00:00 | <twb> By analogy, it's fantastically useful to be confident that all configuration is in /etc (or at a pinch, /var).
| |
00:00 | rcy`` has joined #ltsp | |
00:00 | <twb> If you get crazy bastard stuff like "oops, FDS config is in /opt/fedora-ds/etc" then you have more pain to do backups.
| |
00:01 | rcy``: yarr, I am everywhere!
| |
00:01 | <rcy``> twb: sup
| |
00:02 | <twb> rcy``: nuffin' much.
| |
00:02 | Just trying to get some sucky wx/cairo stuff to work on xterms for a customer
| |
00:03 | <rcy``> sounds like fun
| |
00:03 | <twb> rcy``: they require SuSE 10
| |
00:03 | <rcy``> using kiwi ltsp then?
| |
00:04 | <twb> kiwi apparently requires SuSE 11
| |
00:04 | So this will probably be yet another seat-of-pants op
| |
00:04 | <rcy``> ah
| |
00:04 | <twb> I have ordered extra rolls of gaffer tape especially
| |
00:05 | <rcy``> can you not just throw a debian box down somewhere?
| |
00:05 | <twb> I may be able to trick them into doing that, we'll have to see.
| |
00:05 | <rcy``> just dont tell them
| |
00:08 | <twb> That's not sufficient; I also have to make sure they don't find out on their own
| |
00:08 | Probably that just means something like "chsh -s tcsh" as root
| |
00:09 | <johnny> and /etc/lsb-release :)
| |
00:10 | <twb> They won't know to look there
| |
00:10 | <johnny> well who knows where hte output goes to
| |
00:10 | <twb> Our customers are by definition, dumb. Otherwise, they wouldn't need to hire us
| |
00:10 | <johnny> in some random gui program..
| |
00:11 | hmm..
| |
00:20 | <rcy``> there should be some debian packages to make the system look like any other distro, or os
| |
00:30 | F-GT has quit IRC | |
00:35 | <johnny> rcy``, you mean like that distro that is more macosx like?
| |
00:35 | directorywise
| |
00:35 | <rcy``> maybe, i dont really know
| |
00:36 | <twb> johnny: gobolinux?
| |
00:37 | <johnny> ueaj
| |
00:37 | err yeah
| |
00:37 | ok.. sleeptime..
| |
00:43 | F-GT has joined #ltsp | |
00:56 | hanthana has joined #ltsp | |
01:15 | twb has quit IRC | |
01:31 | Pascal_1 has joined #ltsp | |
02:02 | Pascal_1 has left #ltsp | |
02:05 | phantom has joined #ltsp | |
02:06 | F-GT has quit IRC | |
02:15 | tjikkun_work has joined #ltsp | |
02:38 | Q-FUNK has joined #ltsp | |
02:48 | alekibango has quit IRC | |
02:57 | alekibango has joined #ltsp | |
02:57 | bobby_C has joined #ltsp | |
03:31 | Q-FUNK has quit IRC | |
03:43 | bobby_C has quit IRC | |
04:05 | Q-FUNK has joined #ltsp | |
04:23 | nubae has quit IRC | |
04:37 | Q-FUNK has quit IRC | |
05:23 | otavio has joined #ltsp | |
05:36 | hanthana has quit IRC | |
05:37 | lucascoala has joined #ltsp | |
05:51 | alkisg has joined #ltsp | |
05:53 | alkisg has quit IRC | |
05:53 | alkisg has joined #ltsp | |
05:58 | Q-FUNK has joined #ltsp | |
06:02 | Q-FUNK has quit IRC | |
06:02 | Q-FUNK has joined #ltsp | |
06:10 | PG has joined #ltsp | |
06:11 | PG is now known as PG_away | |
06:19 | <Q-FUNK> well, that's it. today was my last day at Artec.
| |
06:32 | cliebow has joined #ltsp | |
06:32 | <cliebow> Good morning BTS!
| |
06:33 | <Q-FUNK> heh
| |
06:34 | <cliebow> just passing info..The Kozy cove should be open for breakfast
| |
06:54 | jarias has joined #ltsp | |
07:00 | <sbalneav> Morning all!
| |
07:01 | siki has joined #ltsp | |
07:02 | chrisinajar_ has quit IRC | |
07:02 | chrisinajar_ has joined #ltsp | |
07:02 | nubae has joined #ltsp | |
07:03 | warren has joined #ltsp | |
07:03 | warren_ has quit IRC | |
07:08 | <cliebow> sbalneav:imchecking the old Kozy Kove for their times,,just on general principle
| |
07:09 | pmatulis has joined #ltsp | |
07:10 | chrisinajar has joined #ltsp | |
07:17 | petre has joined #ltsp | |
07:20 | Blinny has joined #ltsp | |
07:20 | <sbalneav> We're already here and Hackerating, where the heck are you?
| |
07:21 | <nubae> sbalneav: will u be blogging about it some place?
| |
07:21 | <Blinny> Sorry, I couldn't make it. And, my C is a bit rusty.
| |
07:21 | (;
| |
07:23 | <nubae> Blinny: only ldm is in C, IIRC
| |
07:23 | <johnny> better be careful where you leave it laying then
| |
07:23 | <sbalneav> I dunno. I suppose I could fire up my blog. Haven't updated it in 1.5 years
| |
07:23 | <nubae> probably left it out in the rain
| |
07:23 | <sbalneav> nubae: ldm's in C, cdpinger and ltspfs is too.
| |
07:23 | <nubae> heh, I would appreciate it
| |
07:23 | <chrisinajar> i like your decimal precision in blog updating times...
| |
07:24 | <nubae> as I am sure would the others who cant make it :-(
| |
07:24 | <stgraber> hey sbalneav
| |
07:24 | <johnny> like me
| |
07:24 | i have to keep the store open..
| |
07:24 | <chrisinajar> who else is going to be here?
| |
07:25 | <sbalneav> Warren's a pretty consistent blagger, I think
| |
07:25 | Me, personally, I usually don't consider my life interesting enough to talk about
| |
07:25 | <nubae> sbalneav: yeah but you are more ubuntu friendly ;-)
| |
07:26 | well, we'd hope it was about ltsp, rather than the intracacies of your every day meanderings ;-)
| |
07:26 | <sbalneav> day 14685: got up ate breakfast, went to work, came home went to sleep.
| |
07:26 | day 14686: see 14685
| |
07:27 | <nubae> lol
| |
07:29 | johnny has left #ltsp | |
07:29 | johnny has joined #ltsp | |
07:30 | <sbalneav> http://ltspthinclient.blogspot.com/
| |
07:30 | Surprised I remembered my password.
| |
07:31 | <nubae> sbalneav: which geek is u in the photo?
| |
07:33 | <stgraber> nubae: it's: fgiraldeau, sbalneav, eric and myself
| |
07:33 | it was at UDS-Boston
| |
07:34 | <nubae> ah cool
| |
07:42 | hanthana has joined #ltsp | |
07:43 | <Q-FUNK> heh
| |
07:44 | Blinny has quit IRC | |
07:47 | <nubae> why does samba still suck so much to get working right?
| |
07:49 | <jimcooncat> nubae: moving targets take more ammo :-)
| |
07:49 | but I very much agree with you
| |
07:58 | cliebow_ has joined #ltsp | |
07:58 | mikkel has joined #ltsp | |
08:02 | lucascoala has quit IRC | |
08:03 | <alkisg> I want to run an xterm upon ldm. I used to do that with "DISPLAY=:7 xterm" on a client console. But in intrepid it doesn't work, I get "This program is an suid-root program...". What gives?
| |
08:03 | I see that X is loaded using some authorization file, does this have anything to do with the problem? Can I re-use this authorization file to launch xterm?
| |
08:05 | * got it! export XAUTHORITY=...
| |
08:07 | nubae has quit IRC | |
08:13 | Blinny has joined #ltsp | |
08:14 | nubae has joined #ltsp | |
08:14 | Gadi has joined #ltsp | |
08:16 | <Blinny> Anybody here have experience getting iTalc to work? Ideally with multiple servers?
| |
08:16 | <nubae> Blinny: its pretty straight forward, just install italc-master and italc-client
| |
08:16 | no need for anything on the desktops
| |
08:17 | <Blinny> nubae: Ok. I did that -- but when I run iTalc I only see one of my clients. Is there a special addressing requirement for adding new computers?
| |
08:17 | (and I shouldn't bother with adding the start_ica perl script and running from the chroot's /etc/rc.local ?)
| |
08:17 | <nubae> yes, localhost:11000 where the last 3 0s should be the last ip
| |
08:18 | no need for start_ica, at least I didnt need it
| |
08:18 | <Blinny> Ok great. Thanks.
| |
08:18 | And change localhost to otherservername for multiple servers?
| |
08:18 | <alkisg> Blinny, you should also transfer the keys for multiple servers
| |
08:18 | <nubae> yeah well I suppose that should work... not tried that, but on standalone computers, u can just put in their ip address
| |
08:18 | <Blinny> OK.
| |
08:19 | Transfer keys from where to where?
| |
08:19 | <alkisg> Blinny, man ica => look for the key paths
| |
08:20 | Both servers should have the same keys if the same clients are to be connected to them
| |
08:20 | <Blinny> OK, good. That's what I needed
| |
08:20 | I knew they were in /etc/italc
| |
08:20 | <alkisg> So transfer from server A to server B
| |
08:20 | <Blinny> Cool. Thank you very much.
