg: but thanks for the post :)
11:52
<cyberorg>
vagrantc, first point is about xinetd conf, second one is about the script that fetches ltspinfo
11:53
tmpwatch is one cron script used to clean up swap
11:53
theme file is gtkrc
11:54
about local devices, i wanted to know how to get the hard disk which is on the client mounted too
11:54mccann has joined #ltsp
11:54
<ogra_cmpc>
you need udev rules for that
11:55
(might be that ist suffices to pull out the (removable) from teh existing ones)
11:55
<cyberorg>
ogra_cmpc, i see those in ltsp_fstab
11:55
but not mounted
11:55
<Gadi>
are they ntfs?
11:56
<cyberorg>
Gadi, nope
11:56
but i do have ntfs-3g installed
11:56
<Gadi>
are they ext3?
11:56
<ogra_cmpc>
cyberorg, you rarely see an ltspfs device mounted
11:57
they are not mounted 95% of the time
11:57
<cyberorg>
Gadi, yes, ext3 and reiserfs may be
11:58
ogra_cmpc, yup, but they doesnt show up in /tmp/.user-ltspfs
11:58
<ogra_cmpc>
is your delayed_mounter working ?
11:58
<Gadi>
on the client, make sure that those filesystems are in /proc/filesystems
11:58
<ogra_cmpc>
thatss actually the script that cares for everything thats plugged in before anynoe logs in
11:59
<dfullo>
got link?
11:59
<ogra_cmpc>
(apart from CDs for which you need to monitor a kernel ioctl)
11:59
which is what cdpinger does
12:00
dfullo, to what ?
12:01
<cyberorg>
ogra_cmpc, there is no delayed_mounter in ltspfs-trunk
12:01
<ogra_cmpc>
its an ldm rc.d script
12:01
but i thought we moved it to the ltspfs source already
12:03
<cyberorg>
ogra_cmpc, would dealyed_mounter mount the local hard disks?
12:04
<ogra_cmpc>
it would mount everything it finds in ltspfs_fstab
12:06
<vagrantc>
cyberorg: in current ltspfs-trunk it's scripts/ldm/S*
12:07
cyberorg: which should be installed to /usr/share/ldm/rc.d/
12:07
<ltsppbot>
"ogra" pasted "delayed mounter" (16 lines) at http://pastebot.ltsp.org/462
12:07
<vagrantc>
cyberorg: which should get called by /usr/share/ldm/ldm-script
12:07
cyberorg: at login time
12:08
ah!
12:08
that's a patched version... i didn't apply the patch upstream because nobody else commented on it... i just applied it to debian and i'm guessing ubuntu pulled it in
12:09
<ogra_cmpc>
yeah
12:09
i thought you applied it
12:09
push it, its a needed bugfix
12:09
<cyberorg>
ah, found it S10-delayed-mounter
12:10
<warren_2>
what does it do?
12:10
<ogra_cmpc>
walk the entries in ltspfs_fstab during login
12:10
and call ltspfsmounter for them
12:10
<vagrantc>
well, in the patched version, i have both a while loop and a for loop ... the original just has a while loop
12:11
<ogra_cmpc>
scotts main concern was floppies when we wrote that
12:11
<vagrantc>
and for reasons i cannot fathom, it didn't work on debian without the while and for loop ... calling the ssh directly within the while loop made only the first entry get added
12:11
<ogra_cmpc>
i dont think we even took local disks into account back then
12:11
<vagrantc>
i've since disabled local disks by default due to a variety of reasons
12:11plamengr has joined #ltsp
12:12
<ogra_cmpc>
that code is originally from the detroit hackfest :)
12:13
<vagrantc>
main reason to disable local disks by default is when people suspend-to-disk and then ltspfs mounts the filesystem ... you can easily get some ugly filesystem corruption.
12:13
<ogra_cmpc>
well, the same goes for swap
12:13
<vagrantc>
yes, but that's disabled by default as well
12:13
<ogra_cmpc>
right
12:13
<cyberorg>
ok, so hopefully i'll have local hard disks mounting tonight
12:14* ogra_cmpc was wondering about suspend to swapfiles once o_O
12:14
<vagrantc>
well, the in-kernel suspend would likely fail badly.
