|00:00||alkisg has joined #ltsp|
|00:00||Ahmuck-Jr has joined #ltsp|
www.adobe.com is causing a logout and a keyboard lockup
requiring a hard reboot of the client pc
|00:06||Ahmuck-Jr has quit IRC|
|00:49||vagrantc has quit IRC|
|02:00||twinprism has quit IRC|
|02:12||Q-FUNK has joined #ltsp|
|02:39||puff has quit IRC|
|02:42||puff has joined #ltsp|
|03:12||puff has quit IRC|
|03:15||puff has joined #ltsp|
|03:31||plamengr has joined #ltsp|
|03:37||Q-FUNK has quit IRC|
|03:37||Q-FUNK1 has joined #ltsp|
|03:37||Q-FUNK1 is now known as Q-FUNK|
|03:38||Q-FUNK has quit IRC|
|03:38||Q-FUNK1 has joined #ltsp|
|03:38||Q-FUNK1 is now known as Q-FUNK|
|03:38||Q-FUNK1 has joined #ltsp|
|03:39||Q-FUNK1 is now known as Q-FUNK|
|03:39||Q-FUNK1 has joined #ltsp|
|03:39||Q-FUNK1 is now known as Q-FUNK|
|03:39||Q-FUNK1 has joined #ltsp|
|03:39||Q-FUNK has joined #ltsp|
|03:43||Q-FUNK has quit IRC|
|03:56||alkisg has quit IRC|
|04:18||alkisg has joined #ltsp|
|04:31||rslrdx has quit IRC|
|04:32||Q-FUNK has joined #ltsp|
|04:34||alkisg has quit IRC|
|04:51||alkisg has joined #ltsp|
|04:51||alekibango has quit IRC|
|04:56||alekibango has joined #ltsp|
|05:04||K_O-Gnom has joined #ltsp|
|05:18||lucascoala has joined #ltsp|
|06:04||nubae has joined #ltsp|
|06:11||cosf has joined #ltsp|
|06:19||Basti_dash has joined #ltsp|
|06:40||alkisg has quit IRC|
|06:54||hanthana has joined #ltsp|
|06:56||Q-FUNK has quit IRC|
|07:23||cliebow has joined #ltsp|
good MOrning BTS1
|07:29||cliebow has quit IRC|
|07:30||cliebow has joined #ltsp|
|07:39||Egyptian[Home] has joined #ltsp|
|07:40||Eghie has joined #ltsp|
|07:48||Egyptian[Home]1 has quit IRC|
|07:53||lucascoala has quit IRC|
|07:57||Eghie has quit IRC|
|08:04||jarias has joined #ltsp|
|08:13||hanthana is now known as hanthana|afk|
|08:17||mssmss has joined #ltsp|
I recenty installed Edubuntu 8.04.1 for a classroom ltsp setup
but I can't get more than two clients up at a time ..looks like dhcp freezes
looks like other users have seen the same issue ...
any idea what could be causing this problem ?
If I turn on multiple clients, not more than two will boot .. others time out at DHCP
but if I wait until one systems boots to screen, then start the next one, then many clients can come up
this is not a good time for this channel.. really.
best are during weekdays..
|08:37||nd___ has joined #ltsp|
I just need a quick yes/no if possible: When running applications with LTSP they run on the server and I only get the view via X11? Or do they run on the client and performance of applications is manipulated by the power of the client too?
(Well, so I need two answers... dang)
nd___: both are possible,
by default everything runs on the server
but we have 2 technologies that allow for local apps, one called, ltsp-localapps
and the other fatclient
ltsp-localapps allows u to choose individual apps that run on the client, whereas fatclient makes everything run on the client
ah, I remember reading that
|08:40||Ahmuck has joined #ltsp|
i understand i can install a chip on the nic. does the chip have boot rom already on it or does one flash the nic to get the boot rom there?
good morning. i'm looking for "boot rom?". is anyone familiar with this method?
