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


Channel log from 17 November 2008   (all times are UTC)

00:11Q-FUNK has quit IRC
00:20alkisg has quit IRC
00:39captain_1agnus has joined #ltsp
00:52captain_magnus has quit IRC
01:12Q-FUNK has joined #ltsp
01:16johnny has quit IRC
01:16johnny has joined #ltsp
01:17johnny is now known as Guest1354
01:29Guest1354 is now known as johnny
01:36alkisg has joined #ltsp
01:36Pascal_1 has joined #ltsp
01:46Pascal_1 has left #ltsp
02:14tjikkun_work has joined #ltsp
02:16selffik has joined #ltsp
02:21try2free has joined #ltsp
03:09BrunoXLambert has quit IRC
03:09BrunoXLambert has joined #ltsp
03:10hanthana has joined #ltsp
03:15try2free has quit IRC
03:19mikkel has joined #ltsp
03:58rcy``` has joined #ltsp
03:58rcy`` has quit IRC
04:07selffik has quit IRC
04:09johnny has quit IRC
04:10johnny has joined #ltsp
04:59rcy``` is now known as rcy``
05:12Q-FUNK has quit IRC
05:30Q-FUNK has joined #ltsp
05:32alekibango has quit IRC
05:33selffik has joined #ltsp
05:34Q-FUNK has quit IRC
05:34Q-FUNK1 has joined #ltsp
05:34Q-FUNK1 is now known as Q-FUNK
05:35Egyptian[Home]1 has joined #ltsp
05:39Egyptian[Home] has quit IRC
05:44alekibango has joined #ltsp
06:19BrunoXLambert has quit IRC
06:23Q-FUNK has quit IRC
06:35selffik has quit IRC
06:39exodos has joined #ltsp
06:56pmatulis has joined #ltsp
07:02Basti_dash has joined #ltsp
07:03try2free has joined #ltsp
07:05jammcq has quit IRC
07:07Blinny has joined #ltsp
07:10cliebow has joined #ltsp
07:17try2free has quit IRC
07:19vvinet has quit IRC
07:34bjs_redhat has joined #ltsp
07:34vvinet has joined #ltsp
07:45BrunoXLambert has joined #ltsp
07:48lucascoala has joined #ltsp
07:57Blinny has quit IRC
08:01six2one has joined #ltsp
08:06hanthana_ has joined #ltsp
08:11mikkel has quit IRC
08:14Gadi has joined #ltsp
08:14dmaran has joined #ltsp
08:15
<Gadi>
morning, all
08:16
ogra: put intrepid on my eeepc901 on friday, and yesterday got bluetooth+dialup+cell phone working :)
08:16
its quite solid
08:16
and very nice
08:16
<ogra>
:)
08:17
<Gadi>
couldn't find dialup in the network configs, so I did it all by hand
08:17
<ogra>
oh ?
08:17
<Gadi>
not sure where the desktop folks hid it :)
08:17
<ogra>
its in NM applet
08:17
<Gadi>
they have VPN, DSL, ...
08:17
<ogra>
and 3G
08:17
<Gadi>
but, I was looking for just dialup
08:17
right - but only 3G for certain providers
08:18
and not mine (EVDO)
08:18
<ogra>
right, only the ones we got user submissions for
08:18
<Gadi>
ah
08:18
<ogra>
file a bug and add info :)
08:18
<Gadi>
too bad there was no generic
08:18
I would have liked to just fill it in
08:18
with all the params
08:18
<ogra>
hmm, not sure how that works, i havent used 3G yet
08:19
<Gadi>
with mine, it is just dialup with noauth and number #777
08:19
<ogra>
i just saw the list showing up when i tested my phone bluetooth
08:19
<Gadi>
yeah - bluetooth is 3x slower than USB cable
08:19
so, its only if I dont have the cable on me
08:19
<ogra>
ah
08:20
<Gadi>
but, 128kbps is still useable somewhat
08:20
750kbps is much better, tho
08:20
:)
08:20
<ogra>
heh, yeah
08:20
<Gadi>
intrepid is quite nice
08:21
feels more stable than hardy
08:21* Gadi needs to ask sbalneav how hardy is as a terminal server
08:21
<Gadi>
with all the bugfixes
08:21
<ogra>
hardy is still missing a bunch of SRU backports
08:22
not all intrepid fixes are/will be backported
08:22
<Gadi>
better to do intrepid for a terminal server, you think?
08:22
<ogra>
i would think so
08:22
unless you actually need LTS
08:22
hardy was very conservative wrt ltsp
08:23
<Gadi>
well, I am less concerned about the ltsp part
08:23
more about the memory usage, etc
08:23
I know there were quirky server-side things
08:23
<ogra>
then its definately intrepid
08:23
more client side actually
08:24
clients use compcache by default
08:24
<nubae>
intrepid IS more stable than hardy
08:24
<ogra>
in intrepid you should actually be able to boot 32M clients
08:24
<Gadi>
oh - can I enable that on my eeepc?
08:24
<ogra>
yes
08:24
<Gadi>
is it a kernel arg?
08:24hanthana has quit IRC
08:24
<ogra>
initramfs.conf
08:24
<Gadi>
do you know the magic word off hand?
08:25
<ogra>
COMPCACHE_SIZE=
08:25
values are documented in the conf file
08:25* Gadi was kinda reluctant to give the lil guy swap space on the flash
08:25
<Gadi>
ah, cool
08:25* ogra implemented it :)
08:25
<Gadi>
is there a rule of thumb size?
08:25
given the amount of RAM
08:25
<ogra>
25% is the default
08:26
i takes % M K G as units
08:27alkisg has quit IRC
08:35mccann has joined #ltsp
08:37
<Gadi>
"you can optionally install the compcache package" - but it is not in my apt-cache?
08:37
is it called something else?
08:38
<ogra>
no, it didnt get through REVU in time
08:38
it just adds a debconf frontend
08:38
<Gadi>
ah
08:38
so, I dont need it
08:38
<ogra>
so you get a ble screen with one entry field
08:38
*blue
08:38
and OK button
08:39
just edit initramfs.conf and run update-initramfs
08:39
<Gadi>
running...
08:44
and, it works as advertised :)
08:44
<ogra>
:)
08:48bjs_redhat1 has joined #ltsp
08:49CAN-o-SPAM_ has joined #ltsp
08:49
<CAN-o-SPAM_>
#dlw
08:53Basti_dash has quit IRC
08:53Basti_dash has joined #ltsp
08:54BrunoXLambert has quit IRC
08:54BrunoXLambert has joined #ltsp
09:05bjs_redhat has quit IRC
09:32warren has quit IRC
09:35
<sbalneav>
Good morning all
09:38
<_UsUrPeR_>
morning Mr. Balneav
09:38
!s
09:38
<ltspbot>
_UsUrPeR_: "s" is Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
09:38
<_UsUrPeR_>
:D
09:41
<sbalneav>
_UsUrPeR_: Balneaves, but I thank you for the honorific :)
09:41
<_UsUrPeR_>
I try. I think the !s says it all
09:42
<sbalneav>
Well, last hardy upgrade here went off fairly smoothly
09:42
<_UsUrPeR_>
awsome
09:42
err awesome
09:42
my wife tried to use the automatic upgrade for ubuntu 8.04 on her laptop
09:42
it hosed many a thing :/
09:43
I'm sure a new install will work out better for her, but until then she's trying Fedora w/ KDE
09:43
<sbalneav>
I'm still getting a problem here where once in a while, when people print out from OpenOffice.org, the printout's completely fubared. The formatting's all there, but it's like the character map's gotten randomized, and all the letters on the page are substituted for random ones.
09:43
Power cycling the printer seems to fix it.
09:43
<_UsUrPeR_>
what type of printer is it?
09:44
nm, irrelevant
09:44
<sbalneav>
Several different kinds, all HP's though
09:44
ranging from HP 4100's to HP 4250's
09:44
<_UsUrPeR_>
are they using Ooo 3 or a later version of 2?
09:45
<sbalneav>
2.4.1
09:46
<_UsUrPeR_>
The weirdest thing about that is how the problem is fixed.
09:46
<sbalneav>
It's almost like openoffice sends down something that gets the printer into a confused state
09:47
printing out of Firefox, or other applications doesn't exhibit the problem
09:47
Only OO.o
09:47
<_UsUrPeR_>
do HP printers have font sets in memory?