12:14
<ogra_cmpc>
but that was after lots of beer :)
12:14
<vagrantc>
oh, you mean across the network
12:15
<ogra_cmpc>
yeah
12:15
<warren_2>
So the next LTSP hackfest is proposed for the weeked after OSCON?
12:15
<ogra_cmpc>
swapfiles on the server
12:15
<vagrantc>
warren_2: before or after ...
12:15* ogra_cmpc cant promise he can make it
12:15
<ogra_cmpc>
but i'll try
12:15exodos has quit IRC
12:15
<vagrantc>
should i make a post on-list ?
12:15
<ogra_cmpc>
its a good reason
12:15
yeah
12:15
<warren_2>
vagrantc, yeah
12:16
<vagrantc>
ogra_cmpc: better for you before or after oscon ?
12:16
<ogra_cmpc>
well, in the gap
12:16
<vagrantc>
i guess i can post both dates as a proposal and get feedback
12:16krishna_ has quit IRC
12:16
<ogra_cmpc>
i'm not sure i have the time to attend oscon
12:16krishna_ has joined #ltsp
12:18* warren_2 slightly prefers before OSCON, because after might be burnt out.
12:19mccann has quit IRC
12:20
<ogra_cmpc>
http://people.ubuntu.com/~ogra/cbm.jpg
12:20
whoops ... that was not intended ...
12:20mccann has joined #ltsp
12:21
<warren_2>
ogra_cmpc, is that your LTSP server?
12:21
=)
12:21
<ogra_cmpc>
hehe
12:21
nah, its classmates grampa
12:22
http://www.indidea.org/gael/blog/wp-content/uploads/2007/03/classmatepc.png
12:23
oh, mine have the coloring the other way round ...
12:23
they look more like a cb,
12:23
cbm
12:25milesd has joined #ltsp
12:30tux_440volt has joined #ltsp
12:31krishna_ has quit IRC
12:32steph_ has joined #ltsp
12:33krishna_ has joined #ltsp
12:37
<vagrantc>
anybody know who "Gürkan Sengün" is? they asked to be added to pkg-ltsp project on alioth.debian.org ... ?
12:38
<ltsppbot>
Someone pasted "/proc/filesystems" (27 lines) at http://pastebot.ltsp.org/463
12:38
<cyberorg>
Gadi, i dont see ext3/reiserfs there
12:39
but when i plug in ext3 usb, it works
12:39
<Gadi>
is that the server or the client?
12:39steph_ has quit IRC
12:39
<cyberorg>
/var/run/ltspfs_fstab dont exist till i plug in usb
12:40
on client
12:40
<Gadi>
weird that it works
12:40
<warren_2>
brb
12:40warren_2 has quit IRC
12:40mccann has quit IRC
12:41
<ltsppbot>
"cyberorg" pasted "filesystem after usb plug" (28 lines) at http://pastebot.ltsp.org/464
12:41tux_440volt has quit IRC
12:42mccann has joined #ltsp
12:43
<ogra_cmpc>
vagrantc, sounds turkish ... never heard from him/her
12:43
<ltsppbot>
"cyberorg" pasted "sh -x /usr/share/ldm/rc.d/S10-delayed-mounter start" (35 lines) at http://pastebot.ltsp.org/465
12:44
<vagrantc>
cyberorg: looks like your socket isn't defined
12:44krishna_ has quit IRC
12:44krishna_ has joined #ltsp
12:44subsume|work has joined #ltsp
12:44bjornh` is now known as bjorn
12:45
<cyberorg>
vagrantc, yeah, i ran that as root on client, it wont work
12:45alekibango has quit IRC
12:45
<cyberorg>
however, no local hdd, in ltsp_fstab
12:45
<vagrantc>
cyberorg: what are your udev rules?
12:46
<ogra_cmpc>
yeah, thats fault of udev
12:46
<vagrantc>
cyberorg: has your kernel loaded the modules for your hard drives?