| |
08:20 | <alkisg> np
| |
08:20 | <nubae> if u do get it working for multiple servers, add a little blurb to the ubuntu ltsp pages
| |
08:21 | <Blinny> You mean in addition to the _no_ blurb that is there now? ;)
| |
08:21 | <nubae> hehe, well there is something now on italc isnt there?
| |
08:22 | johnny has left #ltsp | |
08:22 | <alkisg> nubae, yeah, a screenshot and some headers... :P
| |
08:22 | <nubae> heh, ok, then yeah add a full howto for italc :-)
| |
08:22 | <Blinny> Yeah. https://wiki.ubuntu.com/iTalc is pretty useless
| |
08:23 | So I don't need to copy keys to the chroot?
| |
08:23 | <alkisg> ...and I think there should be a seperate page for LTSP/iTalc, the setup is different
| |
08:23 | Blinny, stgraber proposes local italc installation (to the chroot)
| |
08:23 | It avoids some CPU usage (or network traffic, I'm not sure)
| |
08:23 | <nubae> youre right... its in the wrong place even, it should be here: https://help.ubuntu.com/community/UbuntuLTSP/
| |
08:23 | <alkisg> But my TCs don't have enough RAM, I've never tried it
| |
08:23 | <Blinny> nubae: your link references mine
| |
08:24 | <nubae> I know
| |
08:24 | <alkisg> nubae, I think 2 pages are needed, a teacher may install iTalc without LTSP
| |
08:24 | <nubae> get rid of that link alltogether
| |
08:24 | yeah but then it shouldnt be on the ltsp page
| |
08:24 | UbuntuLTSP is for all LTSP related stuff
| |
08:24 | <Blinny> Let me make sure I can get it working first.
| |
08:24 | <nubae> the other italc should be left alone
| |
08:25 | italc link
| |
08:25 | <alkisg> yeap
| |
08:25 | <nubae> the 11000 thing trips people up
| |
08:25 | because in the official docos, it says 10000
| |
08:27 | can someone do me a favour and tell me if this brings up anything: http://84.112.76.128/moodle
| |
08:27 | <Blinny> Ok. I've got tons of those perl scripts flying around. Let me kill them all and start from scratch with a fresher chroot
| |
08:27 | <nubae> its my home computer, and I need to show it externally to someone
| |
08:28 | <Blinny> Waiting for...
| |
08:29 | Are you on a 2400 baud modem? (;
| |
08:29 | <nubae> heh, no
| |
08:29 | <alkisg> Blinny, if you put it to the chroot, the ports are different (=10000)
| |
08:29 | <Blinny> Looks like it's gonna time out
| |
08:29 | <nubae> firewall might still be blocking stuff... got a stupidly complex router
| |
08:29 | <Blinny> alkisg: I'm going for the easiest route first. I have a few spare CPU cycles on my server.
| |
08:30 | so I'll aptitude install and try adding clients. Be right back - need to reboot this client
| |
08:30 | Blinny has quit IRC | |
08:31 | <alkisg> nubae, firefox timed out
| |
08:32 | nubae, did you ever get avahi autodetection working for a normal italc ltsp installation?
| |
08:32 | <nubae> normal, as in no ltsp?
| |
08:33 | <alkisg> no, normal as in ltsp :)
| |
08:33 | <nubae> sure
| |
08:33 | <alkisg> 11000 port and stuff
| |
08:33 | Blinny has joined #ltsp | |
08:33 | <nubae> worked every time for me
| |
08:34 | <Blinny> Hrm. And what about servers with multiple NICs and multiple subnets?
| |
08:34 | <alkisg> What did you do to make it happen? It never worked for me (4 hardy and intrepid installations)
| |
08:34 | <nubae> just installing italc-master and italc-client
| |
08:34 | not in the chroot
| |
08:34 | <alkisg> Bah... I must be doing something wrong
| |
08:34 | (yeah, just a normal installation, not in the chroot)
| |
08:34 | <nubae> which version of italc?
| |
08:34 | I used .9
| |
08:34 | <alkisg> Intrepid = 1.0.9
| |
08:35 | <nubae> yeah thats the one
| |
08:35 | <alkisg> Blinny, I think these are ssh keys, so they may contain more than 1 known hosts
| |
08:35 | <Blinny> Perhaps new things have happened in Intrepid; I'm still on Hardy
| |
08:35 | Right, but when you're adding clients localhost:11??? doesn't work if you have multiple subnets
| |
08:36 | because the ??? could address multiple clients - see what I mean?
| |
08:36 | <nubae> probably not, unless u try ipofserver:11ipofclient
| |
08:36 | <alkisg> Blinny, no! :) (I don't see what you mean)
| |
08:36 | <nubae> or hostname:11last3digitsofclientip
| |
08:37 | localhost is localhost
| |
08:37 | can only have one subnet
| |
08:37 | <alkisg> Blinny, localhost is the same for *any* number of subnets you may have, no?
| |
08:37 | <Blinny> Say I have two subnets: 192.168.0.0 and 192.168.0.1
| |
08:37 | <nubae> u cant attache mutliple subnets for one ethernet interface
| |
08:37 | <alkisg> Blinny, use 127.0.0.1:11???
| |
08:37 | <Blinny> Clients in both. If I add client localhost:11005 that could be 192.168.0.5 or 192.168.1.5
| |
08:37 | <nubae> thats the same subnet
| |
08:38 | oh 1.1
| |
08:38 | then that would be on a different interface
| |
08:38 | <Blinny> They're on two NICs also. My server has four NICs.
| |
08:38 | <nubae> one is localhost
| |
08:38 | well each one has a hostname
| |
08:38 | <alkisg> Blinny, the real IPs are *not* used in iTalc
| |
08:38 | <Blinny> so I could do 192.168.0.1:11005 and 192.168.1.1:11005 ?
| |
08:38 | <alkisg> (unless you put it in the chroot)
| |
08:38 | <nubae> for example, but Id use hostname
| |
08:38 | <Blinny> instead of localhost?
| |
08:38 | <nubae> no, for extra servers
| |
08:39 | <Blinny> It's the same server.
| |
08:39 | <alkisg> Blinny, 127.0.0.1:11???
| |
08:39 | <Blinny> same server, multiple NICs
| |
08:39 | <nubae> sure, 127.0.0.1 is only attached to 1 nic
| |
08:39 | the others would probably be like 127.0.1.1
| |
08:39 | and so on
| |
08:39 | <alkisg> Guys, I'm not getting this. The *port* number is what matters
| |
08:40 | So with 127.0.0.1:11??? you can reach *any* client, no matter the subnet
| |
08:40 | <nubae> yes, port and server its connecting to
| |
08:40 | alkisg: not really
| |
08:40 | CAN-o-SPAM has joined #ltsp | |
08:40 | <nubae> say u have 4 subnets
| |
08:40 | <alkisg> Blinny it talking about one server with multiple NICs, no?
| |
08:40 | <nubae> each has 256 client possibilities
| |
08:40 | <Blinny> if you're stripping off the 192.168.0 and 192.168.1 parts
| |
08:40 | alkisg: Yes.
| |
08:41 | <nubae> but u need to define the subnet its pointing to
| |
08:41 | thats why we do localhost:11000
| |
08:41 | and not just 11000
| |
08:41 | <alkisg> Blinny, I have 4 NICs in my lab, and iTalc worked just fine when I tried subnetting
| |
08:41 | *in my server
| |
08:42 | <nubae> subnetting how?
| |
08:42 | <alkisg> nubae, why do you need to define *any* ips different from 127.0.0.1 to iTalc?
| |
08:42 | iTalc works locally, it doesn't transfer any data over the network
| |
08:42 | <Blinny> Ok. So forget anything but localhost
| |
08:42 | <nubae> but say u have italy running on one server
| |
08:42 | italc
| |
08:43 | <Blinny> What are the ??? then for two clients 192.168.0.15 and 192.168.1.15 both on the same server
| |
08:43 | <nubae> oh wait... same server
| |
08:43 | never mind...
| |
08:43 | johnny has joined #ltsp | |
08:43 | <Blinny> How would you add both clients with one IP 192.168.0.15 and another 192.168.1.15
| |
08:44 | <alkisg> Blinny, ok, I got what you mean. Maybe ica will refuse to start?
| |
08:44 | (for the second client)
| |
08:44 | <nubae> or u can use the hostname of that card
| |
08:44 | <johnny> uggh
| |
08:44 | maybe you guys can help with this..