google is your friend Ahmuck
|08:43||nd___ has left #ltsp|
google is a dumb friend at that
well either u buy a blank rom, or a rom with the data on it
that should make it obvious enough
k, thx nubae that helps.
newegg has the entire card for 9.00 and that site has boot roms for 15.00
it's probably not a great card
intel cards all do PXE native
as do quite a few more recent cards
almost all cards do now, dont they?
for 8.0 i'd be willing to try
i'm using older cards, 10/100. the newer cards will do 100/1000
even the 10/100 intel cards do pxe
|08:51||pdjbarber has joined #ltsp|
|09:06||Ahmuck has quit IRC|
|09:06||Ahmuck has joined #ltsp|
|09:07||klausade has quit IRC|
|09:15||cyberorg has quit IRC|
|09:17||cosf has quit IRC|
|09:22||Ahmuck has quit IRC|
|09:23||Ahmuck has joined #ltsp|
Ahmuck, many already network boot just fine.
|09:38||chrisinajar has joined #ltsp|
johnny: with pxe, yes
mine don't :(
i do have a bunch od 3c905b-txnnm however
|09:48||mssmss has left #ltsp|
|09:55||warren_x has joined #ltsp|
|09:56||warren has quit IRC|
|09:56||warren_x is now known as warren|
|10:02||johnny has left #ltsp|
|10:07||joergel has joined #ltsp|
|10:14||Basti_dash has quit IRC|
|10:21||joergel is now known as joecze|
|10:23||joecze has left #ltsp|
|10:36||alkisg has joined #ltsp|
|10:52||jarias has quit IRC|
|11:02||hanthana|afk has quit IRC|
|11:13||mistik1_ has joined #ltsp|
|11:14||mistik1 has quit IRC|
|11:14||cliebow_ has joined #ltsp|
|11:14||mistik1_ is now known as mistik1|
|11:16||petre has joined #ltsp|
|11:28||vagrantc has joined #ltsp|
vagrantc: Hey hey
So, which version of Debian does the logout hang with?
We made a change late yesterday to both ltsp-trunk and ldm-trunk
we decided to just ditch the "DISP" enviornment variable, and just set DISPLAY properly
sbalneav: i'm guessing it's debian etch, but i've been testing with debian lenny with an older version of openssh-server
Not the version that comes with lenny?
sbalneav: yeah, i always wondered about that DISP thing
Does it work with the ssh version that comes WITH lenny?
sbalneav: nope, the version that comes with lenny seems fine.
Now, lenny's the latest stable, right?
etch is current stable, lenny's in a hard freeze
i don't intend to backport newer ltsp stuff to etch, so it's not a huge deal for me specifically...
I really, REALLY don't want to go back to the kill, since it leaves so many droppings around. So I'd like to figure out why it doesn't work on Etch when all indications are that it should.
but it suggests to me that it might have this issue if folks want to log into older OSes using ldm
Oh, so are we worrying uneccessarily?
the whole "ldm can log into any server running ssh" thing
We shouldn't need to do the "kill", it was always such a hack
i'm pretty sure that it has to do with being started using dbus-launch ...
i think ogra said that ubuntu doesn't use dbus-launch anymore
but for both icewm and gnome, it uses dbus-launch
wonder if i can find a window manager that doesn't use dbus-launch to log in
Maybe fedora doesn't either?
warren: You there?
Does Fedora use dbus-launch
|11:35||bjs_redhat has joined #ltsp|
Fedora uses dbus-launch, I just checked with Warren
the part that i find *really* strange, is even if i kill *all* running processes of the user, it *still* doesn't log out
Rel 5 does as well
whereas during the pdx hackfest, killing dbus-launch would actually cause it to finish logging out
Is that after the mods I did yesterday, or just the <> /dev/null patch?
sbalneav: with all the patches applied to current ldm-trunk, and i also tried applying your "2> /dev/null" addition with no luck
Yeah, yesterday I managed to get in proper sigchld handling. It's going to be a bit buggy for sure, and when I get back to winnipeg, they'll be some more patches from that.