09:48selffik has joined #ltsp
09:48
<sbalneav>
I beleive so, but we use Bistream Vera Sans usually around here, so that will normally be downloaded to the printer anywho.
09:49
<_UsUrPeR_>
Have you attempted changing the font to print to prior to rebooting the printer?
09:49
god that's weird
09:49
It's like the printer forgets how to normally interpret a cetrain font type or something :P
09:50
err... certain. Especially strange when it prints web pages properly.
09:50
Have you considered a script that would convert a document to .pdf prior to printing? :)
09:50
<sbalneav>
Yeah, and it's one of these things that happens infrequently - once or twice a week.
09:50CAN-o-SPAM_ has quit IRC
09:51CAN-o-SPAM_ has joined #ltsp
09:51
<sbalneav>
Enough to be annoying to the users, but not something you can trigger reliably, so you can wireshark the datastream, or look at the postscript produced, or whatever.
09:52
<_UsUrPeR_>
guh.
09:52
that reminds me of issues we'd experience while installing Asterisk phone systems.
09:53
by the time we would get to a user's desk to hear the problems with voice quality, the user will have already unplugged their phone
09:53
!pastebot
09:53
<ltspbot>
_UsUrPeR_: "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.
09:54CAN-o-SPAM_ has quit IRC
09:54Basti_dash has quit IRC
09:55Q-FUNK has joined #ltsp
09:58selffik has quit IRC
09:58selffik has joined #ltsp
10:06primeministerp has quit IRC
10:07pimpministerp has joined #ltsp
10:07primeministerp has joined #ltsp
10:09
<_UsUrPeR_>
I'm having a problem getting a client to mount an NFS dir on boot.
10:10
here's the current /etc/fstab
10:10
<ltsppbot>
"_UsUrPeR_" pasted "NFS mount problem" (3 lines) at http://pastebot.ltsp.org/109
10:10
<_UsUrPeR_>
does anyone see any problems in there?
10:10
if I mount -n the dir, it works fine. Just not from boot
10:10
adding auto to the mount options has no effect
10:12alkisg has joined #ltsp
10:14davidj has quit IRC
10:14
<ogra>
?? where do yu put that fstab ?
10:14
it gets generated on boot usually
10:15
<_UsUrPeR_>
fstab's in /opt/ltsp/i386/etc/fstab
10:16
written to the client ltsp image
10:16
<ogra>
right
10:16
<_UsUrPeR_>
good question though. I noticed that I had not specified it was in the chroot
10:16
<ogra>
make sure to have teh right stuff for teh rootfs in there as well and set CONFIGURE_FSTAB=False
10:17
(at least in ubuntu and debian that should work)
10:17
<_UsUrPeR_>
where does configure_fstab go?
10:17* _UsUrPeR_ is using fedora 9
10:17
<ogra>
lts.conf
10:17
ah, not sure warren ported that bit
10:17
check the initscript
10:17
in ubuntu and deban fstab is generated from ltsp-client-setup
10:18mattwalston has joined #ltsp
10:18
<_UsUrPeR_>
ahh
10:19
<mattwalston>
Anyway to make a second client image for diffrent hardware? I don't see an option in the build client script
10:19
<ogra>
why would you do that ?
10:19
unless its a totally different CPU architecture
10:20
<mattwalston>
ogra: I gave up fixing the issue with the via devices so I put an Xorg.conf in the client image and I have other machines that I don't want to have the "customized"
10:21
<ogra>
well, you create /opt/ltsp/i386/etc/X11/xorg.conf-via and define X_CONF=/etc/X11/xorg.conf-via in lts.conf for the clients with via chipsets
10:22
the rest will use the normal autodetection
10:22staffencasa has joined #ltsp
10:22
<ogra>
thats what lts.conf is for :)
10:23
<mattwalston>
ogra: sounds like a plan... i will give it a go
10:27dirigeant has joined #ltsp
10:31mattwalston has quit IRC
10:35vagrantc has joined #ltsp
10:36
<selffik>
Hello! How can i boot LTSP with PC without PXE?
10:38
<sbalneav>
selffik: You could use etherboot
10:38
<vagrantc>
or gPXE
10:38
!bootfloppy
10:38
<ltspbot>
vagrantc: "bootfloppy" is http://etherboot.anadex.de or ftp://k12linux.mesd.k12.or.us/pub/K12LTSP/Universal_boot_floppy
10:38
<selffik>
better with CD
10:38
i think
10:39
<vagrantc>
selffik: the etherboot.anadex.de has an .iso image too, i think.
10:39
<sbalneav>
There you go
10:39
<vagrantc>
you can also generate a CD, floppy, whatever, from rom-o-matic.net
10:39
<selffik>
yeah i will try thx guys
10:39
<sbalneav>
Or, you could do like I did, and go out and buy yourself an eprom burner :)
10:39
<vagrantc>
be sure to select "gpxe:all-drivers"
10:40
and then you'd get to call yourself old-school. :)
10:40
<sbalneav>
vagrantc: Does freegeek burn eproms?
10:40
<vagrantc>
sbalneav: no. we tried ages ago, and it just was too much hassle.
10:40
<sbalneav>
Kids these days.
10:40
<vagrantc>
sbalneav: by the time we got the hang of it, PXE implementations were getting pretty reliable
10:40
<sbalneav>
:)
10:41
The second time I booted an LTSP workstation, I went and burned an eprom for it :)
10:41
<vagrantc>
heh
10:42
because of some weird switch configuration though, we're actually back to floppies.
10:42* vagrantc grumbles
10:43
<sbalneav>
So, very first time I booted LTSP, I had a gutted Pentium 133 with 64 megs of memory.
10:43
No case, just the floppy
10:43siki has joined #ltsp
10:43
<sbalneav>
worked like a charm but I *REALLY* wanted to see LTSP boot with absolutely NO storage device at all.
10:44
<ogra>
thats what made you invent PXE and sell it to intel ?
10:44
<sbalneav>
So I grabbed an eprom, and the rtl8139 .rom image, and trekked down to the university, borrowed their burner, and birned the 'prom
10:44
<selffik>
NIC/ROM type where i can see it?
10:44* ogra considers an eprom a storage device
10:44tjikkun_work has quit IRC
10:45
<sbalneav>
Next morning, 8:30, I yanked the floppy, plugged the eprom into the rtl8139, and watched it boot automatically.
10:46
You know when you get one of those epiphanies when, like, angels are singing in the background? I has a smile on my face a mile wide. This was in October of '99
10:46
And the rest, as they say, is history. Here I am, 9.1 years later :)
10:47
selffik: What kind of network card is it?
10:47
<selffik>
Realtek
10:47
<sbalneav>
ogra: eprom storage device?
10:47
rm <file>
10:48
Kernel message: Please unplug eprom and place it under UV lamp for 20 minutes
10:48
:)
10:48hanthana_ has quit IRC
10:48
<sbalneav>
selffik: realtek... what? 8140? 8200?
10:48
<ogra>
well, does "storage device" define removability of files ?
10:49
<sbalneav>
No, guess not :)
10:49
<selffik>
81xx
10:49
<sbalneav>
selffik: Well, look at it with lspci, that will tell you
10:50
<selffik>
ok
10:51
<vagrantc>
the nice thing about the etherboot floppy/iso from anadex.de, is it detects your network card and uses the appropriate etherboot image automatically...
10:51
ditto for gPXE
10:53
<sbalneav>
vagrantc: Wonder if we should include a link to them in the doci?
10:53
err, doco
10:53
<vagrantc>
sbalneav: gPXE ?
10:53
<sbalneav>
Sure
10:53
and the anadex.de link where you can download the image
10:53
<vagrantc>
well, it's the recommended choice at rom-o-matic.net if you reference them
10:54
the etherboot project seems to have moved all development over to gPXE, as far as i can tell...
10:56CAN-o-SPAM_ has joined #ltsp
10:57
<_UsUrPeR_>
ogra: when you said "check the initscript", what were you referring to?
10:57
<sbalneav>
Hmm, ok, we should probably make mention of that in the docs.
10:58
I'll look at updating that.