| |
08:44 | AUDIT: client $num rjected from local host
| |
08:44 | <nubae> anyway, blinny, experiment and let us know, I think we just dont know the answer :-)
| |
08:44 | <johnny> and my mouse isn't working
| |
08:44 | <alkisg> Blinny, you could modify ica-launcher to add the 3rd byte of the IP, e.g. 12015 for the second client
| |
08:45 | <johnny> the ctrl + alt + F1 shortcut works, but i can't type either
| |
08:45 | <nubae> ah thats an idea
| |
08:45 | <johnny> in the gdm login field
| |
08:45 | <Blinny> but then I couldn't go higher than 99
| |
08:45 | <nubae> johnny: I've had that happen, turned out my permissions were wrong somewhere
| |
08:45 | Egyptian[Home] has quit IRC | |
08:45 | <nubae> but cant remember how I fixed it
| |
08:45 | <johnny> somewhere? :)
| |
08:45 | lol
| |
08:45 | thanks :)
| |
08:45 | lol
| |
08:45 | <nubae> hehe
| |
08:46 | <alkisg> Blinny, 3rdbyte * 256 + 4th byte
| |
08:46 | <nubae> /tmp or /home or /root or some such
| |
08:46 | <Blinny> Ok. Starting from scratch and purging libitalc italc-client and italc-master
| |
08:46 | <nubae> alkisg: in theory one should be able to do 127.0.1.1:11015
| |
08:47 | and 127.0.0.1:11015
| |
08:47 | <alkisg> nubae, I got what you were talking about. I don't know that much about tcp/ip and ports, I don't know what would happen if the same port is open on 2 different nics of the server
| |
08:48 | But I doubt ica-laucher would work ok with that
| |
08:48 | <Blinny> OK. installed libitalc italc-client and italc-master
| |
08:48 | started up, no autodetected computers
| |
08:49 | Egyptian[Home] has joined #ltsp | |
08:49 | <alkisg> nubae, 127.0.0.1 is *the same* with 127.0.1.1 in every way, no? Can someone install a service to 127.0.0.1:11000 that *doesn't* answer to 127.0.1.1:11000?
| |
08:49 | <Blinny> added localhost:11015 for client 192.168.1.15 with MAC and says 'host is unreachable'
| |
08:50 | <nubae> right because thats probably not localhost
| |
08:50 | <Blinny> AFAIK 127.0.0.0/8 is the same
| |
08:50 | <nubae> check your /etc/hosts
| |
08:50 | <alkisg> Blinny, you have to login to the client first
| |
08:50 | <Blinny> alkisg: 192.168.1.15 is the client I'm logged in from. Let me log out & back in
| |
08:50 | brb
| |
08:50 | Blinny has quit IRC | |
08:50 | <alkisg> no
| |
08:50 | ...too late
| |
08:51 | A restart is needed somewhere... I don't remember where, maybe on the server?
| |
08:51 | <nubae> i have 127.0.0.1 and localhost and 127.0.1.1 homeserve
| |
08:51 | in /etc/hosts
| |
08:51 | but maybe I'M doing something wrong :-)
| |
08:52 | six2one has joined #ltsp | |
08:52 | <nubae> also all my clients are defined in /etc/hosts
| |
08:52 | ok, see what u mean alkisg, shouldnt make a difference, you're right
| |
08:52 | Blinny has joined #ltsp | |
08:52 | warren_ has joined #ltsp | |
08:53 | <Blinny> Bummer. No love.
| |
08:53 | <alkisg> I saw ahavi "advertisements" for each italc clients, but I never got it to work
| |
08:53 | Blinny, try 10015
| |
08:53 | warren has quit IRC | |
08:53 | warren_ is now known as warren | |
08:53 | warren_ has joined #ltsp | |
08:53 | <Blinny> alkisg: That worked.
| |
08:53 | <nubae> hah.. damn i got it confused it is 10000 for ltsp
| |
08:54 | 11000 in normal desktop mode
| |
08:54 | <Blinny> alkisg: So this shows the desktop of client 192.168.1.15
| |
08:54 | Excellent. How would I then show the desktop of 192.168.0.15 :)
| |
08:54 | <alkisg> Blinny, ok, and what about 192.168.0.15??? :)
| |
08:55 | <johnny> what is the privileges of the contents of /var/lib/gdm ?
| |
08:55 | <alkisg> I think ica-launcher will refuse to start
| |
08:55 | <johnny> uggh..
| |
08:55 | <Blinny> johnny: drwxrwx--T
| |
08:55 | (Hardy Heron)
| |
08:56 | <johnny> can you give me more info
| |
08:56 | like the directory and contents/
| |
08:56 | and the usernames and groups
| |
08:56 | <Blinny> 192.168.1.1:10015 also works for 192.168.1.15 (where 192.168.1.1 is the server's IP)
| |
08:57 | <nubae> so try 192.168.0.1
| |
08:57 | Subhodip has joined #ltsp | |
08:57 | <Blinny> I think I need someone from that subnet to log out and log in, yes?
| |
08:57 | Cannot operate on non-logged in computers?
| |
08:57 | <nubae> right
| |
08:57 | <johnny> i could swear that it works if i install the nvidia-driver
| |
08:57 | which is weird
| |
08:58 | maybe it's not related
| |
08:58 | <nubae> yeah
| |
08:58 | <Blinny> OK. All my clients logged in before I did the purge/install of italc
| |
08:58 | <nubae> could be X related then
| |
08:58 | Blinny: when u do the sshkeys, u have to update the image
| |
08:59 | well if its using chroot only I guess
| |
08:59 | * nubae shuts up... | |
08:59 | <alkisg> Blinny, to work without logging in you have to install it to the chroot
| |
09:00 | <cliebow_> ogra:whta did we do to tell the thin client mac to look for nbdroot on 2001..i altered ineted.conf..but the kernel needs to know..
| |
09:01 | <alkisg> Blinny, change this line of ica-laucher: PORT=`echo $LTSP_CLIENT | awk -F . '{print \$4}'`
| |
09:01 | <Blinny> I haven't done anything in the chroot yet
| |
09:01 | <nubae> can someone try nubae.selfip.com/moodle
| |
09:01 | <alkisg> so it doesn't work without logging in. :)
| |
09:01 | <Blinny> For now that's fine with me.
| |
09:01 | <alkisg> about ica-laucher, to take $3 into account
| |
09:01 | <Blinny> I think the wiki should have two sections then - one for installing to chroot, one for installing to server
| |
09:02 | <alkisg> Blinny, both 192.168.0.15 and 192.168.1.15 work simultaneously?
| |
09:02 | * nubae hopes Blinny can remember all this to put into the ubuntu wiki :D | |
09:02 | <johnny> uggh
| |
09:02 | <Blinny> alkisg: I haven't had a 192.168.0.15 log in yet.
| |
09:02 | <johnny> my stuff is totally screwed
| |
09:02 | i can't even open the store :(
| |
09:02 | guess it's paper and pen for a bit
| |
09:02 | <nubae> rebuild the image??
| |
09:02 | <Blinny> alkisg: For me the wiki is easy:
| |
09:02 | sudo aptitude install libitalc italc-client italc-master; Add SERVERIP:10??? where ??? is the last 3 digits of the client IP
| |
09:02 | <alkisg> Blinny, yeah, piece of cake... :)
| |
09:03 | <Blinny> Let me go start up a client over there.
| |
09:03 | One second - running to different building
| |
09:03 | <nubae> ^^ can someone try http://nubae.selfip.com:81/moodle
| |
09:03 | <alkisg> eeek...
| |
09:03 | <cliebow_> nubae:just a sec
| |
09:03 | <alkisg> nubae, doesn't look good...
| |
09:03 | about to timeout again
| |
09:03 | <nubae> with port 81 also not?
| |
09:04 | jcastro_ is now known as jcastro | |
09:04 | <nubae> darn... stupid router...
| |
09:04 | <cliebow_> not good here..
| |
09:05 | pmatulis_t60 has joined #ltsp | |
09:05 | <alkisg> I wonder why avahi advertizing is needed for iTalc when it is used in a normal (=no chroot) LTSP setup... ps aux gives all the logged on clients
| |
09:05 | <stgraber> it's not needed, it's just not turned of :)
| |
09:06 | and it makes it work with workstation outside of the ltsp server
| |
09:06 | <johnny> weird
| |
09:06 | it's telling me that it's doing stuff that isn't in my config file
| |
09:06 | like loading DRI
| |
09:06 | even though i don't have a capable card without nvidia driver
| |
09:07 | <alkisg> stgraber, thanks for the confirmation! I thought it was weird... :)
| |
09:07 | Subhodip has quit IRC | |
09:09 | <nubae> someone try now http://nubae.selfip.com/moodle
| |
09:09 | <alkisg> nubae, better!
| |
09:09 | You don't have permission to access /moodle on this server.
| |
09:09 | <nubae> forbidden
| |
09:09 | cool
| |
09:10 | stupid router only allows 4 mapped ports...
| |
09:11 | <Blinny> 192.168.0.1:10??? works also
| |
09:11 | YEA!
| |
09:12 | <nubae> so its for different subnets like we said?
| |
09:12 | subnet:10000 and subnet2:10000 ?
| |
09:12 | <alkisg> Blinny, both of them logged on simultaneously?
| |
09:14 | <nubae> http://nubae.selfip.com/moodle/
| |
09:14 | wroks now?
| |
09:14 | works
| |
09:14 | <alkisg> nubae, yeap
| |
09:15 | Duh... an S3 virge on a lab with Celerons 600 gives 30 fps. The *same* card on another lab with Celerons 450 gives 3 fps!!! Not only with video, it's slow even with scrolling text in gedit or MS notepad! It's not a CPU problem, something's wrong with the boards of the second lab... :'( arrrggg...
| |
09:16 | <Blinny> Yes, works.
| |
09:16 | <alkisg> cool!
| |
09:17 | <Blinny> using address 192.168.1.1:10015 and 192.160.0.1:10015
| |
09:17 | davidj has joined #ltsp | |
09:18 | <monteslu> warren, you arround?
| |
09:18 | <Blinny> Unfortunately it is horrifically slow, even on 100MB client NICs on a 1000MB server NIC & switch
| |
09:18 | <monteslu> ldm-scritp is taking over 3 minutes to run on my thin clients
| |
09:18 | this makes the login horribly long
| |
09:18 | Ahmuck has joined #ltsp | |
09:19 | <alkisg> Blinny, you mean like in video broadcasting or in just displaying a demo screen of the server?