So I'm wondering if that's causing some problems.
It seems to work here, under Ubuntu and Fedora
Buth that doesn't mean it doesn't need some more work
|11:44||* warren tests it on fedora again|
sbalneav: oh, looks like i didn't test with your most recent patch to ldm-trunk
looks like the DISP -> DISPLAY switch, plus a change with handle_entry(GTK_ENTRY(entry), window);
in the greeter
|11:54||cliebow_ has quit IRC|
ok, build current ldm-trunk still fails with debian lenny using an older version of openssh-server
fails to log out on the second and subsequent logins
first goes just fine.
|12:15||bjs_redhat1 has joined #ltsp|
|12:15||pmatulis_t60 has joined #ltsp|
in the logs, i see "ending ssh session" and "shutting down ssh session", the socket is still open (note: it seems to be missing the vtNN part)
|12:20||bjs_redhat has quit IRC|
|12:23||davidj has joined #ltsp|
now ldm is hanging on logout with newer versions of openssh ...
yes, now it is consistantly hanging for me.
vagrantc: do you see the ttyname call in the code?
sbalneav: i see some code mentioning ttyname
what's this about: ldm.vty = ttyname + 5;
Well, ttyname returns a "/dev/ttyxxxx" name
ah, i see a FIXME around it too
Yeah, we should be doing that a little better
so ttyname_r isn't returning a ttyname?
well, ldm.vty doesn't contain anything when it creates the socket ...
ok, lemme add some code to parse the string a little more intelligently
i've gotta eat and then run to work soon ...
|12:58||GodFather has joined #ltsp|
|12:58||warren has quit IRC|
|12:59||pmatulis_t60 has quit IRC|
vagrantc: ok, check out r939
lets at least see if it gives you an error message now
|13:20||alkisg1 has joined #ltsp|
|13:20||alkisg has quit IRC|
|13:20||alkisg1 is now known as alkisg|
|13:31||Q-FUNK has joined #ltsp|
sbalneav: well, interesting ...
sbalneav: that worked fine the first login, as usual, but when ldm respawned, it kept spitting out error messages ...
sbalneav: "couldn't determine the tty name"
|13:45||Basti_dash has joined #ltsp|
|14:12||GodFather has quit IRC|
|14:16||cliebow_ has joined #ltsp|
|14:16||Basti_dash has quit IRC|
|14:17||vagrantc has quit IRC|
|14:47||cliebow__ has joined #ltsp|
|14:52||alkisg has quit IRC|
|14:55||vagrantc has joined #ltsp|
|15:05||cliebow has left #ltsp|
|15:21||johnny has joined #ltsp|
He is the afk.
iv'e got a number of thin clients that fail to boot to a login screen. here is the error and the specs. should i be filing a bug?
Disconecting: que, disconnect, sock, done
[36.883906] hbd0: Recieve control failed (result -32)
before the Disconec... line, is this line Setting up LTSP client ...
the mobo is an AOpen AX6BC with a AOpen PG80/DVD video card an either a Acer G722 or a AOC 7Gir monitor
a number of them fail with the above error messages one simply fails with many messages so is suspect something odd is going on with the hardware in it, however as all the others are failing with the same error i suspect it's unsupported hardware ?
these are 450mhz, 256mb, and 8 or 16 mb video clients
i need to step out for the afternoon, as i'm exhausted trying to get this solution to work. msg me if you have any idea what it might be. i'm using ubuntu ltsp server installation, and i'll not cross post. if i there is no reply, i'm going to file a bug in both bug lists. i would have suspected this hardware was supported as a thin client
doesn't sound like a hardware problem to me.