10:58
<ogra>
_UsUrPeR_, whatever initscript fedora uses instead of ours :)
10:58
<_UsUrPeR_>
oh :|
10:58
<vagrantc>
sbalneav: i haven't really tried gPXE extensively, but overall it seems to be as capable as etherboot, with a whole lot more features.
10:59
sbalneav: and a little better overall design
10:59
i.e. it handles munging support for multiple card types into a single image better
10:59
<selffik>
where i can choose all drivers for gPXE?
10:59
<vagrantc>
selffik: are you using rom-o-matic.net ?
11:00
<selffik>
yes
11:00
<vagrantc>
there should be a drop-down menu, probably 3c5 something
11:01
http://rom-o-matic.net/gpxe/gpxe-0.9.5/contrib/rom-o-matic/
11:01
drop-down menu: choose NIC/ROM type
11:01
<selffik>
yeah i found!
11:01
<vagrantc>
selffik: select gpxe:all-drivers
11:01
<selffik>
why they just dont put it at begin or end :)
11:05warren_ has joined #ltsp
11:06warren_ is now known as warren
11:07
<johnny>
ubuntu only has ltsp-client-launch _UsUrPeR_ iirc
11:07
err sorry
11:07
fedora
11:08
<_UsUrPeR_>
johnny: ok, thanks
11:08
<johnny>
ubuntu uses two init scripts ltsp-client and ltsp-client-core
11:08
warren's stuff is alot shorter :)
11:08
finally nvidia drivers showed up..
11:09
damnit pcspkr .. why did you want to become so annoying
11:09
<warren>
johnny: I just blacklist pcspkr
11:10
<johnny>
yeah.. done.
11:10
gone
11:10
<ogra>
poor pcspkr
11:11
<johnny>
so.. banshee 1.4 is pretty incredible
11:11
it has the tag editor i want..
11:16mattwalston has joined #ltsp
11:16
<mattwalston>
ogra: no luck with the change... it could be a config file issues, I will retry using MAC address of a test client
11:17
ogra: is it X_CONF or XF86CONFIG_FILE? I have conflicting docs
11:17
!docs
11:17
<ltspbot>
mattwalston: "docs" is For the most current documentation, see http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
11:19
<johnny>
both
11:19
X_CONF is neweer
11:19
so it is suggested you use that
11:19
the old way is for ltsp4 migrations
11:21alkisg1 has joined #ltsp
11:24Barbosa_ has joined #ltsp
11:25japerry has joined #ltsp
11:27
<sbalneav>
K, updated docs, pushed rev.
11:28
I'm going to also spend about 20 minutes right now making all web references in the docs clickable. We've got quite a number of links which are just text.
11:31CAN-o-SPAM_ has quit IRC
11:32hallmant has joined #ltsp
11:33
<hallmant>
Hello! I'm having some trouble getting an external HDD to show up on an LTS desktop. Is anyone available to help?
11:34alkisg has quit IRC
11:37spectra has joined #ltsp
11:37Barbosa has quit IRC
11:37
<selffik>
hallmant, HDD by USB?
11:40
<hallmant>
selffik: Hi! Yes. Sorry for the delay.
11:41
<sbalneav>
hallmant: is it an NTFS volume?
11:41
<hallmant>
The client is seeing the drive okay (ran dmesg)
11:41
No
11:41
It's ext3
11:41
I'm currently looking at https://wiki.edubuntu.org/DebugLocalDev
11:41
<sbalneav>
Is the user in the fuse group?
11:41
<hallmant>
Yup
11:41
And I can manually mount the drive on the client
11:42
<Gadi>
prolly isnt considered removable
11:42
<hallmant>
But it's not "automatically" happening
11:42
ohh
11:42
<sbalneav>
hallmant: So, this drive's in an external USB case?
11:42
<Gadi>
or it sucks too much current and is killing the kernel (if ur thin client has a weak power supply)
11:43
<johnny>
people often want non removable drives to be accessible Gadi
11:43
<Gadi>
johnny: ok
11:43
:)
11:43
<hallmant>
sbalneav: Yeah, that's what it appears to be. (It's a friend's device.)
11:43
Gadi: I don't think it's a weak supply
11:43
<Gadi>
ok
11:44
you can edit the udev rules to allow removable
11:44
(or maybe there is another way these days)
11:44* Gadi lives a bit in the past
11:44
<sbalneav>
Well, sounds as if the udev rule isn't firing, I agree.
11:45
Whice version of LTSP and host OS are you running?
11:45
<hallmant>
sbalneav: Yeah, that's right. I just checked
11:45
/var/run/ltspfs_fstab only lists the floppy
11:45
But not the USB HDD
11:46
<sbalneav>
So, something about that drive isn't triggering the udev event.
11:46
<hallmant>
sbalneav: It's ltsp5, and Ubuntu Hardy
11:48
Any ideas as to how to get this working? Or do I just tell my friend to try a different device? :)
11:48
He's basically using it as a glorified thumb drive (except the thing's heavy and ugly) *shrug*
11:50
<sbalneav>
Well, you might want to try changing the add line to be something like:
11:50
ACTION=="add", SUBSYSTEM=="block", ENV{ID_FS_TYPE}=="?*" RUN+=....
11:51
that might help
11:52
<hallmant>
sbalneav: I'm sorry; I don't follow. What add line is that?
11:53
<sbalneav>
in the ltsp udev rules in /opt/ltsp/i386/etc/udev/rules.d
11:53
You'll have to update the image afterwards with ltsp-update-image
11:54
<hallmant>
sbalneav: In that dir there are a lot of "rules" files... which one should I look at? (Sorry I'm not up on my udev :))
11:54
oh nm
11:54
Sorry... I see it now
11:55
<vagrantc>
ACTION=="add", SUBSYSTEM=="block", ENV{ID_BUS}=="usb", ENV{ID_TYPE}=="disk", ATTRS{removable}!="1", RUN+="ltspfs_entry add %k"
11:55
but you may need to change "ltspfs_entry add" to "add_fstab_entry"
11:56
<hallmant>
Whoo boy :) I'm glad you guys know what you're doing! It's greek to me...
11:57
Thanks very much for your help
11:57
I've got to run off here, but I'll give it a shot as soon as I'm back
11:58
<johnny>
vagrantc, is there a reason to have ENV{ID_BUS} == "usb" ?
11:59
<vagrantc>
johnny: because many USB disks show up as non-removeable, and we don't want real non-removable drives, only external USB drives
11:59
<johnny>
how come?
11:59
many people ask for local hard drive support
11:59
<vagrantc>
and many people asked to not have it
11:59
<johnny>
could be a configurable option :)
11:59
<vagrantc>
yes, that would be better...
12:00
i've been meaning to look at Gadi's configurability options ... for months.
12:00
johnny: basically, teachers would plug in laptops, and then students would have root access to their laptop's hard drive
12:00
<johnny>
aha
12:00
<vagrantc>
it's not expected behavior :)
12:00
<johnny>
i tried to use my gf's work laptop as an ltsp client
12:01
sadly the bios locks out changing to allow netboot :(
12:01
<vagrantc>
so, as a safety mechanism, the default is to not allow non-removable drives
12:01
i wonder if firewire drives have the same issue...
12:01
<loather>
need one to test?
12:01exodos has quit IRC
12:02
<vagrantc>
loather: wouldn't hurt ...
12:02
loather: plug it in, and what's it say for /sys/block/$DEVICE/removable ?
12:02
<loather>
ok. i have this firewire chassis and a 12G disk i'm not using for anything in particular.
12:02
oh, hang tight
12:03
<vagrantc>
i wonder what BUS type firewire comes up as...
12:03
<loather>
removable = 0
12:04
sbp2 emulates scsi
12:04
<vagrantc>
ok, so i think the current udev rules won't work for external firewire drives, then ... *sigh*
12:05
<loather>
:(
12:06
well, i'm gonna head into work. i'll be back around later
12:06
<vagrantc>
loather: what's udevinfo -qenv -n /dev/$DEVICE report for ID_BUS ?
12:06
or udevadm info -qenv -n /dev/$DEVICE
12:06
<loather>
ieee1394
12:06
<vagrantc>
ok, so we'd have to add another rule for that ... although is firewire sometimes mounted internally?
12:07
<loather>
the only time i've ever seen that done is on an obscure storage server
12:09
<vagrantc>
ok, so that should be pretty safe to add ...