| |
09:19 | <Blinny> remote control of a client - ~30 seconds between screen refreshes
| |
09:19 | <johnny> going back to the old school way of configuring X fixed the problem
| |
09:19 | <nubae> italc is not fast... thats normal
| |
09:19 | <johnny> :(
| |
09:19 | <warren> monteslu: are you sure it isn't a network problem?
| |
09:19 | <johnny> this makes me sad
| |
09:19 | <nubae> unless u have gigabit
| |
09:19 | I guess
| |
09:20 | <monteslu> warren, I am not
| |
09:20 | <johnny> but whatever.. it works
| |
09:20 | now i need to make sound work
| |
09:20 | jammcq_ has joined #ltsp | |
09:20 | <warren> monteslu: if you plug one client directly into the server, same problem?
| |
09:20 | <jammcq_> hey all
| |
09:20 | <monteslu> but why is the cpu pegged on that script
| |
09:20 | <nubae> johnny: with restricted drivers?
| |
09:20 | <johnny> no restricted drivers
| |
09:20 | <alkisg> Blinny, something's wrong, 30 seconds is too much... 5 seconds I think is about normal
| |
09:20 | <johnny> if i want glx on the clients.. can't use restricted drivers
| |
09:20 | iirc
| |
09:20 | <nubae> well, at least thas a plus right :-)
| |
09:20 | <johnny> restricted on the server that is..
| |
09:20 | <nubae> right
| |
09:20 | <Blinny> Ah wait - came back now
| |
09:20 | <johnny> hmm.. i wish i had another machine to run ltsp on
| |
09:21 | <warren> monteslu: I've seen that and other issues before, it was always a problem with the switch or cables
| |
09:21 | <johnny> or another to run the POS software even .
| |
09:21 | * nubae wishes he had a bunch of test servers to run LTSP on | |
09:21 | <johnny> i'm gonna have some pics of my terminals soon tho :)
| |
09:21 | and my store :)
| |
09:21 | <monteslu> warren, I'll give that a shot now
| |
09:22 | <nubae> today I passed by a second hand shop.... they had a motherboard with 128mb ram + ethernet for 6.99 €, a monitor for €19.99 and keyboard and mouse for little more
| |
09:22 | that means a thin client could be built for under 40€
| |
09:22 | well, getting the box is the hard part
| |
09:22 | <Blinny> Is there something that requires clicks to be doubleclicks through iTalc?
| |
09:23 | * alkisg doesn't know, always used gnome or windows with italc | |
09:24 | <Blinny> Yeah - something with mouse clicks isn't quite right
| |
09:24 | They're not being passed through to the client correctly
| |
09:25 | siki has quit IRC | |
09:25 | <Blinny> Mouseovers bring up the gnome panel popups but I can't click on the logout/shutdown button or the menus
| |
09:26 | <alkisg> Blinny, try closing one of these 2 clients
| |
09:26 | cosf has joined #ltsp | |
09:26 | <alkisg> *logging off
| |
09:27 | <Blinny> No same effect
| |
09:27 | dtrask has joined #ltsp | |
09:27 | <Blinny> I even double-clicked the client name in the left taskbar to disconnect fully, then reconnect, then re-remote controlled.
| |
09:27 | Hrm.
| |
09:28 | <dtrask> Hey!
| |
09:28 | <cliebow_> Ho!
| |
09:28 | <Blinny> D'oh!
| |
09:29 | closing iTalc and re-opening lost all the computers I'd added
| |
09:29 | <dtrask> hey if you guys need anything like spare computers...servers...laptops....access points...switches....pizza...beer....scotch....let me know and I'll bring it
| |
09:29 | <jammcq_> dtrask: hey
| |
09:29 | Basti_dash has joined #ltsp | |
09:29 | <jammcq_> we've got awesome connectivity here, thanks to Warren
| |
09:29 | <ltsppbot> "cliebow" pasted "Introduction: ==== Jim McQuill" (33 lines) at http://pastebot.ltsp.org/94
| |
09:30 | Basti_dash has quit IRC | |
09:30 | <cliebow_> dtrask: this is what we are talking about
| |
09:30 | <dtrask> jammcq_: What'd Warren do to make it awesome?
| |
09:30 | <cliebow_> rewired the whole place
| |
09:30 | <warren> not really
| |
09:30 | <cliebow_> ;-]
| |
09:31 | <warren> I told the owner what to do, and he & son-in-law rewired it last night in the dark
| |
09:31 | <chrisinajar> the best time to do rewiring...
| |
09:31 | <warren> also figured out that the router they have in the office is crap, so they're buying new hardware to make it more reliable
| |
09:31 | <cliebow_> weehee!
| |
09:32 | <Blinny> It _looks_ like the button is being depressed but nothing comes up
| |
09:32 | <dtrask> very cool...talked to Dave yesterday...he's excited to have us all there again :-)
| |
09:32 | He get's a network upgrade each time ;-)
| |
09:33 | <jammcq_> dtrask: what time do you plan on coming?
| |
09:33 | <warren> I told him how to wire the rest of the motel and his house, 5 or 6 routers in total, all connected with ethernet.
| |
09:35 | <dtrask> late....I'll be arriving close to midnight or maybe a little earlier
| |
09:35 | <ninkendo> anyone with a working setup want to pastebin their /etc/initramfs-tools/initramfs.conf?
| |
09:35 | mine seems to be not working
| |
09:36 | spectra has joined #ltsp | |
09:36 | <ninkendo> http://arslinux.pastebin.ca/1247910
| |
09:36 | <dtrask> jammcq_: I get to leave here around 10 pm....then depending on how much beer and pizza I have to pick up.... ;-)
| |
09:36 | <jammcq_> we'll prolly have plenty of both long before you get here
| |
09:37 | <dtrask> jammcq_: LOL
| |
09:37 | jammcq_: I'll call when I leave just to make sure....that is if anyone can still converse with me.. ;-)
| |
09:37 | <Blinny> Looks like starting iTalc wipes out ~/.italc settings every time. Should this be?
| |
09:38 | <monteslu> warren, plugging directly to the server doesn't seem to be working, I think I need to use another stand alone switch in the middel or make a crossover. There's alot of people logged in so i think I'll do that after school
| |
09:38 | warren, in the meantime, is there any debuggin i can do on ldm-script to find the exact line that is taking long?
| |
09:39 | DonSilver has joined #ltsp | |
09:39 | <monteslu> I guess i could just echo messages and pipe them to the /var/log/something huh?
| |
09:39 | <warren> sounds like you have 100mbit interface on the server?
| |
09:39 | <ltsppbot> "Blinny" pasted "initramfs.conf" (52 lines) at http://pastebot.ltsp.org/95
| |
09:39 | <alkisg> Blinny, no, it's a bug
| |
09:39 | <monteslu> nope gb
| |
09:40 | clients are all 10/100
| |
09:40 | <Blinny> alkisg: Wow. Ok, yuck. That's a pain when you have 50 terminals... but I guess for now I can connect when-needed
| |
09:40 | alkisg: Where are those bugs listed?
| |
09:40 | <warren> monteslu: was this working well before with the same hardware but different software?
| |
09:40 | <monteslu> warren, yup
| |
09:40 | <dtrask> jammcq_: where'd ya'll go for breakfast?
| |
09:40 | <alkisg> Blinny, I just saw this in intrepid, it wasn't there before
| |
09:40 | <monteslu> been using ltsp for 6 or 7 years now
| |
09:40 | <Blinny> I'm in Hardy
| |
09:41 | <monteslu> this is the first time I've seen this long login
| |
09:41 | <alkisg> Blinny, when I used Hardy it still had 1.0.7... it has 1.0.9 now?
| |
09:41 | <ninkendo> Blinny: ah, nm, I was looking at my update-initramfs.conf, and not my intramfs.conf, my initramfs.conf looks like yours
| |
09:42 | <Blinny> alkisg: I added the http://ppa.launchpad.net/edubuntu-italc-devel/ubuntu source
| |
09:42 | <alkisg> Blinny, try rm ~/.italc - or make a copy of globalconfig.xml and personalconfig.xml while you added the clients and italc is still running
| |
09:42 | * ninkendo is still stumped as to why eth0 is not being found on his thin client | |
09:42 | Q-FUNK has quit IRC | |
09:42 | <Blinny> alkisg: Right. Let me try the plain sources from hardy without the other edubuntu repo
| |
09:42 | <jammcq_> dtrask: donuts and coffee at the hotel
| |
09:43 | tomorrow, we'll do that place in bar harbor
| |
09:43 | seemed to be some interest in that
| |
09:44 | <dtrask> jammcq_: Yay! Love Jordans....been there several times already this year...never tire of it ;-)
| |
09:44 | <monteslu> warren, /usr/share on the chroot is read only?
| |
09:44 | <ninkendo> I have an intel 82562 onboard card in here (should be using the e100 driver), and ipconfig is not finding any cards on bootup... LTSP 5 on hardy, stock 2.6.24-21-generic kernel with up-to-date initrd
| |
09:44 | <monteslu> i want to add some debugging to ldm-script, do i have to rebuild the client for each change?