[13:40] [vagrantc(+ei)] [3:#ltsp(+cnt)] [Act: 9]
[13:40] [vagrantc(+ei)] [3:#ltsp(+cnt)] [Act: 9] [#ltsp]
Ahmuck: unless it didn't detect your network card
vagrantc: it detects it, as i've been using these cards in other machines and ltsp on them to verify they work
one other note, if i put a 4.x G hard drive in and install xubuntu they work fine.
albeit in low graphics mode
but does the version of ltsp you're currently using detect the network card?
wouldn't it if the network card works with ltsp in another machine?
vagrantc: i'm confused a bit. if the network card is recognized in one machine, wouldn't it be recognized in this machine as well ?
it's a pxe 3com card and it boots a ltsp client in another machine
|15:49||Q-FUNK has quit IRC|
probably would be.
but it's not impossible for machines that seem identical to contain different parts on the inside.
they are identical. i put them togather this afternoon
though i *think* you'd get a different error message
i'll try a different video card when i get back next week and see if that is the problem
|15:51||alkisg has joined #ltsp|
on a nother note, we have a number of imac's osx 9.2 and i assume we can use those as well?
ah, it's already booted to the root filesystem ...
it's more complicated ..
and running the init scripts
you have to build the chroot on the imacs
and then copy it to your server
Ahmuck: sounds more like nbd is dying part way through the boot process ...
network boot dameon ?
anywho, i've not got the server ordered yet, thankfully, as i was hoping to use the older hardware, but may need to consider something else
network block device
it's one of the more recent ways ltsp mounts it's root filesystem over the network
arg, i'll stay. i'm going to cancel my afternoon
and try to de-bug this
vagrantc, there are some new strings in ldm (untranslated), how are translation-updates handled? I just generate a new ldm.po and send it to you? :)
k, swapped video card. it's the board
it appears that board is not compatible with ltsp though it does work for xubuntu
evne with a new video card, the error is the same
i'll post it as a bug for non-compatible hardware
|16:06||warren has joined #ltsp|
before i do, i'm going to pull both video cards and try them in a good machine i know that boots to ltsp client to verify it's not the video cards
alkisg: translation updates haven't been handled much at all lately :)
vagrantc, should I send you an updated version, or just wait for the new version you're all preparing?
alkisg: not sure ...
alkisg: might want to wait a bit, as there may be some significant changes in the near future ... so strings may change.
vagrantc, ok, I'll stay tuned! :)
alkisg: i've been meaning to take a look at making calls for translations soonish
vagrantc, general question about translations: I've seen a nice interface for translations in translations.launchpad.net (is that rosetta?). It's very easy for the translators, because it also proposes translations from other packages. How easy or hard would it be for the developers of some application to use that? (or, if there are any other reasons not to use it...)
|16:26||plamengr has quit IRC|
vagrantc, ok, ty
alkisg: at least, it's a closed project, and the debian translation teams have really found the translations recieved by way of rosetta to be of varying quality- often very disconnected and disjointed.
alkisg: very inconsistant translations, etc.
alkisg: but i haven't used it directly, merely heard word from various teams that have been working on translation projects in debian for years...
vagrantc: I see... I've had a long talk with others about the translation quality vs translation quantity, and I thought "well, better to have not-so-good translations than no translations at all", but I see the point here
|16:33||otavio has quit IRC|
It must become an open project first, and easy for the developers, and maybe with a ranking system for the translators, so that the developers would be able to estimate the quality...
...and "better" translators would be able to override "worse" translations... well, somethink like this would take time, I guess
|16:40||johnny has left #ltsp|
alkisg: it mainly seems to need to require an overall process for making consistant translations for a given application, and a good review process, most likely.
|16:44||loather has joined #ltsp|
vagrantc: Can you try 940?
Just pushed it.
As a further test:
disable localdevs, here's why:
We're <> /dev/null the Xsession, but NOT the ltspfsmounter remove all
i've actually been testing with localdev, local_app and sound disabled.