12:09
<loather>
i'm thinking so
12:09
<vagrantc>
although Gadi's technique is way better ...
12:09
allows you to specify that stuff in lts.conf and such
12:10
<loather>
definitely would be a nice option
12:15Q-FUNK has quit IRC
12:16
<vagrantc>
eeyk. those patches are from early september...
12:17
<Gadi>
yeah, Ive been meaning to set aside some time to work on ltspfs
12:21
<vagrantc>
IFS tricks always scare me a little ...
12:24
<Gadi>
be not afraid
12:24
unless you tell me it is not POSIX compliant
12:24
:)
12:25
fwiw - spaces scare me more
12:25
<vagrantc>
i think it's fine, i just have to turn my head around a couple times to get used to it :)
12:30mattwalston has quit IRC
12:42
<Gadi>
vagrantc: do you know if virtual hard drives in virtual machines have unique serial numbers?
12:43
<vagrantc>
nope.
12:43* Gadi assumes it may depend on the vm tool
12:43
<vagrantc>
as in, i don't know.
12:50warren has quit IRC
12:54warren has joined #ltsp
13:10jammcq has joined #ltsp
13:10
<jammcq>
hellooooo #ltsp
13:12bjs_redhat1 has left #ltsp
13:13
<vagrantc>
Gadi: two minor bugs with your LOCALDEV_DENY code, but i've got it working pretty good i think
13:14
<Gadi>
vagrantc: cool - its been so long I forgot what I coded ;)
13:14jammcq has quit IRC
13:14
<vagrantc>
Gadi: need to move the LOCALDEV_DENY code above the cdrom handling
13:14
Gadi: and there was a small typo that made the script totally bork.
13:15jammcq has joined #ltsp
13:15
<jammcq>
hellooooo #ltsp
13:15
see if this one works better :)
13:15
<Gadi>
hey, jammcq
13:15
<jammcq>
hey gadi
13:15
<Gadi>
both workes
13:15
:)
13:15
<jammcq>
yeah but something goofy happens with my internet connection at my office
13:15
it's like they are throttling IRC or something
13:16
I can connect to the channel and send a message, but then that's it
13:16
I can't get any more IRC traffic
13:16
<Gadi>
speaking of internet, I put intrepid on my eeepc, and also got bluetooth and USB + cell phone as modem working
13:16
:)
13:16
and very solid
13:16
<jammcq>
wow
13:16
sounds way cool
13:16
<Gadi>
on xandros, the connection kept dropping
13:16
not on ubuntu
13:16
<jammcq>
sounds like it works better than intrepid on my Dell
13:16
<Gadi>
oh yeah?
13:17
i thought dell and ubuntu were lovers
13:17Guest7765 has joined #ltsp
13:18CAN-o-SPAM_ has joined #ltsp
13:20
<Guest7765>
testing from web interface
13:20
testing again
13:20
1
13:21Guest7765 has left #ltsp
13:21
<johnny>
neat
13:21
now .. if only joining was faster.. i could recommend it
13:21
too much whois on startup..
13:23CAN-o-SPAM_ has quit IRC
13:32brendan_ has joined #ltsp
13:38rcy`` is now known as rcy
13:38Guaraldo has joined #ltsp
13:39Guaraldo has left #ltsp
13:39bobby_C has joined #ltsp
13:41lucascoala has quit IRC
13:43alkisg1 has quit IRC
13:50mccann has quit IRC
13:55johnny has left #ltsp
13:55mccann has joined #ltsp
14:07
<vagrantc>
Gadi: seems like the ID_BUS:ata+ID_TYPE:disk syntax doesn't work.
14:07
Gadi: it just denies all things of the bus or disk type, of that case
14:08
in that case
14:08CAN-o-SPAM_ has joined #ltsp
14:08
<vagrantc>
Gadi: also, i'm curious why you used : instead of just = ?
14:09
LOCALDEV_DENY=ID_BUS=usb
14:09
think that should work fine.
14:09
<Gadi>
vagrantc: parsing nightmare
14:09
one sec
14:15
<vagrantc>
ahhh... eval and such
14:19
<jammcq>
getltscfg should handle the double '='
14:19
in the past, we could specify module options that contains '='
14:20johnny has joined #ltsp
14:22dmaran has quit IRC
14:24rcy` has joined #ltsp
14:24
<vagrantc>
jammcq: yeah, but actually writing code to handle the = in the ltspfs scripts is a little tricky.
14:25rcy has quit IRC
14:25
<vagrantc>
actually, i don't see how it's difficult...
14:25* vagrantc will try it
14:28nubae has quit IRC
14:30
<vagrantc>
seems like it has to be quoted in lts.conf, though
14:30
<_UsUrPeR_>
hey, what's the best way to pull the local ip of a machine in bash and assign it to a variable? I.E. localip = `some command that gives the ip`
14:34
<vagrantc>
what's the context?
14:35spectra has quit IRC
14:42ogra has quit IRC
14:43ogra has joined #ltsp
14:44
<vagrantc>
seems like current ltspfs requires quoted values when there's = in the data
14:45
<Gadi>
= makes things look ugly, too
14:45
:)
14:45
why make life difficult?
14:45
<rjune>
!g
14:45
<ltspbot>
rjune: "g" is Gadi!!!!!!!!!!!!!!!!!!!!!!!!
14:46rcy` is now known as rcy
14:46
<cliebow>
someone say difficult??
14:49
<vagrantc>
Gadi: i think the = make it look more clear ... but it appears that getltscfg doesn't properly handle that.
14:49
Gadi: at any rate, this patch would allow us to simplify the default rules a lot!
14:52
<johnny>
_UsUrPeR_, look in ltsp-client-launch for one method
14:52
using the ip command
14:52
or you can just awk the ifconfig
14:53
<sbalneav>
ip -o -f inet addr show eth0 | sed -nr 's|^.*inet[[:blank:]]*(.*)/.*$|\1|p'
14:55
That assumes your active interface is on eth0
14:55
and you're not interested in an IPV6 interface
14:56
If you wanted to do things "properly", i.e. look at which interface has a default route:
15:01
IFACE=$(route -n | awk '$1 == "0.0.0.0" {print $8}')
15:01
ip -o -f inet addr show $IFACE | sed -nr 's|^.*inet[[:blank:]]*(.*)/.*$|\1|p'
15:02
!seen davidj
15:02
<ltspbot>
sbalneav: davidj was last seen in #ltsp 18 hours, 29 minutes, and 2 seconds ago: <davidj> The point of this was to make it easier on you, not harder ;-)
15:02
<vagrantc>
although to *really* do it properly, you should use "ip route" and use magic to figure out if there happen to be multiple default routes...
15:03
<Gadi>
vagrantc: will you push your patch upstream or are you going to share an separate branch?
15:04
<vagrantc>
Gadi: the LOCALDEV_DENY stuff only does anything if it's set, so i think it's safe to push upstream.
15:04
Gadi: i might make a couple minor code cleanups, but it's basically ready.
15:04
<Gadi>
ok
15:04
will you make any aliases for common deniable things?
15:04
or just keep it generic?
15:05
<vagrantc>
it would be easy to do
15:05
<Gadi>
s/will you/should we/
15:05
<vagrantc>
although exactly what aliases to have
15:05
that's the hard part.
15:05
<sbalneav>
vagrantc: well, "more properly" :)
15:05
<vagrantc>
sbalneav: :)
15:06
<sbalneav>
I shouldn't have said "properly"
15:06
<vagrantc>
sbalneav: as should i
15:06
<sbalneav>
"slightly more general case" is what I should have said.
15:06* vagrantc says that sbalneav should have said that vagrantc would have said that ...
15:06
<sbalneav>
heh, the question "what's my IP address" is always a surprisingly complex one
15:07
<vagrantc>
no doubt.
15:07
seems so simple.
15:07
<sbalneav>
yeah
15:09
<vagrantc>
Gadi: i'm wondering if it shouldn't exit 0 instead of exit 1 ... udev might treat that as a failed rule and move on to the next...
15:09
Gadi: i seem to recall that when i was implementing the cdpinger from udev stuff
15:10cliebow has quit IRC
15:11rjune_ has quit IRC
15:14
<Gadi>
vagrantc: do we not want it to move on to the next rule if it fails?
15:14* vagrantc wonders what ID_BUS serial ata devices have
15:15
<vagrantc>
Gadi: i seem to recall weird things happening...