| |
09:45 | <alkisg> heh... scrolling text in gedit does take 100% CPU in a 450 Pentium client
| |
09:45 | <Blinny> brb
| |
09:45 | Blinny has quit IRC | |
09:47 | <warren> monteslu: the entire chroot is read-only from the client's perspective
| |
09:47 | <monteslu> that makes it difficult then :)
| |
09:48 | but id does log to /var
| |
09:48 | <warren> monteslu: /opt/ltsp/i386 is where the chroot lives, you can edit things in there
| |
09:48 | <nubae> monteslu: or set to nfs
| |
09:48 | <warren> monteslu: and you are probably using NFS root mode, so no need to build any images after you edit things
| |
09:48 | nubae: nfs is default in fedora
| |
09:48 | <nubae> ah
| |
09:48 | <warren> you can switch to nbd easily if you want
| |
09:48 | <monteslu> naw, nbd gave me hell on ubuntu
| |
09:49 | <nubae> once the environment is built it is faster
| |
09:49 | but for development, its hell indeed
| |
09:50 | so u can still set nfs to be read/write
| |
09:51 | Blinny has joined #ltsp | |
09:51 | <Blinny> Okay, awesome. With the vanilla Hardy .debs it Just Works
| |
09:51 | (with the exception of saved states on iTalc quit)
| |
09:51 | CAN-o-SPAM has quit IRC | |
09:52 | * Blinny goes off to edit the wiki | |
09:54 | chrisinajar has quit IRC | |
09:56 | <Blinny> Now should I really create an LTSP-specific wiki page for iTalc and not reference wiki.ubuntu.com/iTalc ?
| |
09:57 | <fgiraldeau> hi all
| |
09:58 | <alkisg> Blinny, yeah! :)
| |
09:58 | cosf has quit IRC | |
09:58 | <johnny> nubae, you use habari?
| |
09:58 | i have it installed right now.. trying to set it up myself..
| |
09:59 | <monteslu> warren, you sure all i have to do is edit files in /opt/ltsp/i386?
| |
09:59 | script didnt get my changes when i restarted a client
| |
10:01 | nevermind it did
| |
10:01 | it just didn't update the my log file
| |
10:01 | <Blinny> Ok.
| |
10:01 | <monteslu> so i guess /var/log is read only too ?
| |
10:01 | on the client
| |
10:02 | gm04030276 has quit IRC | |
10:02 | <Blinny> Is there a bug page for not-saving .italc that I can reference in the wiki?
| |
10:03 | <alkisg> Blinny, I didn't see any, but I got mine to save configuration files,
| |
10:03 | I don't remember what I did though, some deleting, some manual editing... then it just worked
| |
10:03 | <Blinny> heh
| |
10:03 | I'll leave that part of the wiki up to you, then (;
| |
10:04 | <alkisg> Blinny, you should delete ~/.italc anyway, because it contains 1.0.9 version files and now you have 1.0.7
| |
10:04 | Maybe it'll work if it gets recreated
| |
10:06 | <nubae> johnny: I do
| |
10:06 | its a super little piece of soft
| |
10:06 | very easy to manage and add stuff to
| |
10:07 | alkisg: that would be pretty important, I couldnt figure out how to save settings either
| |
10:09 | <monteslu> hmmm.... I had updates running in the this morning, and ldminfod was updated....
| |
10:09 | things are logging in faster now, magically
| |
10:09 | maybe there was a fix
| |
10:10 | or maybe my script is bombing in just the right way :)
| |
10:10 | i'm replacing it with the backup and trying again
| |
10:11 | <alkisg> nubae, I'll try to reproduce how I did it some time later
| |
10:11 | <monteslu> ok, logins have gone from 3-4 minutes, down to about 30 seconds
| |
10:11 | I'm confused
| |
10:11 | <dtrask> monteslu: be thankful
| |
10:11 | :-)
| |
10:11 | <monteslu> yes, that too!
| |
10:12 | hope it sticks
| |
10:13 | I would like to be better able to debug this stuff though, and get a better knowledge of everything that happens with ltsp5
| |
10:13 | <dtrask> monteslu: agreed
| |
10:14 | <monteslu> no i gotta figure out how to kill the bridge thing and get this back on one NIC
| |
10:14 | Q-FUNK has joined #ltsp | |
10:15 | <monteslu> also how to run localapps for firefox & flash, and figure a way to log which kids are going to which sites
| |
10:15 | should be a little easier now that i can possibly be tied to a terminal ip
| |
10:16 | s/i/it
| |
10:17 | <dtrask> monteslu: logging which sites is easy if you do proxy authentication....
| |
10:17 | <monteslu> dtrask, not a bad idea. I'm already using ipcop for a filter. I guess I could just make it a captive portal
| |
10:17 | <Blinny> OK, https://help.ubuntu.com/community/UbuntuLTSP/iTalc created
| |
10:18 | <monteslu> as long as i can do an ACL for updates
| |
10:19 | <nubae> warren is ubuntu the only distro doing nbd then?
| |
10:19 | <dtrask> monteslu: I use IPCop and can tie it to my Samba/LDAP server (you can use AD or whatever you use) and simply use authentication in such a way that when a kid clicks Firefox...they are challenged to log in...now log files are bound to the username
| |
10:20 | <nubae> Blinny: nice :-)
| |
10:21 | <warren> nubae: fedora's nbd works great
| |
10:21 | nubae: we just don't do it by default
| |
10:21 | <nubae> yeah neither does debian
| |
10:21 | <Blinny> Let me know if you have suggestions for improvements. Or make them yourself! (;
| |
10:21 | <monteslu> dtrask, my smbldap is no more :(
| |
10:21 | <nubae> so its just ubuntu doing that? Suse uses kiwi, and I guess gentoo is nfs too
| |
10:22 | <dtrask> monteslu: what are you using for auth?
| |
10:22 | <monteslu> I've been thinking about doing another, though I really miss the script
| |
10:22 | dtrask, its all local
| |
10:22 | one big box right now
| |
10:22 | <dtrask> aah
| |
10:22 | <monteslu> 8core amd64 16gb
| |
10:22 | ~70 workstations
| |
10:22 | <dtrask> monteslu: do you use enforced passwords or do the users pick their own?
| |
10:22 | <monteslu> though i have two of these boxes
| |
10:23 | dtrask, we assign them
| |
10:23 | <dtrask> monteslu: me too...so you have a list of users and pw's....you can simply put them in the IPCop box
| |
10:23 | <nubae> monteslu: wow thats a lot of ram...
| |
10:24 | <dtrask> and use the IPCop local auth....and do proxy auth that way
| |
10:24 | <monteslu> dtrask, cool stuff
| |
10:24 | nubae, yeah, and they were cheap too
| |
10:24 | about 4K a piece
| |
10:24 | with raid5
| |
10:24 | <dtrask> monteslu: one of the big benefits of assigned passwords....that and not having to reset forgotten ones all the time
| |
10:24 | staffencasa has joined #ltsp | |
10:25 | <monteslu> dtrask, also, when i moved from ubuntu back to fedora, i just copied the homes, rand a create user script, and had the lab teacher set the passwords
| |
10:26 | * alkisg just found out that 16 bit color depth in s3virge = 10 times faster than 24 bit.... | |
10:26 | <nubae> monteslu: I'm curious, whats your feeling buntu vs fedora?
| |
10:27 | mccann has joined #ltsp | |
10:27 | <nubae> for ltsp that is
| |
10:27 | <dtrask> monteslu: curious...not insinuating anything....why'd you switch?
| |
10:27 | <monteslu> nubae, well... I liked 7.04
| |
10:27 | but 8.04 gave me lots of problems
| |
10:27 | <dtrask> monteslu: you feel the ltsp implementation is better in Fedora?
| |
10:27 | <monteslu> i could never get things working right
| |
10:28 | <nubae> yeah that shouldnt have been a LTS
| |
10:28 | <monteslu> I've been a redhat user for 10 years
| |
10:28 | and i wanted to go back to fedora on ltsp
| |
10:28 | but k12ltsp died
| |
10:28 | <nubae> the first day was problems with the entire firefox debacle, and then just bug after bug
| |
10:28 | <monteslu> then as soon as i heard warrend picked it back up, i was all over it
| |
10:28 | <nubae> :-)
| |
10:29 | <monteslu> dtrask, i dont know if its better, but things seem to be working. And I'm just more comfortable with the system
| |
10:29 | i cant wait for a centos6 based version
| |
10:29 | I'll have that thing here for the next couple of years and hopefully it just works :)
| |
10:30 | there's also a factor of wanting to help out, with testing if nothing else...
| |
10:31 | lots of people using ubuntu, and it seemed like warren needed testers
| |
10:31 | <dtrask> monteslu: I hear ya'
| |
10:32 | <nubae> in the end it should all be quite similar
| |
10:32 | <dtrask> monteslu: I used to be a dyed in the wool Fedora guy....now I'm used to Ubuntu....but...now I use Resara which is kinda' Kubuntu based. I'm very encouraged and happy with what Warren and the Fedora guys are doing.
| |
10:32 | <nubae> I mean its an upstream project
| |
10:33 | <dtrask> looking forward to seeing Warren and the guys again this evening
| |
10:33 | :-)
| |
10:34 | dtrask is now known as dtrask-is-out-to | |
10:34 | dtrask-is-out-to is now known as dtrask | |
10:35 | dtrask is now known as dtrask-at-lunch | |
10:45 | tjikkun_work has quit IRC | |
10:45 | <ninkendo> so what exactly does ltsp-update-kernels do, aside from copy the kernels to the tftp root directory?