Hm, that was my other thought, that maybe ssh was waiting for output from ltspfsmounter
ok, then try enabling them :)
well, i've tested with them enabled as well.
i've been trying with all possible combinations :)
|16:52||jammcq has joined #ltsp|
Try 940, see if that makes any difference
though i've erred on the side of disabling stuff mostly, trying to get to the root cuase.
sbalneav: i'll see if i can wrangle some time to do so
I'll set up a lenny box when I get home.
sbalneav: could you cut-and paste your commit messages for your merge commit messages?
|16:53||johnny has joined #ltsp|
sbalneav: would be helpful :)
bzr log in trunk shows my local tree log messages
why should I duplucate?
bzr log -r -1
940 Scott Balneaves 2008-11-08 [merge]
Merged from Scott's tree
that's what i see
in mine, I see:
so i have to specify --long
to see the actual commit messages
Then if someone specifies --long, they're going to get duped messages
that seems not right
well, bzr tells me it's a merge and from where, so it's not right either to specify that in the commit message.
if redundancy of information is your primary concern.
it was just a small request, take it or leave it as you please.
alternately, briefly summarize what the merged commits contain
No, I'm fine with doing it, it's just that it seems redundant
both ways carry redundancy
ok, no problem
We were only using the vt tag in the control socket as a differentiator between multiple ldm instances running on the same thin client
so I just ditched the ttyname_r and per stgraber's suggestion, just using getpid()
so that should solve the whacky
control socket name problem
when I get home, I'll put together a Lenny box, and try to track down the ssh problem
now, does debian still use nfs + bindmount only, i.e. is the overlay mount in the kernel still broken?
default is nfs + bindmounts, yes.
there's options to use NBD and such
Does that work now? I remember Debian used to have a problem with either the ramfs, or the mounting of the ramfs over the nfs
it'd REALLY be sweet if we ditched the bindmounts, and could just get the ramfs mounting over either nfs or nbd
For me, anyway, the ram overlying the nfs would be really nice.
i've heard people have sucess with it using aufs
aufs with tmpfs/NFS overlays
vagrantc: Oh, also, I'm going to add a trap handler to the ldm screen script so that killing the screen script cleans up the XAUTHORITY file.
You'll see that pushed soon
i haven't tried it in a while, but NBD+aufs+squashfs worked fine.
We should look at ditching the bindmounts then, at some point.
i've also done NBD+ext2+bindmounts, which was nice.
allowed for updating the chroot without rebuilding it
that'd be good too.
NBD+aufs+ext2+tmpfs would probably also work.
we're just about to have dinner here, I'll be afk for a bit.
then I'll get the trap handler in.
|17:39||alkisg has left #ltsp|
sbalneav: 940 hung initiating the second login attempt
i'm supicious of the XAUTHORITY changes in that one.
i seem to recall having issues with XAUTHORITY needing to only be set when LDM_DIRECTX=true
seems like the XAUTHORITY file isn't being created
ah, it's only created ONCE in the screen script, but then deletes it somewhere else...
|18:06||K_O-Gnom has quit IRC|
|18:09||johnny has left #ltsp|
|18:17||captain_magnus has quit IRC|
|18:26||captain_magnus has joined #ltsp|
|18:28||killsalad has joined #ltsp|
hi all - how can I use terminal's local hard drive, somebody told me that local /tmp would help performance?
|18:51||Gadi has joined #ltsp|
|18:58||petre has quit IRC|
|18:58||nubae has quit IRC|
|18:58||jc2it has quit IRC|
|18:58||rcy`` has quit IRC|
|18:58||daduke_ has quit IRC|
|18:58||Hyperbyte has quit IRC|
|18:58||jbrett has quit IRC|
|18:58||artista_frustrad has quit IRC|
|19:01||gentgeen__ has quit IRC|
|19:03||gentgeen__ has joined #ltsp|
|19:03||petre has joined #ltsp|
|19:03||nubae has joined #ltsp|
|19:03||jc2it has joined #ltsp|
|19:03||rcy`` has joined #ltsp|
|19:03||artista_frustrad has joined #ltsp|
|19:03||Hyperbyte has joined #ltsp|
|19:03||daduke_ has joined #ltsp|
|19:03||jbrett has joined #ltsp|
|19:08||gonzaloaf_laptop has joined #ltsp|
|19:13||phantom has quit IRC|
|19:13||F-GT has joined #ltsp|
|19:50||Ahmuck has quit IRC|
|19:58||nubae has quit IRC|
|20:03||japerry has joined #ltsp|
|20:09||daduke_ has quit IRC|
|20:09||Hyperbyte has quit IRC|
|20:09||jc2it has quit IRC|
|20:09||petre has quit IRC|
|20:09||jbrett has quit IRC|
|20:09||artista_frustrad has quit IRC|
|20:09||rcy`` has quit IRC|
|20:10||petre has joined #ltsp|
|20:10||jc2it has joined #ltsp|
|20:10||rcy`` has joined #ltsp|
|20:10||artista_frustrad has joined #ltsp|
|20:10||Hyperbyte has joined #ltsp|
|20:10||daduke_ has joined #ltsp|
|20:10||jbrett has joined #ltsp|
|20:18||Ahmuck has joined #ltsp|
|20:25||warren has quit IRC|
|20:37||warren has joined #ltsp|
Gadi, sbalneav: If you have any good photos of the group I could use for my blog, please let me know.