12:46
<cyberorg>
checking "<ogra_cmpc> (might be that ist suffices to pull out the (removable) from teh existing ones)"
12:47
vagrantc, yes
12:47
<vagrantc>
cyberorg: so /dev/[hs]d* exists?
12:47plamengr has quit IRC
12:47dfullo has quit IRC
12:48
<cyberorg>
sda sda1 sda2 sda5 sda6 sda7 sdb sdb1 sdb2 sdb3 sdb5 sdb6
12:49
<ogra_cmpc>
i guess he has some rules overruling the ltsp rules here
12:49
else all of these would bebin ltspfs_fstab
12:49
<vagrantc>
sure
12:50alekibango has joined #ltsp
12:50
<vagrantc>
cyberorg: well, you commented out the whole line, or you removed the ATTRS{removable]=="1" part from the line(s) in your ltspfs udev rules?
12:50
<Gadi>
not all - only those with FSTYPEs
12:51
<ltsppbot>
"cyberorg" pasted "udev" (49 lines) at http://pastebot.ltsp.org/466
12:51
<vagrantc>
well, yes.
12:51
<Gadi>
for i in $(ls /dev/sd*); do vol_id $i; done
12:51
<cyberorg>
what am i missing?
12:52
<vagrantc>
cyberorg: which of those sd* is your local device, and which are the usb sticks?
12:52
<Gadi>
cyberorg: can you: mount /dev/sda1 /mnt
12:52
on the client
12:52
in other words, try mounting them
12:53
<cyberorg>
sda = local sdb = usb, usb shows up
12:53
<Gadi>
if you get an error about specifying filesystem type, then perhaps the fs modules need to be loaded on the client
12:53
on boot
12:53
(ie add them to $chroot/etc/modules or equivalent)
12:54
<cyberorg>
mount /dev/sda1 /mnt works
12:54
Gadi, if module was the problem, usb stick wouldnt mount too, it is ext3
12:54
<Gadi>
right
12:54
<subsume|work>
Anyone tell me a common cause of ARP timeouts?
12:55
My netboot was working just fine until I changed some settings in DHCP
12:55
Now my clients get ARP timeout.
12:55
<cyberorg>
Gadi, vol_id: command not found
12:55
<subsume|work>
Googling produces nothing useful, I am guessing it is something super stupid and simple.
12:55
<ogra_cmpc>
thats rather on the phys layer
12:55
do you have a manageable switch or so ?
12:56
<subsume|work>
ogra_cmpc: not sure what you are talking about.
12:56
<Gadi>
cyberorg: try: /lib/udev/vol_id
12:56
<subsume|work>
<-- not a network guru
12:56
<-- more like a network neophyte.
12:56
<ogra_cmpc>
subsume|work, arp is part for the ethernet protocol ....
12:57
s/for/of/
12:57
<ltsppbot>
"cyberorg" pasted "vol_id" (69 lines) at http://pastebot.ltsp.org/467
12:57
<ogra_cmpc>
the IP protocol (and dhcp) sit on top of that
12:57
<subsume|work>
ogra_cmpc: ok... well I'm not sure. I know its seeing the DHCP server.
12:57
<ogra_cmpc>
having arp timeouts indicates a problem on the physical layer
12:57krishna_ has quit IRC
12:57
<subsume|work>
ogra_cmpc: It was net booting perfectly before I changed the gateway
12:57
<ogra_cmpc>
cables, switches, netwrok cards
12:58krishna_ has joined #ltsp
12:58
<subsume|work>
ogra_cmpc: I don't think that's it because I can get rid of the ARP timeout problem and produce a TFTP not found error by changing the DHCP settings.
12:58
<cyberorg>
vagrantc, i con't find ATTRS{removable]=="1" anywhere
12:59
<vagrantc>
cyberorg: ah, that's right. we ship a different file in debian
13:00
<cyberorg>
vagrantc, where can i find that?
13:00
<vagrantc>
cyberorg: seems a little odd that your /dev/sda reports a filesystem, yet it also has a /dev/sda2
13:00
<cyberorg>
or better still can the one in ltspfs-trunk be fixed with yours?
13:00
<subsume|work>
Is there some way to inherit urls?
13:00
woops.
13:00
wrong channel.