| |
10:45 | I'm trying to debug why LTSP isn't seeing my NIC on bootup... and when I chroot into the ltsp chroot, manually run update-initramfs, then copy the initramfs over, I get a certain set of problems
| |
10:46 | but when I do ltsp-update-kernels, I get an entirely different set of problsm
| |
10:46 | *problems
| |
10:47 | Blinny has quit IRC | |
10:47 | DonSilver has quit IRC | |
10:47 | <ninkendo> specifically, when I manually build my initramfs, it sees my NIC, but doesn't mount the NBD root properly. When I run ltsp-update-kernels, it doesn't see the NIC, and kernel panics trying to read /tmp/net-eth0.conf
| |
10:53 | <_UsUrPeR_> has anyone had any success installing an nvidia driver in chroot? There's a discrepancy between the kernels (server is running i686 while client is running i586) and it's causing installation problems with ./nvidia-installer
| |
11:03 | <johnny> you can make it think you're using whatever via setarch
| |
11:03 | setarch <whatever> chroot /path
| |
11:05 | setarch i586 chroot /path/to/ltsp
| |
11:05 | _UsUrPeR_, ?
| |
11:05 | <_UsUrPeR_> got it :P
| |
11:05 | had to uninstall the drivers from the server before I could install them in chroot
| |
11:06 | then back to the server afterwards
| |
11:06 | PITA
| |
11:06 | <johnny> that's what you get
| |
11:06 | for having nvidia
| |
11:06 | * johnny hate nvidia | |
11:06 | <_UsUrPeR_> are you an ATI man?
| |
11:06 | <johnny> or intel
| |
11:06 | anything with proper free drivers :)
| |
11:07 | are all your clients nvidia?
| |
11:07 | <_UsUrPeR_> My last ATI was a radeon 9800 pro, and it's support in hardy was atrocious. At least, while I used it.
| |
11:07 | <johnny> if so.. you'll get lucky
| |
11:07 | <_UsUrPeR_> no, this is for a customer
| |
11:08 | <johnny> whatever
| |
11:08 | <_UsUrPeR_> and all his clients use nvidia cards
| |
11:08 | <johnny> aha
| |
11:08 | then you're lucky
| |
11:08 | any other wonky combinations with non free drivers are likely to cause trouble
| |
11:08 | due to the way they ati and nvidia nonfree driveres handle libGL
| |
11:12 | <_UsUrPeR_> oh
| |
11:12 | yeah, I should probably thank my lucky stars for that one :)
| |
11:14 | <nubae> I just bought a ati hd3850, and its beautiful, though ubuntu didnt detect it properly at first
| |
11:14 | needs non-free drivers
| |
11:15 | but its well priced and does the job
| |
11:15 | <alkisg> johnny, what if someone has nvidia everywhere but 64bit OS on the server and 32bit on the chroot?
| |
11:17 | <johnny> i have no idea
| |
11:17 | i don't think it is a problem
| |
11:17 | <_UsUrPeR_> alkisg: I just worked through the kernel discrepancy thing
| |
11:17 | if that's what you're talking about...
| |
11:18 | <alkisg> _UsUrPeR_, something like that, yes
| |
11:18 | <johnny> sure.. but both of yours were 32bit
| |
11:18 | <_UsUrPeR_> no, server was 64-bit
| |
11:18 | <alkisg> Did it work OK?
| |
11:18 | <_UsUrPeR_> though it was a non-issue because the kernels were different anyway. Had to work past that too
| |
11:19 | alkisg: I got it to install on the chroot...
| |
11:19 | still working on getting client to boot
| |
11:19 | server's back up though
| |
11:19 | nice
| |
11:19 | got it
| |
11:19 | <alkisg> ...and use it to see if it does work ok... :)
| |
11:19 | <_UsUrPeR_> yes. works fine
| |
11:19 | <johnny> the weather is great here right now..
| |
11:19 | <alkisg> You still have to run some GL apps to see if it's ok
| |
11:19 | <_UsUrPeR_> there's a script I used to get it working. Quite useful. Do you need help right now? I have a minute
| |
11:19 | true
| |
11:19 | though...
| |
11:20 | <alkisg> _UsUrPeR_, no, not right now, just for the future, I'll need it sometime....
| |
11:21 | <_UsUrPeR_> ok, well one thing that I just figured out
| |
11:21 | the nvidia installer will put an xorg.conf file in
| |
11:22 | and under section "device" it calls for the driver "nvidia"
| |
11:22 | except for the driver's name is "nv"
| |
11:22 | I had to switch the driver name to nv
| |
11:22 | <alkisg> _UsUrPeR_, nv is the non-commercial driver
| |
11:22 | it's used even if you don't install anything
| |
11:22 | and it doesn't have much acceleration
| |
11:23 | <_UsUrPeR_> according to xorg's log, it's using the proper modules...
| |
11:23 | I'll need to check some 3d stuff for sure :)
| |
11:23 | * _UsUrPeR_ installs glxgears | |
11:23 | <Q-FUNK> http://q-funk.blogspot.com/2008/11/available-for-new-job.html
| |
11:24 | <jimcooncat> working on pxelinux.cfg
| |
11:26 | <_UsUrPeR_> q-funk: holy crap
| |
11:26 | <Q-FUNK> indeed
| |
11:26 | <nubae> Q-FUNK: I wish you luck... the education market is soon booming... just have to keep spirits high... in these time of economic crisis...
| |
11:26 | <_UsUrPeR_> congratulations? Are they in order? I can't really tell...
| |
11:28 | <jimcooncat> working on pxelinux.cfg, "C0A80025 the hexadecimal form of the client's IP address". How do I calculate this? example 10.0.0.202
| |
11:28 | <nubae> Q-FUNK: did u see shogunx project?
| |
11:28 | http://www.linux.com/feature/151362
| |
11:28 | might be opportunities there
| |
11:29 | <alkisg> jimcooncat, gnome-calculator for each byte? e.g. 10 => 0A
| |
11:30 | <jimcooncat> alkisg: i guess so. hoping for an easier way. maybe there's a calculator online, I'll look
| |
11:30 | <alkisg> jimcooncat, 0A0000CA
| |
11:30 | it only takes some seconds
| |
11:31 | <jimcooncat> thanks alkisg! Actually i found one, and your answer matches. So I guess I can trust them
| |
11:32 | oh, I see now how you did that
| |
11:32 | <alkisg> yeap, you only had to convert 202 to hex
| |
11:32 | Gadi has left #ltsp | |
11:36 | <nubae> anyone familiar with mahara eportfolio
| |
11:37 | dtrask-at-lunch is now known as dtrask | |
11:37 | <alkisg> jimcooncat, just curious, what do you want to do with that? (maybe IPAPPEND 3 will be useful for you)
| |
11:37 | <dtrask> back
| |
11:37 | jammcq_: when is Gadi coming up?
| |
11:40 | vagrantc has joined #ltsp | |
11:41 | <jimcooncat> alkisg: pxelinux will select a configuration based on MAC, a full or partial IP address, or a default. I want different ones for different computers.
| |
11:42 | alkisg: I originally had it based on MAC address, but I'm switching to IP address as I keep changing my mind which NIC to use.
| |
11:43 | <alkisg> jimcooncat, I see, ty
| |
11:44 | hanthana has quit IRC | |
11:44 | <jimcooncat> alkisg: what is IPAPPEND 3?
| |
11:45 | <alkisg> jimcooncat, if you needed a couple of configuration for *all* of your PCs, you could eg. put a 0A file (for 10.x.x.x.) and put IPAPPEND 3 to pass the ip back to the kernel, so no second dhcp request would be made... Never mind, that's why I use these files, your needs are different
| |
11:47 | <jimcooncat> oh! that would be helpful, my second dhcp request hangs half the time (on netboot installs, haven't tried ltsp with this config)
| |
11:48 | <alkisg> jimcooncat, you lose the dhcp assigned hostname this way, and lts.conf path will be "changed" to /var/lib/tftpboot
| |
11:48 | <vagrantc> sbalneav: looks like you've got another $PPID in commit log for all eternity :)
| |
11:49 | <sbalneav> Sigh
| |
11:49 | * sbalneav slaps self with a wet haddock | |
11:50 | <sbalneav> There's a REASON why I shouldn't be allowed to touch the trunks
| |
11:50 | However, we just got syslogging going too.
| |
11:50 | So that's cool
| |
11:50 | <vagrantc> gives us the opportunity to remenisc for old PIDs that had significant historical relevence :)
| |
11:50 | gm04030276 has joined #ltsp | |
11:51 | <vagrantc> stripped out X initialization from LDM!
| |
11:51 | * vagrantc remembers suggesting that in 2005 | |
11:51 | <sbalneav> Well, now it's done!
| |
11:51 | <Q-FUNK> :D
| |
11:52 | <sbalneav> We're logging to LOG_DAEMON, that seem fine with you, vagrant?
| |
11:52 | chrisinajar has joined #ltsp | |
11:53 | <vagrantc> sbalneav: what did you do to drop "kill -1 $PPID" reliably?
| |
11:53 | krishna has joined #ltsp | |
11:53 | <vagrantc> seemed like there were a fair number of cases where it seemed to work, but plenty that didn't.