15:17
<sbalneav>
I've always wondered, with all the ballyhoo about /proc, that there isn't a /proc entry somewhere with NAME=VALUE pairs, that you could just read/grep
15:18
and get the IP address that way.
15:18
<vagrantc>
might be ...
15:18
or something close enough
15:18
<Gadi>
mac address
15:18
in /sys/class/net
15:18
<sbalneav>
rrrreaaaaly?
15:19
<Gadi>
sure
15:19
<laga>
do you want some python code to get the ip adress?
15:19
<Gadi>
/sys/class/net/eth0/address
15:19mccann_ has joined #ltsp
15:20
<sbalneav>
/sys/class/net/eth0/address appears to be the mac addrss
15:20
unless I'm reading it wrong
15:21
<vagrantc>
oooh.
15:21
<Gadi>
right
15:21
<vagrantc>
that looks *way* easier.
15:21
<Gadi>
IP must be in arp somewhere
15:22
I suppose you can use it with /proc/net/arp and crossreference
15:22
<sbalneav>
find doesn't report an arp in /sys
15:22
<Gadi>
:)
15:22
but ur still using awk or some such
15:22
hehe
15:23
ah, and no guarantee that it'll be in the arp table
15:23
<sbalneav>
right, might as well do my one liner.
15:23
It just seems to me you'd have something like /sys/class/net/eth0/ip_address or something
15:23
<johnny>
yeah.. why isn't there..
15:23
i notice that
15:24
that'd be sweet
15:24
<laga>
as i said, i do have some python code for that
15:24
<sbalneav>
With the completely anally-retentive way that they seem to stick everything ELSE in /sys or /proc, you'd figgure that'd be there for sore.
15:24
laga: I've got C code, python code, and I just pasted shell code :)
15:25
<Gadi>
well, you can have multiple IPs attached to one interface
15:25
<sbalneav>
We can get it, it's just that there's no ELEGANT way to get it :)
15:25
<Gadi>
and fancy multi-home crap
15:25
<laga>
sbalneav: yeah, that's very annoying
15:25
<Gadi>
and trunking
15:25
prolly too much to think about
15:25
:)
15:26
<sbalneav>
Gadi: /sys/class/net/eth0/ip_addr, /sys/class/net/eth0:1/ip_addr...
15:26
<Gadi>
right
15:26
what do I know? I just work here
15:26
<sbalneav>
but yeah, I suspect it's just one of those "oh, this is so hard, meh, let 'em use the ip command" type things.
15:27
<Gadi>
either that or it's a debate between kernel mode and usermode
15:27
and its all political
15:27
so no one does it
15:27
<sbalneav>
That too.
15:27
Or a combination of the above. :)
15:28
Short answer: "what's my ip address" is still an NP-Complete question :)
15:29
<johnny>
the question is really something like.. what is the ip address i use to get to $target
15:30
you have to know where you want to go, before you know how can get there..
15:30
<sbalneav>
bingo
15:30
<johnny>
profound.. yes .. i know
15:30
lol
15:30
<Gadi>
arp -a
15:30
:P
15:31mccann has quit IRC
15:33BrunoXLambert has quit IRC
15:33
<vagrantc>
something about this localdev_deny seems to be breaking my cdpinger from udev magic when i'm denying cdpinger devices ...
15:33
or devices that would get called with cdpinger
15:33
i get one cdpinger for the raw device, /dev/hdc, and one for /dev/cdrom
15:34
<sbalneav>
latest copy of the docs are up.
15:35
Tonight I'm going to make sure all the variables are up to date, and maybe shuffle around a few sections.
15:36
<vagrantc>
ah, it might be getting started from ltsp-client-* init scripts
15:36
cdpinger, that is
15:37
<sbalneav>
yeah, it is.
15:37
<vagrantc>
normally, udev calls it first, and it checks if it's started ... but the initscripts aren't respecting LOCALDEV_DENY ... so they start it up
15:37
or something like that, is my guess.
15:42primeministerp has quit IRC
15:43CAN-o-SPAM_ has quit IRC
15:44primeministerp has joined #ltsp
15:54
<vagrantc>
hmmm....
15:54
my cdpinger from udev isn't working quite like it used to ... i wonder if it's borked in other places as well
15:56
it's that IFS stuff!
15:56
i knew that was wonky.
15:57
we have to actually unset it, not set it to nothing
15:57
although, if it was previously set for some reason, we need to restore it
15:59
<Gadi>
ah, right
16:00
I was bit by that at some point since writing that code
16:00
I always unset it now
16:00
<vagrantc>
i've got it going in my branch ... almost ready to merge upstream
16:01
<Gadi>
nice
16:01
<sbalneav>
I'm looking forward to seeing that.
16:02
What I'd really like to see too is launching ltspfsd from udev as well
16:02
<vagrantc>
so i think some boolean aliases, too: LOCALDEV_DENY_USB ... LOCALDEV_DENY_LOCAL_DISK ?
16:03
<sbalneav>
That could be fairly simple too. I should just create a lock file for it somewhere in /var/run
16:03
<vagrantc>
sbalneav: i did that a long time ago :)
16:03
<sbalneav>
lol, I haven't had to look at ltspfs for so long :)
16:03
so, if you don't plug anything in, -fsd doesnt launch?
16:03
Saweet
16:03
<vagrantc>
pretty much
16:04
<sbalneav>
Is there much need/call for c-ifying jetpipe?
16:04
<vagrantc>
although i think we still call it from the init scripts if it isn't started ... but we can remove that now, in my opinion
16:05
sbalneav: if it reduces the memory footprint by a huge amount, maybe... but it makes coding and testing a lot trickier
16:06
<sbalneav>
Somewhere out there, I hope Richard Stallman was paying attention: I went through the docs and made all references to "Linux" read "GNU/Linux" where a distro was being talked about.
16:06hallmant has quit IRC
16:06
<dberkholz>
gentoo goes by gentoo linux
16:07
i don't care what rms says, that's what we call ourselves
16:07
similar holds true for everyone but debian, i suspect
16:07
<vagrantc>
sbalneav: Debian GNU/Linux
16:07
GLTSP
16:08
<sbalneav>
jammcq passed judgement on that one a long time ago: the "Linux" in LTSP just stands for the kernel :)
16:08
<jammcq>
:)
16:08
<vagrantc>
but was judgement then passed on jammcq ?
16:09
<jammcq>
prolly
16:09
<sbalneav>
Anyone looked at the docs lately?
16:09* Gadi thought kernel starts with 'k'
16:09
<vagrantc>
sbalneav: i look over your commits, but not the docs themselves :)
16:10
Gadi: no doubt. that's why it's called Debian GNU/kFreeBSD
16:10
so really, it should be GkLTSP
16:10
<sbalneav>
I'm going to work with LaserJock and get some packaging made up.
16:10
<vagrantc>
er, GNUkLTSP
16:10
<Gadi>
heh
16:10
<dberkholz>
so perhaps uGNU/kLinux then.
16:10* Gadi thought Gnuk meant "nothing"
16:10
<Gadi>
;)
16:11
gornischt
16:11
<sbalneav>
we can have a -docs package that installs the html and xml files that can be read by yelp or... what's the K desktops's XML helpfile reader called?
16:12* vagrantc always prefers plain-text docs
16:12
<johnny>
vagrantc, well transforming is easy enough :)
16:12
<jammcq>
sbalneav: i'm sure it starts with a 'k'
16:13
<vagrantc>
jammcq: you're just guessing!
16:13
<jammcq>
well, it's a safe bet
16:14
meanwhile, I need to head out. gonna interview a programmer to see if he's got what it takes to come work for me
16:14
<sbalneav>
I think docbook can produce plaintext as well.
16:15
We could add those to the make targets.
16:15jammcq has quit IRC
16:17bobby_C has quit IRC
16:20mccann_ has quit IRC
16:32
<vagrantc>
Gadi: proposals for some default LOCALDEV_DENY aliases?
16:33
<Gadi>
vagrantc: when I add my ghotofs patch to ltspfs, should I do a cdfs one, too?
16:33
<vagrantc>
what's cdfs?
16:33
<Gadi>
another fuse filesystem for turning audio cds into files
16:33
<vagrantc>
ooooh.