13:00
<vagrantc>
cyberorg: bzr get http://bzr.debian.org/bzr/pkg-ltsp/ltsp-debian-packaging/
13:01
cyberorg: ltspfsd.udev
13:01
<cyberorg>
vagrantc, sda would be MBR?
13:01
<ogra_cmpc>
MBR !+ filesystem
13:01
!=
13:01
damned keyboard
13:01mccann has quit IRC
13:02
<subsume|work>
ogra_cmpc: that to me?
13:02
<ogra_cmpc>
subsume|work, are you fiddling with filesystems ? :)
13:02
<cyberorg>
ogra, i know, but like vagrantc it should have any FS on it?
13:02
<vagrantc>
cyberorg: well, on all my devices, i get nothing when calling /lib/udev/vol_id on /dev/sd? , but i do when i run it on /dev/sd??
13:02
<subsume|work>
ogra_cmpc: sorry. this is how unversed I am regarding my above question to ya. =)
13:03
<Gadi>
it doesnt
13:03
those are errors
13:03
<cyberorg>
vagrantc, i called on /dev/sd*
13:03
<ogra_cmpc>
subsume|work, so if you turn everything back as it was before, does it work ?
13:03
<Gadi>
in other words, sda and sda2 do NOT have filesystems
13:03
vol_id failed
13:03
the rest of the output is from the other parts
13:03
<subsume|work>
ogra_cmpc: I will have to check. Its at my night job. =(
13:03
<ogra_cmpc>
by just putting back your backups you did before changing the configs ...
13:03
:)
13:04
<subsume|work>
ogra_cmpc: backups? what are those? =)
13:04
<ogra_cmpc>
hehe
13:04
<Gadi>
so, it must be delayed_mounter's fault
13:04
<subsume|work>
ogra_cmpc: but based on what I am saying, do you still think its phys or can it be config?
13:04
<vagrantc>
delayed_mounter can't be expected to mount a non-existant filesystem
13:05
<Gadi>
no, it shoul mount: sda1, sda3, and sda4
13:05
<vagrantc>
subsume|work: one way to find out if it's configuration is to switch the configuration back exactly as it was...
13:05
<subsume|work>
ogra_cmpc: because at the phys level I can see that the netbooting client responds to changes in th DHCP server.
13:05
<Gadi>
as they are ext3, vfat, and ext3
13:05
<cyberorg>
vagrantc, delayed_mounter does'nt come into picture till ltsp_fstab is populated, only usb shows up there
13:05
<subsume|work>
vagrantc: I hope you are up tonight. =)
13:05
<ogra_cmpc>
vagrantc, lets just formeat everything we find during client boot ...
13:05
<Gadi>
if you see those 3 parts, then thats right, cyberorg
13:05
<vagrantc>
ogra_cmpc: pfft.
13:06
<subsume|work>
I read this as far as PXE errors
13:06
"Setting the DHCP Class Identifier (option 60) on the DHCP server and installing the proxyDHCP on a separate machine."
13:07
Could someone translate this into human, ignorant dialect?
13:07
<cyberorg>
Gadi, it doesnt mount sda* at all
13:07
<ogra_cmpc>
subsume|work, "run away screaming" ?
13:07
<Gadi>
cyberorg: but you see it create folders on the server, right?
13:08
<subsume|work>
grr
13:08
<cyberorg>
Gadi, no, just the usb ones
13:08
<ogra_cmpc>
subsume|work, why do you want a dhcp proxy ?
13:08
<subsume|work>
heh. just forget about it, it probably doesn't apply.
13:08
<ogra_cmpc>
thats maintenance hell
13:08
<Gadi>
..and you rerolled the image and rebooted since you made changes, yeah?
13:08
<ogra_cmpc>
ah
13:09
<vagrantc>
cyberorg: can you run vol_id on each of your local devices, and see if any of them have filesystems?
13:09
<Gadi>
he did
13:09
and they do
13:09
<cyberorg>
vagrantc, i just pasted vol_id on /dev/sd*
13:09
<vagrantc>
ok.
13:10
the output was rather hard to make heads or tails of.