| |
11:54 | * vagrantc wonders if we shouldn't bump ldm version to 2.1.0 sometime soonish | |
11:55 | mikkel has quit IRC | |
11:56 | <sbalneav> vagrantc: it's the < /dev/null > /dev/null that does it.
| |
11:56 | <vagrantc> nice!
| |
11:56 | <sbalneav> any problem if I push syslog mods to trunk?
| |
11:57 | * vagrantc doesn't know the changes | |
11:57 | <stgraber> sbalneav: the ltsp-localappsd change is already in trunk, so just upload the ldm one :)
| |
11:57 | <vagrantc> sbalneav: i thought there was already a different parameter for syslogging ... SYSLOG_HOST, maybe?
| |
11:58 | <sbalneav> Yeah, that's right, syslog_host's what we log to, but LOG_DAEMON's the facility (i.e. logs show up in daemon.log)
| |
11:59 | Gadi_eeepc has joined #ltsp | |
11:59 | <sbalneav> Gadi_eeepc!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
11:59 | <Gadi_eeepc> hehe
| |
11:59 | <sbalneav> How are you!
| |
11:59 | Long time no see!!!
| |
11:59 | <Gadi_eeepc> har har har
| |
11:59 | <vagrantc> sbalneav: so you talking about ldm logging to a syslog ?
| |
11:59 | <chrisinajar> har har har
| |
11:59 | <sbalneav> yeah
| |
11:59 | so LDM files pop out on the server
| |
12:00 | <vagrantc> nice.
| |
12:00 | <sbalneav> I'ts done and works. stgraber and I just tested it
| |
12:00 | OK if I push
| |
12:00 | ?
| |
12:00 | <stgraber> yeah, go ahead
| |
12:00 | <Gadi_eeepc> agreed
| |
12:00 | as does vagrantc
| |
12:00 | ;)
| |
12:01 | * vagrantc trusts that it'll be ok. | |
12:01 | <dtrask> Gadi_eeepc: Hey...you coming to Maine?
| |
12:01 | <vagrantc> sbalneav: go ahead and push ... i'm only keeping up with upstream in debian experimental anyways
| |
12:02 | <chrisinajar> isn't gadi already here?... I mean, i'm terrible with names, let alone screennames but...
| |
12:02 | <Gadi_eeepc> dtrask: pinch yourself - I
| |
12:02 | <dtrask> look for a guy in a yarmulke
| |
12:02 | <Gadi_eeepc> Im already here
| |
12:02 | :)
| |
12:02 | <chrisinajar> yay!
| |
12:02 | I'm right!
| |
12:03 | <dtrask> Gadi_eeepc: LOL! Oops....I'll see you tonight
| |
12:03 | :-)
| |
12:03 | * chrisinajar rejoices | |
12:03 | <dtrask> Gadi_eeepc: what kind of beer shall I bring...I owe you one
| |
12:03 | * Gadi_eeepc should have worn his basball cap | |
12:03 | <chrisinajar> dtrask: you should do something about your ETA :-P.... Arive earlier...
| |
12:03 | <Gadi_eeepc> ;)
| |
12:03 | dtrask: really?
| |
12:03 | cool
| |
12:04 | what did I do?
| |
12:04 | was I decent?
| |
12:04 | <dtrask> Gadi_eeepc: Are you ever decent?
| |
12:04 | <Gadi_eeepc> good point
| |
12:04 | <Q-FUNK> Gadecent_PC
| |
12:04 | <Gadi_eeepc> hehe
| |
12:05 | hey, Q-FUNK
| |
12:05 | <Q-FUNK> hey :)
| |
12:05 | <dtrask> chrisinajar: I'll drive "wicked" fast
| |
12:05 | <chrisinajar> awesome.
| |
12:05 | wicked awesome, even.
| |
12:05 | <Gadi_eeepc> tubular
| |
12:05 | whoops
| |
12:05 | <chrisinajar> eww.
| |
12:05 | <Gadi_eeepc> wrong coast
| |
12:05 | <chrisinajar> haha
| |
12:05 | east coast FTW!
| |
12:05 | <dtrask> chrisinajar: seriously though....what time did everyone finally turn in last night?
| |
12:05 | <chrisinajar> we're WICKED AWESOME.
| |
12:06 | dtrask: we weren't up too late... brendan and i departed for sleep around 11-ish...
| |
12:06 | <dtrask> chrisinajar: OMG! That's just wrong!
| |
12:06 | <chrisinajar> dtrask: last night wasn't really a party night :-P
| |
12:06 | <dtrask> chrisinajar: I'm banking on getting there and having at least 3 more hours before bedtime ;-)
| |
12:06 | <chrisinajar> dtrask: damn straight.
| |
12:07 | <dtrask> I'll bring midnight pizza
| |
12:07 | <chrisinajar> dtrask: i'm sure most of us wont argue with that :-P
| |
12:08 | brendan0powers has joined #ltsp | |
12:09 | <dtrask> chrisinajar: unless I bring only anchovie and sauerkraut pizza
| |
12:09 | chrisinajar: ask Chuck...do they have "lopstah" pizza?
| |
12:09 | <chrisinajar> dtrask: from what i hear, at that point in the night they wont care :-P
| |
12:10 | <dtrask> LOL!
| |
12:10 | If it's like last year....yer right
| |
12:10 | <brendan0powers> if they do, i hope it turns red when you cook it
| |
12:10 | <dtrask> Ragnar had a long happy night...and a not so happy morning
| |
12:11 | <brendan0powers> ha
| |
12:11 | when he walked in i almost blerted out "Hey I heard about you!..."
| |
12:11 | <dtrask> warren_: hey...did you ever end up getting those headsets last year? Remember...I think you ordered them at BTS
| |
12:11 | brendan0powers: LOL!
| |
12:11 | brendan0powers: ask him about hiking all over Montreal looking for a Scotch
| |
12:12 | <brendan0powers> mayby later
| |
12:12 | <dtrask> Actually Dave Johnston can tell you about it
| |
12:12 | <chrisinajar> hey i think i know who that one is!
| |
12:12 | Weew!
| |
12:12 | One of the 3 people who I remember their name :-P
| |
12:13 | <dtrask> Hey....how are you guys set on power strips...and extension cords. I have bunches of them from FOSSED...just let me know and I can bring some
| |
12:14 | <stgraber> power strips and extension cords would be great, networking is quite good :)
| |
12:14 | <dtrask> k....I'll bring some
| |
12:14 | got whole big Rubbermaid things...full of em
| |
12:28 | spectra has quit IRC | |
12:29 | <stgraber> dtrask: when do you plan arriving here ?
| |
12:29 | <chrisinajar> stgraber: if i remember correctly, he said ~midnight
| |
12:32 | jacd has joined #LTSP | |
12:36 | Basti_dash has joined #ltsp | |
12:46 | * vagrantc pities the poor translator who needs to translate "SIOCGIFCONF" | |
12:47 | <vagrantc> we should probably split out the untranslateable parts of our strings :)
| |
12:50 | krishna has quit IRC | |
12:52 | <dtrask> stgraber: Late this evening...prolly close to midnight
| |
12:53 | stgraber: nonetheless....I suspect you guys will all be wide awake at that time....
| |
12:53 | krishna has joined #ltsp | |
12:55 | <sbalneav> vagrantc: I don't: SIOCGIFCONF's the same in all languages :)
| |
12:55 | <stgraber> dtrask: yeah, that's very likely
| |
12:55 | <davidj> sbalneav: Not in Russian...
| |
12:55 | <vagrantc> heh
| |
12:56 | <warren> dtrask: yes
| |
12:59 | jacd has left #LTSP | |
12:59 | jacd has joined #LTSP | |
12:59 | jacd has left #LTSP | |
12:59 | jacd has joined #LTSP | |
12:59 | jacd has left #LTSP | |
12:59 | jacd has joined #LTSP | |
12:59 | jacd has left #LTSP | |
13:00 | jacd has joined #LTSP | |
13:00 | jacd has left #LTSP | |
13:02 | <sbalneav> vagrantc: OK, I've now down the void headers like pere wanted.
| |
13:02 | I'm gonna push another revision of ldm-trunk
| |
13:02 | OK by you?
| |
13:02 | <vagrantc> sbalneav: go ahead and push freely. :)
| |
13:03 | <sbalneav> danke
| |
13:03 | <vagrantc> no need to hold things up on my account :)
| |
13:04 | debian's in a hard freeze, so i'm still using a patched ldm 2.0.6, ltsp 5.1.10, and ltspfs 0.5.3
| |
13:04 | although it's basically ltspfs with all patches from 0.5.5 applied :)
| |
13:04 | so i'll have plenty of time to fix any bugs introduced now :)
| |
13:05 | <Q-FUNK> yikes
| |
13:06 | phantom has quit IRC | |
13:06 | <vagrantc> though i'd hope any development doesn't make backports to lenny insanely difficult
| |
13:07 | phantom has joined #ltsp | |
13:07 | alkisg1 has joined #ltsp | |
13:09 | <sbalneav> I don't think we're doing anything out of the ordinary.
| |
13:10 | the X in the screen script thing's fairly simple.
| |
13:10 | Does the syslog stuff cause you any difficulty?
| |
13:10 | <vagrantc> i doubt it
| |
13:11 | <_UsUrPeR_> alkisg: ok, glxgears works on the client, everything is wonderful
| |
13:11 | <vagrantc> main thing would be mandatory changes requiring the use of newer hal or xorg or glib or some such
| |
13:11 | <_UsUrPeR_> alkisg: if you need help, let me know.