16:33
<Gadi>
there may be something newer these days
16:34
I gotta check the repos
16:34
:)
16:34
but, might be worthwhile?
16:35
I like the two aliases you came up with so far
16:35
I would say, alias stuff that is common and/or complicated
16:35
we can start with those 2
16:35
and add if needbe
16:35
but, the fewer lts.conf vars to support in the future, the better
16:36
so, lets not go too crazt
16:36
*crazy
16:36selffik has quit IRC
16:36
<vagrantc>
Gadi: any proposals for some LOCALDEV_DENY_* aliases?
16:37
<Gadi>
FLOPPY
16:37
maybe?
16:37
<vagrantc>
oh, we may need that at freegeek. yes!
16:38
although floppy is tricky...
16:38
<Gadi>
hehe
16:38chrisinajar has joined #ltsp
16:38
<Gadi>
right - internal, external, all..
16:38
<vagrantc>
as *real* floppies don't set ID_TYPE=floppy
16:38
though usb floppies do
16:38
<Gadi>
I would say, have it kill all
16:38six2one has quit IRC
16:38
<Gadi>
hmm...
16:38
so, how do you determine real floppy?
16:38
fro dev name?
16:39
<vagrantc>
yeah
16:39
<Gadi>
can we access that?
16:39
<vagrantc>
not in the deny rules
16:39
at least, not without more coding ...
16:39mccann has joined #ltsp
16:40
<vagrantc>
well, maybe it does ...
16:40alkisg has joined #ltsp
16:40
<vagrantc>
we could play tricks, i think ... i.e. case $device in ; fd?) ID_TYPE=floppy ;; esac
16:41
i wish the removeable value was as easy to test for
16:42
we can emulate it with ID_BUS=ata+ID_TYPE=disk,ID_BUS=scsi+ID_TYPE=disk ... and that should catch most internal drives
16:43
<Gadi>
right
16:43
<vagrantc>
but i'm not confident it'll catch everything
16:44
<Gadi>
btw
16:44
call it LOCALDEV_DENY_LOCAL_HDD
16:44
DISK is vague, I think
16:44
actually, it should be internal
16:44
right?
16:44
<vagrantc>
sounds good
16:44
<Gadi>
LOCALDEV_DENY_INTERNAL_HDD
16:45
and then you can pick off ide, scsi, sata
16:45
<vagrantc>
sata isn't hot-pluggable?
16:45
<Gadi>
even if it is - it's internal
16:45
if I am a user, I would distinguish between internal/external
16:45
<stgraber>
Gadi: no, there are motherboards with an external S-ATA connector
16:46
<Gadi>
not removable/not-removable
16:46
<vagrantc>
i've seen external sata ports ... but i'm not really familiar with the technology
16:46
<chrisinajar>
there are mobos with external
16:46
<Gadi>
right
16:46
but, the admin does not expect someone to walk up with an external sata drive?
16:46
<vagrantc>
so how do you distinguish external from internal sata, then?
16:46
<Gadi>
to a thin client
16:46
remember our use case
16:46
<vagrantc>
common use cases
16:47* vagrantc has no idea how common external sata drives are or will be
16:47
<vagrantc>
i'm always about 5 years behind the times, remember :)
16:47
<Gadi>
would you want someone to have access to internal sata drive but not external?
16:47
or vice versa?
16:47
<chrisinajar>
Gadi: on the other hand, a lot of places convert their desktops to thin clients as apposed to upgrading them. That's what we run into really really regularly...
16:47
<Gadi>
right
16:47
I dont dispute that
16:48
I am just saying, a person using LOCALDEV_DENY is doing so for a reason
16:48
<chrisinajar>
so if a place supports external drives, say an artsy place...
16:48
<vagrantc>
LOCALDEV_DENY_USB, LOCALDEV_DENY_FIREWIRE ... those are easy.
16:48
<chrisinajar>
i think external drives should be catagorized with USB sticks and stuff
16:48
<vagrantc>
LOCALDEV_DENY_FLOPPY will require a little work, but actually not too bad.
16:48
<Gadi>
fine - but artsy place+external sata drive?
16:48
<vagrantc>
Gadi: well, i will set a default for LOCALDEV_DENY on Debian
16:49
<Gadi>
that wants to not use internal sata drive?
16:49
<vagrantc>
chrisinajar: how do you distinguish external from internal sata on a programmatic level?
16:49
<chrisinajar>
does HAL know?
16:50
<vagrantc>
and i guess, technically there's external scsi and even ata ... but i'm willing to let that slide
16:50
<Gadi>
you know, there are internal USB ports, too :P
16:50
I think it is an overcomplication
16:50
<chrisinajar>
only modern boards have the external sata, so we can assume as much...
16:50
Gadi: i think if it's possible, we should support it...
16:50
<Gadi>
this is not a qu of supporting it
16:50
<vagrantc>
chrisinajar: patches gladly accepted :)
16:50
<chrisinajar>
haha, that's true :-P
16:50
<Gadi>
this is a qu of supporting the ability to disable an external sata but keep an internal one enabled
16:51
or vice versa
16:51
<vagrantc>
LOCALDEV_ALLOW ?
16:51
that would be tricky to implement, though...
16:51
<Gadi>
do you really see an admin requiring that ? or not willing to work around it?
16:51
<vagrantc>
well, actually, we just need to change the LOCALDEV_DENY rules into a function ...
16:52
<Gadi>
if I am the admin, I either disable all SATA and use external USB/FW drives OR allow all SATA and remove the internal sata drive
16:52
<vagrantc>
let's implement it when they want it. :)
16:52
<Gadi>
:)
16:52
<vagrantc>
LOCALDEV_DENY_LOTSA_DISK_TYPES=true
16:56captain_1agnus is now known as captain_magnus
16:58
<vagrantc>
Gadi: what sort of problems did you have with space-separated values?
17:01warren has quit IRC
17:04
<vagrantc>
seems like SATA comes out as ID_BUS scsi anyways ...
17:05
not sure on anything other than Debian etch, though (only machine i have access to with sata
17:05
i bet a lot of device types look like scsi devices
17:06
<Gadi>
vagrantc: space-separated values also need to be quoted in lts.conf
17:06
<vagrantc>
Gadi: ok.
17:07
<stgraber>
E: ID_TYPE=disk
17:07
E: ID_BUS=scsi
17:07
that's for my iinternal HDD in my lappy
17:07
*internal
17:07
<Gadi>
but, in general, they can be hard to work with when you need a for loop
17:07
<vagrantc>
stgraber: OS and kernel?
17:07
<stgraber>
2.6.27 Ubuntu
17:07
<vagrantc>
ok.
17:07
that's pretty much the same as what i saw ...
17:07
so nothing special to handle serial ata
17:08
LOCALDEV_DENY_INTERNAL_HDD seems kind of long for a variable name...
17:08
lot of room for typos
17:08
and foggy memories
17:10
what about simply LOCALDEV_DENY_HARDDISK ?
17:10
<Gadi>
does that deny usb hdd?
17:10
<vagrantc>
no
17:10
<Gadi>
im confused
17:10
;)
17:10
<vagrantc>
ok.
17:10
<Gadi>
hehe
17:10
<vagrantc>
i figured we could handle that in documentation, then.
17:11
but maybe it's way too vague
17:11
LOCALDEV_DENY_DOWHATIMEANNOTWHATISAY
17:12chrisinajar has quit IRC
17:12
<Gadi>
aliases will be boolean, right?
17:12
<alkisg>
How about LOCALDEV_DENY=pata,scsi etc.... (comma seperated values, not just True/False)
17:12
<vagrantc>
Gadi: yes
17:13
alkisg: already have that, these are "simple" aliases to extend on that with
17:13
<Gadi>
alkisg: LOCALDEV_DENY is already more flexible than that
17:13
these are the simpler aliases
17:13
<alkisg>
Ah, ok, shutting up... :)
17:14
<Gadi>
I assume LOCALDEV_DENY_INTERNAL_HDD will default to True
17:14
upstream
17:14
<vagrantc>
it's really just trying to come up with a simple enough name to warrant a boolean that handles ata, scsi and sata drives ...
17:14
don't know if pata are any different
17:14
Gadi: i'm not sure about upstream, but definitely in Debian.