13:10* ogra_cmpc is sure there is a udev rule firing before ltsp even comes into play
13:10
<Gadi>
cyberorg: you see them in /var/run/ltspfs_fstab?
13:10
<cyberorg>
Gadi, no
13:10
<ogra_cmpc>
that steals away the devices before ltsp would care
13:10
<cyberorg>
ogra, there is one rule i suspect that one
13:10
<Gadi>
oh, really?
13:10
then, yeah its udev
13:10
<vagrantc>
"for i in $(ls /dev/sd*); do echo trying $i ; /lib/udev/vol_id $i; done" would have made it a little easier for me to read :)
13:10
<Gadi>
hmmm...
13:11
<ogra_cmpc>
its even more likely that they (as block devices) come from initramfs
13:11
<ltsppbot>
"cyberorg" pasted "cat /etc/udev/rules.d/60-persistent-storage.rules" (59 lines) at http://pastebot.ltsp.org/468
13:11
<Gadi>
cyberorg: what happens if you restart udev on the client?
13:11
<cyberorg>
ogra, true, they also come from initrd
13:11
<ogra_cmpc>
right
13:12
in case of ubuntu they would be created in the initramf's /dev during boot ... (i dont think we have any block device drivers in initramfs though)
13:13
which then is bind mounted to the real /
13:13
<cyberorg>
we have udev in intrd which creates all the devices
13:13
<ogra_cmpc>
yeah, thats a prob then
13:14
can you build the initrd without any blockdevice drivers ? then a later udev instance that runs in / should cover them
13:15
<cyberorg>
i'll try removing the persistant device rule
13:16mccann has joined #ltsp
13:16krishna_ has quit IRC
13:25steph_ has joined #ltsp
13:27steph_ has quit IRC
13:40
<cyberorg>
removing other udev persistent device rules did not change anything
13:40
<Gadi>
did you remove from initrd?
13:41
<cyberorg>
Gadi, hmm, those rules are there in initrd too :(
13:42
<Gadi>
so dont change the rules
13:42
just blacklist the drivers in initrd
13:42
and reroll it
13:43
just be sure not to blacklist them in the OS
13:43
and by drivers, I mean ide-disk, libata and friends
13:44
<cyberorg>
Gadi, donno how to blacklist them
13:45
Gadi, do i remove drivers/ide/* ?
13:45
<Gadi>
does novell have a tool to regenerate the initrd?
13:45
<cyberorg>
Gadi, yes
13:45
<Gadi>
is there an /etc/.... file to control it?
13:47
<cyberorg>
Gadi, i've removed drivers/ide/* from my config
13:47
we list what drivers/packages we want to install in initrd in a xml file
13:49
http://svn.berlios.de/wsvn/kiwi/kiwi-head/system/boot/netboot/suse-11.0/config.xml see the diskless profile
13:50
<Gadi>
are they ide drives or ata?
13:51
<cyberorg>
Gadi, now all are ata?
13:51
<Gadi>
well, sata, for example
13:51
uses libata and not ide-disk
13:51
<ogra_cmpc>
<drivers type="drivers" profiles="diskless">
13:51
<!-- DRIVERS START DISKLESS -->
13:51
<file name="drivers/ide/*"/>
13:52* ogra_cmpc laughs loudly
13:52
<Gadi>
ogra_cmpc: what are you? the playground bully?
13:53
<ogra_cmpc>
thats like putting a sign with arrow near your car saying "This is a ship !"
13:53
...
13:53
this is the diskless profile ... the first module we need is the disk driver :)
13:53
<Gadi>
lol
13:54
that is ironic
13:54
<ogra_cmpc>
yeah
13:54
ironic sourcecode
13:54
<cyberorg>
ok, drivers gone, but still no disks in ltsp_fstab
13:54
ironic thing is i am trying to get local disk since last few hours :)
13:55
<ogra_cmpc>
Gadi, if you print it out it could probably run a stand-up-golem
13:56
ok, i clarly need sleep ....