|20:49||Ahmuck has quit IRC|
|20:50||cliebow_ has quit IRC|
|20:52||daduke_ has quit IRC|
|20:52||Hyperbyte has quit IRC|
|20:52||jc2it has quit IRC|
|20:52||petre has quit IRC|
|20:52||jbrett has quit IRC|
|20:52||artista_frustrad has quit IRC|
|20:52||rcy`` has quit IRC|
|20:57||petre has joined #ltsp|
|20:57||jc2it has joined #ltsp|
|20:57||rcy`` has joined #ltsp|
|20:57||artista_frustrad has joined #ltsp|
|20:57||Hyperbyte has joined #ltsp|
|20:57||daduke_ has joined #ltsp|
|20:57||jbrett has joined #ltsp|
|21:05||vagrantc has quit IRC|
|21:13||cliebow__ has quit IRC|
|21:18||davidj has quit IRC|
|21:22||vagrantc has joined #ltsp|
|21:27||jammcq has quit IRC|
|21:27||bjs_redhat1 has quit IRC|
|21:31||* vagrantc looks at all the pushes today|
good to doublecheck them - folks here have been drinkin
Gadi: mostly, it's my usual nitpicking :)
localapps menu stuff, though .. wheee
|21:39||petre has quit IRC|
vagrantc: Gadi: any idea what version of GNOME that will work with?
warren: i have no confirmation that it works at all :)
nor have i even tried it
|22:02||CaScAdE^1arAway has joined #ltsp|
|22:09||CaScAdE^FarAway has quit IRC|
|22:16||johnny has joined #ltsp|
|22:21||Ahmuck has joined #ltsp|
|22:38||hanthana has joined #ltsp|
|22:39||vagrantc has quit IRC|
|22:59||Gadi has quit IRC|
|23:10||hanthana has quit IRC|
|23:15||dtrask has joined #ltsp|
|23:15||dtrask has joined #ltsp|
|23:23||Ahmuck has quit IRC|
yum is really annoying on missing repositories..
|23:31||Ahmuck has joined #ltsp|
|23:38||alkisg has joined #ltsp|
johnny: how many distros do you use? :)
on which computers? :)
web developers tend to work with alot of different OS
or rather.. alot of different distros
|23:45||chrisinajar has quit IRC|
some servers are debian, some centos, some rhel, some ubuntu, some gentoo
on the desktop tho.. ubuntu, gentoo, and now fedora.. so i could understand rhel stuff better
that would confuse me way too much.
being a web developer is uber complicated..
the amount of variables that are completely out of your control is unumerable..
|23:49||Ahmuck has quit IRC|
|23:52||Ahmuck has joined #ltsp|
Ryan52, you should be lucky any web pages you look at even work :)
some even tell me they work! http://ryan52.info/ :D
to be a web developer.. you really have to understand the plumbing of the internet...
altho.. mail is often the most complicated..
i don't own a mac or any win boxes .. that's given me more brain power :)
i probably should be more proficient at bsd in general tho..