17:14
<Gadi>
I think upstream should strive to have secure defaults
17:15
<vagrantc>
sounds reasonable, and makes it easier on me :)
17:15
<Gadi>
much like defaulting to ldm with directx off
17:15
does your patch implement the mount options, too?
17:15
<vagrantc>
of course, then we automatically enable this feature and the whole "it shouldn't break anything" doesn't quite hold true
17:15
Gadi: nope
17:16
Gadi: one feature at a time :)
17:16
<Gadi>
:)
17:16
yes, yes
17:17
<vagrantc>
so far i've got simple code for LOCALDEV_DENY_USB, LOCALDEV_DENY_FLOPPY ... and still struggling for a name for the LOCALDEV_DENY_INTERNAL_DRIVES_OF_YOUR_TYPICAL_TYPES
17:18
<Gadi>
vagrantc: do the floppies and cdroms get triggered in initramfs's udev?
17:18
or are they OS-present?
17:18
<vagrantc>
Gadi: i don't think so
17:18
haven't seen it, anyways.
17:18
<Gadi>
ok
17:19
because we may not have an lts.conf in /etc at that point
17:19
:)
17:19rcy has quit IRC
17:19
<Gadi>
oh
17:19
it's handles by delayed mounter anyways
17:19
so we just need to make sure its not in ltspfs_fstab
17:20
<vagrantc>
i have been testing these patches :)
17:25
<Gadi>
hehe
17:27
gotta go - vagrantc, I'll keep an eye out for the push notification :)
17:28
l8r
17:28Gadi has left #ltsp
17:28X0d_of_N0d has joined #ltsp
17:34
<vagrantc>
LOCALDEV_DENY_CDROM ?
17:43johnny has left #ltsp
17:46lacqui has joined #ltsp
17:46
<vagrantc>
hmmm... LOCALDEV_DENY_INTERNAL_DISK ... is that sufficient?
18:10warren has joined #ltsp
18:15alkisg has quit IRC
18:16johnny has joined #ltsp
18:16johnny is now known as Guest75644
18:18X0d_of_N0d has joined #ltsp
18:18Guest75644 is now known as johnny
18:19* Ryan52 hates how launchpad sets the to address to me. it's more of a ml than to me directly. and it screws up my searches.
18:20
<X0d_of_N0d>
Is anyone who has used hardy ltsp using intrepid ltsp?
18:21
<stgraber>
yes
18:21
<Ryan52>
X0d_of_N0d: do you have a real question?
18:21* vagrantc commits LOCALDEV_DENY patches
18:21
<Ryan52>
if so, then please do ask.
18:21
<X0d_of_N0d>
Ryan52: well, I was wondering if anyone noticed a performance difference between the two....
18:22
<johnny>
Ryan52, you have to search by the headers
18:22
<X0d_of_N0d>
Ryan52: so if no one in here used it, then there wouldn't be any point in asking my second question
18:22
<johnny>
launchapd has interesting headers
18:22
lots of poeple are using intrepidltsp
18:22
including myself
18:22
i don't know any performances differences
18:23
i noticed it is much more stable
18:23
my gnome sessions don't die anymore
18:23
and leave with just a background and a taskbar
18:23
err
18:23
leave just a background with no taskbars
18:23
i upgraded more for features and stabilizity
18:24
localapps support
18:26
uggh.. i hate nvidia :(
18:26
what am i supposed to do to configure my cards if i don't want an xorg.conf :(
18:27
<stgraber>
I'm still working on it (init script remaining ...) but Jaunty's LTSP will probably be a lot better and likely more stable too :) It's basically Intrepid's feature but well integrated and with I hope working auto-detection (I still have some broken hardware at the office) and a way to manually set most of the settings (xrandr, dpms, .. well Gadi's stuff)
18:27
johnny: with the binary drivers you need a minimal xorg.conf to make it load
18:27
<johnny>
actually you don't
18:27
<stgraber>
johnny: otherwise it'll load the opensource one
18:27
<johnny>
mine loads nvidia just fine :)
18:27
without an xorg.conf
18:27
<stgraber>
oh ?
18:28
<johnny>
the new xorg server seems to do the right thing afaict
18:28
<stgraber>
ok, so Xorg's driver choosing logic is really weird
18:28
<X0d_of_N0d>
stgraber: so have you noticed any performance difference in X?
18:28X0d_of_N0d has quit IRC
18:28
<stgraber>
didn't quite have the time to answer that one I'm affraid :)
18:28
<johnny>
stgraber, this is newer behaviour
18:28
<stgraber>
on my lappy (ATI binary) I need a xorg.conf or it'll load the radeon driver
18:29
<johnny>
you're using intrepid?
18:29
<stgraber>
yep
18:29
<johnny>
sure you didn't have that behaviour from before?
18:29
<stgraber>
nope, I only ran intrepid on that lappy :)
18:29
<johnny>
hmm.. who knows then..
18:29
at least the ati card i have in another machine
18:29
works with xrandr with the radeon driver
18:30
makes me happy
18:30
<stgraber>
yeah, mine too except I just don't get 3D and AIGLX :)
18:30
<johnny>
really?
18:30
:(
18:30
wrong card then
18:30
<stgraber>
yeah, I've one of the new Radeon HD ... so basically I get 2D working
18:30
<johnny>
stgraber, on fedora10.. you'd possibly get flickrfree bootup
18:31
between plymouth and gdm
18:31
<stgraber>
well, I only boot once a month, so I don't really care :)
18:31
the only feature I need is suspend-to-ram and I'm happy
18:31
<johnny>
hmm.. suspend-to-ram doesn't cut it for me sadly..
18:31
it means i'm offline
18:31
i want a beagleboard
18:31
or something similiar..
18:32
i wonder if i could gut this laptop and put a beagleboard in it :)
18:32
i'd probably be able to have something as thin as a macbook air.. :)
18:33X0d_of_N0d has joined #ltsp
18:33
<johnny>
X0d_of_N0d, i don't think any of us have noticed any performance difference..
18:34
seems the same
18:34
the real problem for me.. is all kernels since cfq got merged
18:34
it sucks on a desktop :(
18:35
<stgraber>
X0d_of_N0d: I'm using really big LTSP networks, so that mean the application servers are extremely powerful and so does my thin clients. I always had local-like performance on that setup.
18:37
<johnny>
stgraber, hook it up
18:37
gimme some of your machines :)
18:37rcy has joined #ltsp
18:37
<johnny>
hehe
18:38
<stgraber>
johnny: nah, that's my quadcores, well rather our customer's :)
18:38
but you need that to handle 5k thin clients
18:40
<johnny>
stgraber, i need th thin clients ;)
18:44X0d_of_N0d has quit IRC
18:46X0d_of_N0d has joined #ltsp
18:55staffencasa has quit IRC
18:58
<X0d_of_N0d>
hum...thanx
18:58X0d_of_N0d has quit IRC
19:03_gentgeen_ has joined #ltsp
19:03gentgeen__ has quit IRC
19:03X0d_of_N0d has joined #ltsp
19:31gonzaloaf_laptop has joined #ltsp
19:34
<sbalneav>
Evening all
19:34
<vagrantc>
heya
19:34
<stgraber>
hey sbalneav
19:35
<sbalneav>
Hey stgraber &amp; vagrantc
19:35
heh
19:35
too much docbook
19:35
<stgraber>
sbalneav: did too much html recently ? :)
19:36
<vagrantc>
happy to say i haven't sufferred any docbook today...
19:36
all shell and udev, for me.
19:37
<stgraber>
well, I'd prefer doing some docbook than debuging some udev rules :)
19:37
at least that's the feeling I got last time I had to debug the ltsp udev rules, basically working a day to only change the sequence number ...
19:37
<vagrantc>
it was mostly gutting the udev rules, so it wasn't so bad :)
19:38
<sbalneav>
I'm going to be doing more "manual" labour tonight. :)
19:38
<vagrantc>
heh
19:45japerry has quit IRC
19:57BrunoXLambert has joined #ltsp
19:58X0d_of_N0d has quit IRC
19:59X0d_of_N0d has joined #ltsp
20:02
<vagrantc>
night all
20:02vagrantc has quit IRC
20:50chrisinajar has joined #ltsp
20:51
<chrisinajar>
ugh... the shereton in boston uses actual routing rules to force you to use their guest login page as apposed to just a proxy...