13:56
<- starts getting silly
13:56
<Gadi>
cyberorg: try restarting udev from a shell on the client
13:56
and then check ltspfs_fstab again
13:58mccann has quit IRC
13:58
<cyberorg>
Gadi, no ltsp_fstab still
13:59
<vagrantc>
now that's satisfying: 4 files changed, 12 insertions(+), 81 deletions(-)
13:59
effectively identical code
14:02jammcq has quit IRC
14:06
<johnny>
very much so..
14:06mccann has joined #ltsp
14:08
<Gadi>
cyberorg: if there is no ltspfs_fstab even after restarting udev, then that is suspicious to me
14:08
bec that udev rule should definitely execute
14:09
<cyberorg>
Gadi, strange thing is, it turns up as soon as i plug in usb
14:09
i'll try vagrantc's udev
14:10
<ogra_cmpc>
what would restarting the daemon gain you there ?
14:10
udevd doesnt do anything on its own
14:10
it only acts on events
14:11
make sure the daemon runs and do a: udevadm trigger
14:11
that will send the fake events udev needs to create the devices
14:11
<cyberorg>
cool, will try that as well
14:12
rebuilding chroot..
14:12deavid has quit IRC
14:13
<ogra_cmpc>
oh, and udev will only act on devices that dont exist
14:13
so make sure there atre no /dev/sdX files
14:16mikkel has quit IRC
14:16
<cyberorg>
removed, /dev/sd*, restarted udev, ran udevadm trigger, /dev/sda* back, but ltsp_fstab no there
14:17
<ltsppbot>
"cyberorg" pasted "/etc/udev/rules.d/88-ltsp.rules" (4 lines) at http://pastebot.ltsp.org/469
14:17
<cyberorg>
is this enough?
14:17
<ogra_cmpc>
well, find /dev/sd* in the other rules
14:18
for me in ubuntu thats in 65-persistent-storage.rules
14:19
given that ltsp runs with priority 88 they are already there if ltspfs comes into play
14:20
<cyberorg>
grep sd * -R returns nothing
14:20* ogra_cmpc wonders if suse hardcodes blockdevice creation in initramfs
14:20
<ogra_cmpc>
hmm
14:20
no
14:21
then they wouldnt be creted by trigger
14:21
*created
14:21
grep for block od scsi
14:21
*or
14:23
<ltsppbot>
"cyberorg" pasted "grep scsi and block" (19 lines) at http://pastebot.ltsp.org/470
14:24
<ogra_cmpc>
hmm
14:24
doesnt look like there is anything for /dev/sd
14:24
<cyberorg>
what would happen if i set 88-ltsp to 10-ltsp...
14:25
want access to the TC?
14:26
<ogra_cmpc>
no, suse scares me ... i might wipe it accidentially
14:26
<cyberorg>
haha you can't wipe anything in squashfs image
14:33cliebow_ has quit IRC
14:35mccann has quit IRC
14:36mccann has joined #ltsp
14:40Gadi has left #ltsp
14:50
<cyberorg>
vagrantc, ltspfsd.udev doesnt exist in ltsp-debian-packaging
14:52
<vagrantc>
cyberorg: oops: http://bzr.debian.org/bzr/pkg-ltsp/ltspfs/ltspfs-debian-packaging/
14:52
s,ltsp-debian-packaging,ltspfs/ltspfs-debian-packaging,
14:53
cyberorg: sorry about that
14:53
<cyberorg>
bzr: ERROR: Unknown branch format: 'Bazaar pack repository format 1 (needs bzr 0.92)!!
14:53
Bazaar (bzr) 0.91.0 :(
14:54
<vagrantc>
cyberorg: what version of bzr are you using?
14:54
heh
14:54
<cyberorg>
how can it not be compatible with even it's . version!!
14:54
<vagrantc>
unfortunately, bzr requires keeping pretty up to date.
14:54
<cyberorg>
git/svn rocks
14:54
<johnny>
lol
14:55
not really
14:55
<vagrantc>
actually, i've been quite happy with bzr 1.0
14:55
<cyberorg>
is there online repo?
14:55
<johnny>
monotone ftw
14:55
<vagrantc>
haven't felt the need to upgrade
14:56
<cyberorg>
vagrantc, can you pastebot it please?