20:53
<loather-work>
heh, so it's a real captive portal
20:55davidj has joined #ltsp
20:55
<davidj>
Anyone here using evolution email a lot? I'm seeing weird freezes from time to time, evolution has the CPU pegged at 100%.
20:55
<chrisinajar>
indeed. I was expecting a little ssh -D to let me do whatever I wanted :-P
20:58try2free has joined #ltsp
21:05jammcq has joined #ltsp
21:05johnny has left #ltsp
21:05
<jammcq>
hello #ltsp
21:05
<chrisinajar>
howdy!
21:11
<davidj>
how do, jammcq
21:11
food healing nicely
21:11
<jammcq>
food ?
21:12
<davidj>
food, foot, it seems I eat both lately.
21:12
;-)
21:20
<sbalneav>
Hello davidj
21:20
I'm working on a massive re-org of the docs, to get them into much, MUCH more sane sections.
21:21
<davidj>
I thought you wanted me to do that massive reorg ;-)
21:21
<sbalneav>
I should have ltsp-docs-trunk pushed up in another hour.
21:21
<davidj>
..to match the massive reorg I did of the bones in my foot.
21:21
(doc says it's really only a minor reorg, won't even need a rebuild)
21:22* sbalneav refrains from making the joke that he wants the first reorg to turn out better than the second.
21:22
<davidj>
he
21:22
<sbalneav>
I would never, EVER make a joke like that
21:22
uh, uh
21:22
<davidj>
Of course you wouldn't.
21:22
<sbalneav>
Oh, geez!
21:22
Can you guys see what I'm thinking????
21:22
GET OUT OF MY HEEEEAAAAAADDDDD
21:23
<Ryan52>
heh
21:23
<sbalneav>
"I seem to have lost my internal monologue"
21:23* jammcq knows exactly what sbalneav is thinking
21:23
<davidj>
Afraid we can't do that. Turns out Homeland Security likes being able to see what's in your head.
21:23
Nope, can't do that. Wouldn't be prudent.
21:24
<sbalneav>
Homeland security DEFINITELY shouldn't have any business in my head. CIA, maybe.
21:24
But the tinfoil hat keeps them away from me.
21:24
:)
21:24
<davidj>
:)
21:25
<sbalneav>
Anywho, I'll ping you in a bit.
21:25mccann has quit IRC
21:25
<davidj>
sbalneav: I've got to deliver a big project tomorrow, burning the midnight oil. I think I'll sign off now. Can you send me an email when you're ready? I'll get to it later this week.
21:25
<sbalneav>
gotta get back to majjjjjjjjjjd'a45kP
21:25
Okeidoke
21:26
<davidj>
thanks. Good night.
21:26davidj has quit IRC
21:26
<sbalneav>
NP
21:29chrisinajar has quit IRC
21:42warren has quit IRC
21:53warren has joined #ltsp
22:03CaScAdE^FarAway has joined #ltsp
22:05CaScAdE^1arAway has quit IRC
22:09twb has joined #ltsp
22:19johnny has joined #ltsp
22:27hanthana has joined #ltsp
23:01try2free has left #ltsp
23:02
<Ryan52>
sbalneav: do you just use a regular old editor for the docbook stuff, or do you use somethin specifically for docbook editing?
23:02lacqui has left #ltsp
23:02kaos01_ has joined #ltsp
23:02
<sbalneav>
Ryan52: I'd hardly call vim a "regular old editor" :)
23:03
Just vim.
23:03
<kaos01_>
hi, trying to get ltsp5 workingon fedora, when i add eth0 to the bridge interface networkign seems to die
23:03
<Ryan52>
sbalneav: but does it do anything special or just show you xml?
23:03
<sbalneav>
I haven't ever found an editor that handles docbook, and I want the formatting to be preserved, since it helps me to debug.
23:04
<twb>
conglomerate
23:04
<kaos01_>
do i need to enable ip forwarding also ?
23:04
<sbalneav>
Ryan52: Just shows me the XML. Only niceity's I've got on are syntax hightlighting, and autoindent.
23:04
twb: Looked at conglomorate
23:04
<twb>
Yeah, it's flaky as hell
23:04
<sbalneav>
Threw it out in about 10 seconds.
23:04
<twb>
That's why I use reST instead of docbook
23:06
<sbalneav>
represenational state transfer?
23:06
<twb>
No, restructured text.
23:06
http://docutils.sf.net/
23:07japerry has joined #ltsp
23:08
<sbalneav>
Not for me, thanks.
23:08
Too many years invested in docbook :)
23:10alkisg has joined #ltsp
23:11
<sbalneav>
!doco
23:11
<ltspbot>
sbalneav: "doco" is you can find the link to the latest, and most current upstream documentation at: http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspDocumentationUpstream
23:12
<sbalneav>
Anyone who's interested, massive restructuring to the docs, more stuff documented, and a lot of cleanup
23:14
<Ryan52>
14.4 doesn't mention LDM_AUTOLOGIN, and it probably should.
23:15
hrm. /me sees no example. "An example will serve to illustrate how to use this:"
23:16
see, I would do more than just tell you this, but my writing skillz (if they can even be called skillz) suuuuuuuuuuuuuuuuuuuuuuuuck.
23:17
<twb>
Certainly they would not be called "skillz"
23:18
<Ryan52>
:P
23:18
<sbalneav>
14.7's got more detail, but certainly, an example would be handy in 14.4
23:19
assert(build_rome == 1)
23:19
Error: assertion failed: build_rome != 1
23:20
<Ryan52>
what's that mean? 0_o
23:20* Ryan52 feels confused
23:21
<johnny>
rome wasn't built in a day
23:21
Ryan52, common expression about how huge things take a long time :)
23:21
huge efforts*
23:21
<Ryan52>
oh, ya, I get that. the assert part confused me >.<
23:22
<johnny>
hehe
23:24
<sbalneav>
If I've confused at least one person today, it was a good day.
23:24
<twb>
Why would `build_rome' be an integer in the first place?
23:24
<Ryan52>
exactly!
23:24
<twb>
It sounds like a procedure to me.
23:26
The problem with pseudocode is that not only is it incomprehensible to your readers, it's incomprehensible to the computer, too.
23:29
<Ryan52>
at least the author can (usually) understand it :P
23:30
<twb>
Well, at the time of writing
23:30
Not the author's future self, I expect
23:33
<alkisg>
sbalneav, in chapter 20, `apt-get install italc-client` should be italc-master
23:34
<johnny>
all programming languages are pseudo code to the machine language ;)
23:35
<twb>
johnny: hardly.
23:35
johnny: they are *isomorphic* languages.
23:37
<cyberorg>
it would be nice if the doc were in wiki format, we can go and fix things or add missing parts
23:37
<twb>
cyberorg: but then they wouldn't be printable
23:37
<cyberorg>
twb, there are scripts that convert wiki to docbook
23:37
<twb>
Crappy scripts.
23:39
<johnny>
you can check out the repo and edit it just fine too
23:39
<sbalneav>
cyberorg: What's stopping you creating a branch I can merge from? :)
23:40
<cyberorg>
sbalneav, i was commenting on things like "<alkisg> sbalneav, in chapter 20, `apt-get install italc-client` should be italc-master"
23:40
<sbalneav>
alkisg: ok, I've never used italc, so I'll take your word for it.
23:40
cyberorg: We had docs in wiki. wiki.ltsp.org. Checked out what it looks like lately? :)
23:40
<alkisg>
sbalneav, (thanks for the docs!) - https://help.ubuntu.com/community/UbuntuLTSP/iTalc
23:41
<cyberorg>
sbalneav, yeah, i keep putting in missing information three sometimes
23:42
<sbalneav>
wiki's tend to devolve into a disasterous mess, unless you have a very, VERY committed individual, or team, to CONSTANTLY clean them up.
23:45
committed, pushed, and wiki page updated.
23:45alkisg has quit IRC
23:46
<twb>
sbalneav: what they turn into, above all else, is a bunch of lists
23:57
<sbalneav>
Phhht, well that was silly
23:57
So, vagrantc wanted a text version
23:57
so, run docbook2txt, right?
23:58
bleh, all I get is HTML
23:58
Why? Apparently, the backend converts to html, then passes through lynx to do the text conversion. I didn't have lynx installed.
23:59
but it never gave an error.
23:59
Wierd.