15:02
<ltsppbot>
"vagrantc" pasted "ltspfs udev rules on debian" (5 lines) at http://pastebot.ltsp.org/471
15:02* vagrantc notes that it still says # /etc/udev/rules.d/88-ltsp.rules
15:03
<vagrantc>
even though it's usually installed at z60_ or something
15:07
<cyberorg>
vagrantc, thanks, will try that
15:07
are you running cdpinger directly from tehre?
15:07
<vagrantc>
cyberorg: not yet
15:08
<cyberorg>
ACTION=="add", SUBSYSTEM=="block", ENV{ID_TYPE}=="cd", RUN+="/usr/sbin/cdpinger %k"
15:08
<vagrantc>
i experimented with it, and it seemed to work.
15:08
<cyberorg>
ok, will test that here as well
15:08
<vagrantc>
i don't know if it will catch all devices or not
15:08
i.e. do dvd-player's report as ID_TYPE=dvd ?
15:09mccann has quit IRC
15:10
<cyberorg>
vagrantc, donno, will find out :)
15:11milesd has quit IRC
15:12
<vagrantc>
hmm... at least on an ubuntu gutsy system, i'm getting a DVD to show up as ID_TYPE=cd
15:13
<cyberorg>
vagrantc, wouldnt it be dependent on kernel?
15:14
<vagrantc>
cyberorg: could you be more specific?
15:15
<cyberorg>
device id types, wouldnt they depend on kernel?
15:17
vagrantc, can we use something like ENV{ID_TYPE}=="cd|dvd" ?
15:17
<vagrantc>
not to my knowledge
15:17mccann has joined #ltsp
15:17* vagrantc wishes there was an OR syntax for udev rules
15:18chupa has quit IRC
15:19
<cyberorg>
i see a lot of DRIVERS=="ide-cs|ide-floppy", GOTO="persistent_storage_end
15:19
<vagrantc>
hmmm... maybe it can
15:19mccann_ has joined #ltsp
15:19
<cyberorg>
KERNEL=="ram*|loop*|fd*|nbd*|gnbd*|dm-*|md*", GOTO="persistent_storage_end"
15:19
<vagrantc>
that would be nice.
15:20
couldn't find anything in the documentation about it
15:20
<cyberorg>
vagrantc, we do not have ID_TYPE used in any of our udev rules
15:21
<vagrantc>
cyberorg: well, run udevinfo -qenv -n/dev/hda
15:21
cyberorg: do you see ID_TYPE ?
15:21chupa has joined #ltsp
15:22milesd has joined #ltsp
15:22
<vagrantc>
on ubuntu, i'm seeing DEVTYPE as well as ID_TYPE
15:22
<cyberorg>
vagrantc, yes
15:22
<vagrantc>
cyberorg: then it will likely work, weather other rules use it or not.
15:22
<cyberorg>
so most likely we can use cd|dvd :)
15:23
<vagrantc>
most likely, yes.
15:23
<cyberorg>
want tarball of all our rules?
15:23
<vagrantc>
although with at least one DVD drive is shows itself as cd
15:23
no desire for such a tarball, no.
15:24twinprism has quit IRC
15:30vagrantc has quit IRC
15:34mccann has quit IRC
15:36daduke_ has quit IRC
15:37mccann has joined #ltsp
15:42Blinny has quit IRC
15:46mccann_ has quit IRC
15:49Egyptian[Home] has quit IRC
15:51Egyptian[Home] has joined #ltsp
16:03mccann has quit IRC
16:05mccann has joined #ltsp
16:14mhterres has quit IRC
16:27mccann has quit IRC
16:27mccann has joined #ltsp
16:39ogra_cmpc_ has joined #ltsp
16:40ogra_cmpc has quit IRC
16:49topslakr has quit IRC
16:52Guaraldo has left #ltsp
16:54topslakr has joined #ltsp
16:55daduke_ has joined #ltsp
17:00sutula has quit IRC
17:03subsume|work has quit IRC
17:15ogra_cmpc_ has quit IRC
17:16bobby_C has joined #ltsp
17:17mccann has quit IRC
17:27nicoAMG has quit IRC
17:53Q-FUNK has joined #ltsp
17:59topslakr__ has quit IRC
18:29abadger1999 has joined #ltsp
18:48Q-FUNK has quit IRC
18:58twinprism has joined #ltsp
19:03ogra_cmpc has joined #ltsp
19:03staffencasa has quit IRC
19:14daduke_ has quit IRC
19:24bobby_C has quit IRC
20:05milesd has quit IRC
20:09chupa has quit IRC
20:11chupa has joined #ltsp
20:36hari has joined #ltsp
20:43elisboa_ has joined #ltsp
20:59
<hari>
I'm trying to install ltsp 4.2 into ubuntu 7.10 gutsy but still failed. the lastest failure is TFTP out of time. Does anyone have experience to this problem. ltspadmin status and running is 'yes' . thx
21:01
client is booting using ethernet pxe. when i'm trying to boot using floppy disk. the message is sounde the same.
21:38spectra has joined #ltsp
21:43
<RyanRyan52>
hari: look at the topic of this channel: LTSP-5.0 is the current version that you should be using. Especially if you are new to LTSP.
22:13
<hari>
do you know where i can download ltsp5. i can not find at the same web (twiki ?) with ltsp 4.2. regards
22:21spectra has quit IRC
22:23
<hari>
i found it.. downloading... thanks
22:27
It seems i have to use ubuntu feisty fawn to install ltsp5-0, is it right?
22:28
what ubuntu version do you suggest for installing ltsp 5?
22:50abadger1999 has quit IRC
22:55abadger1999 has joined #ltsp
23:05
<johnny>
hari, gutsy or dapper
23:05
definitely not feisty
23:17
<hari>
i have difficulties this week to install ltsp especially ltsp 4.2 using ltsp-utils and ltspadmin although all status are running well. I think ubuntu give the easiest way to install packages using synaptic. do you think i should install ltsp with synaptic (ltsp-server-standalone. if i use gutsy, do you it's automatically using ltsp 5? regards)
23:24ogra_cmpc has quit IRC
23:24ogra has quit IRC
23:24ogra_cmpc has joined #ltsp
23:24ogra has joined #ltsp
23:25
<johnny>
yes it does
23:28
<hari>
what should I install for ltsp5 running well? i think only dhcp3-server, ltsp-server-standalone and all dependencies should be installed as well.. am I right?
23:28
<johnny>
yes
23:28
ltsp-server-standalone
23:28
not dhcp3-server
23:29
ltsp uses it's own dhcp config
23:29
<hari>
i see, thanks.. will try...
23:50
<dberkholz>
oh that reminds me
23:51
johnny: my plan with gentoo is to use dnsmasq for everything -- dns, dhcp, tftp -- rather than deal with the huge complex piles of other stuff
23:51
<johnny>
dberkholz, cool
23:51
i'm already familiar with dnsmasq
23:51
i use it with ltsp in ubuntu at the store
23:52
<dberkholz>
i've got a working config around somewhere for everything but the tftp
23:52
if you're using it already too, that's great
23:52
<johnny>
i have a working config with everything :)
23:52
only one problem..
23:52
altho not for 99% of use cases
23:53
i wanted to move dnsmasq to another box, but for some reason it doesn't pass the right info to the dhcp classes
23:54
err
23:54
to the dhcp clients
23:54
altho in that case you obviously need a seperate tftp server ..
23:54
<dberkholz>
it should if you just find the right number
23:54
<johnny>
i tried all the recommended ones
23:54
there's a post on the list with no answer
23:54
from q3 2007
23:55
<dberkholz>
frankly i don't really care about 0.1% of people, they can deal with it on their own
23:55
<johnny>
i'm one of those people :(
23:55
<dberkholz>
they can use isc stuff for all i care
23:55
i want an easy config for most people
23:55
and a possible config for the rest
23:55
<johnny>
so there will be a solution in the long term
23:55
<dberkholz>
the solution will be "use isc dhcp/bind if dnsmasq doesn't work for you"
23:56
there's no reason people can't just switch packages around on the server, every day if they want to.
23:56
i'm just not interested in supporting anything but dnsmasq
23:57subir has quit IRC