| |
13:11 | <vagrantc> sbalneav: the changes thus far seem excellent as far as i'm concerned :)
| |
13:12 | <johnny> hmm..
| |
13:12 | <vagrantc> well, after disabling lots and lots of stuff, i managed to get an /opt/ltsp/kfreebsd-i386 built :)
| |
13:12 | <johnny> hmm.. i wonder if anybody is gonna do a ppa for newer ltsp packages
| |
13:13 | once you guys got all this going
| |
13:13 | <vagrantc> missing things on kfreebsd-i386: anything related to alsa or fuse, and getltscfg
| |
13:14 | <johnny> getltscfg shouldn't require much work
| |
13:14 | the rest.. hmm :)
| |
13:14 | oss time i guess
| |
13:14 | guess abstracting the alsa stuff will be easy
| |
13:14 | and fuse.. doesn't that exist?
| |
13:15 | <vagrantc> there's a freebsd port of fuse, but i haven't tried building it yet
| |
13:15 | there's libsalsa, which is an alsa -> oss compatibility layer
| |
13:16 | <Gadi_eeepc> lp:~gideon/ltsp/ldm-trunk-xinitrc
| |
13:16 | warren has quit IRC | |
13:16 | <Gadi_eeepc> lp:~gideon/ltsp/ltsp-trunk-xinitrc
| |
13:16 | <vagrantc> Gadi_eeepc: different from what sbalneav already merged?
| |
13:17 | <Gadi_eeepc> vagrantc: yeah - I just added an export of XINITRC=/usr/lib/ltsp/xinitrc
| |
13:18 | and /usr/lib/ltsp/xinitrc will call all scripts in /usr/lib/ltsp/xinitrc.d/
| |
13:18 | <vagrantc> ah. so it's not specific to ldm ?
| |
13:18 | <Gadi_eeepc> to (hopefully) implement the non-ldm specific init dir
| |
13:18 | we will still have both
| |
13:18 | <vagrantc> 921. By MTkyLjE2OC4wLjExMw <MTkyLjE2OC4wLjExMw@ltspdemo-appserv1> 19 minutes ago
| |
13:18 | <Gadi_eeepc> in case there are ldm-specific init scripts
| |
13:18 | <vagrantc> Add xinitrc script and modify rdesktop to call it
| |
13:18 | nice :)
| |
13:19 | <Gadi_eeepc> hehe
| |
13:19 | shandflm has quit IRC | |
13:19 | <Gadi_eeepc> I am on a guest accoount
| |
13:19 | on a thin client
| |
13:19 | :)
| |
13:19 | sorry
| |
13:19 | hmm...
| |
13:19 | well, if you want to commit it under your name
| |
13:19 | please do
| |
13:19 | :D
| |
13:21 | <Q-FUNK> whot? anonymous coward <guest@samba> is now a valid developer name? ;)
| |
13:22 | <Gadi_eeepc> hehe
| |
13:22 | stupid xandros has old bzr
| |
13:22 | and I am on my eeepc
| |
13:22 | so, now I am an eeepc thin client
| |
13:22 | <Q-FUNK> heh
| |
13:24 | alkisg has quit IRC | |
13:27 | davidj has quit IRC | |
13:27 | Basti_dash has quit IRC | |
13:29 | <dtrask> ok....I'm getting off here....and getting ready to get this stuff outta the way and head to SW Hahbah....see you guys later tonight....looking forward to hanging out with you all :-)
| |
13:30 | <cliebow_> cool!
| |
13:30 | <dtrask> cya l8tr
| |
13:31 | dtrask has left #ltsp | |
13:32 | etyack has joined #ltsp | |
13:32 | <jammcq_> etyack: hey
| |
13:32 | jammcq_ is now known as jammcq | |
13:33 | <etyack> jammcq_: hey i need to talk to ron asap
| |
13:33 | <jammcq> hey's gonna try to call you right now, but cell signal here is almost non-existant
| |
13:33 | troubles in paradise?
| |
13:34 | Lns has joined #ltsp | |
13:41 | Q-FUNK has quit IRC | |
13:53 | bobby_C has joined #ltsp | |
13:54 | etyack has quit IRC | |
13:56 | nubae has quit IRC | |
13:59 | <Lns> Anyone running rdesktop in LTSP with working sound to a windows server?
| |
13:59 | Basti_dash has joined #ltsp | |
14:03 | jarias has quit IRC | |
14:05 | GodFather has joined #ltsp | |
14:06 | <highvoltage> stgraber: I thought you were a motu since ages ago!
| |
14:08 | <Gadi_eeepc> Lns: from the server or from the client?
| |
14:08 | <Lns> Gadi_eeepc: from the client
| |
14:08 | <Gadi_eeepc> as in rdesktop screen script?
| |
14:08 | <Lns> Gadi_eeepc: I saw your ltsp.org wiki remark about all of that actually
| |
14:08 | Gadi_eeepc: no, from Gnome / tsclient
| |
14:09 | <Gadi_eeepc> ah, so thats from the server
| |
14:09 | ;)
| |
14:09 | padsp rdesktop -r sound:local:oss ...
| |
14:10 | CAN-o-SPAM has joined #ltsp | |
14:10 | <Gadi_eeepc> I think thats the only way for it to work because of some bug between pulse and alsa and alsa redirection
| |
14:10 | <Lns> Gadi_eeepc: so that will allow sound in both rdesktop session and linux?
| |
14:11 | <Gadi_eeepc> that will pipe the sound from your Linux desktop when rdesktop is launched from theere
| |
14:12 | so, if you open gnome-terminal in a thin client session and type that
| |
14:12 | you should get sound
| |
14:12 | <Lns> woot!
| |
14:12 | thanks buddy =)
| |
14:12 | <Gadi_eeepc> another way...
| |
14:12 | is to run rdesktop as a local app
| |
14:12 | * Lns hasn't touched localapps yet | |
14:13 | <Gadi_eeepc> you will get better sound as a local app
| |
14:13 | than the above methodd
| |
14:13 | but both work
| |
14:13 | <_UsUrPeR_> using Fedora 9, why does dhcpd sometimes use the dhcpd.conf in /etc/ltsp and at other times use the conf in /etc/? Is there a specific config file I can change to point towards one or the other?
| |
14:13 | <Lns> hmm.
| |
14:13 | _UsUrPeR_: isn't one of them a symlink?
| |
14:14 | Gadi_eeepc: thank you =) did you buy an eee pc ?
| |
14:14 | <Gadi_eeepc> _UsUrPeR_: if it works like Ubuntu, it honors /etc/ltsp first
| |
14:14 | then /etc
| |
14:14 | Lns: yeah, back at the end of the summer
| |
14:14 | <_UsUrPeR_> gadi: that's what I had understoof
| |
14:14 | <Gadi_eeepc> _UsUrPeR_: it is in the init.d script
| |
14:14 | <_UsUrPeR_> hmm
| |
14:14 | <Gadi_eeepc> for dhcp3-server
| |
14:14 | but, again, thats ubuntu
| |
14:15 | <_UsUrPeR_> gadi: yeah, a little different
| |
14:15 | <Gadi_eeepc> i would guess on redhat it is in /etc/default
| |
14:15 | or /etc/sysconfig
| |
14:15 | <_UsUrPeR_> though it appears it SHOULD be using /etc/ltsp due to an if then statement in /etc/init.d/dhcpd
| |
14:15 | <Gadi_eeepc> ah
| |
14:15 | there ya go
| |
14:16 | etyack has joined #ltsp | |
14:20 | krishna_ has joined #ltsp | |
14:20 | <alkisg1> Lns, you could also try the old way to run it locally, = install rdesktop into the chroot and put SCREEN_02=rdesktop.
| |
14:20 | puff has joined #ltsp | |
14:20 | <puff> Hi... anyone home?
| |
14:21 | Q-FUNK has joined #ltsp | |
14:21 | <puff> Anyone from the ubuntu thin-client project around? I was explaining to this user on #ubuntu what "thin client" means and why, etc, and I wondered if the ubuntu thin client might find it useful for me to write that up for their project page.
| |
14:22 | <Lns> alkisg1: that might work if this doesn't seem to
| |
14:22 | <johnny> hmm?
| |
14:22 | alkisg1 is now known as alkisg | |
14:22 | <johnny> puff, point them to the edubuntu handbook
| |
14:22 | maybe?
| |
14:22 | or the ltsp docs listed in the topic of this channel
| |
14:23 | Q-FUNK has quit IRC | |
14:23 | alekibango has quit IRC | |
14:23 | Q-FUNK has joined #ltsp | |
14:24 | <Lns> alkisg: gadi's command wrapper worked...i'll leave it at that since i don't want to spend time on windows compatibility anymore :p
| |
14:25 | <alkisg> Lns, yeah, I've tried both myself and now in my new installation I don't have any of them :)
| |
14:26 | Q-FUNK has quit IRC | |
14:27 | Q-FUNK has joined #ltsp | |
14:30 | <Lns> alkisg: haha, nice. On a sidenote, I think its hillarious that Windows terminal server REQUIRES a sound card in the SERVER for audio to work on client sessions.
| |
14:31 | * Lns wonders if anyone has built a fake audio hardware driver for windows terminal server to eliminate that necessity | |
14:33 | <alkisg> Lns, I've been a windows programmer for 17 years... There are *lots* of hillarious things on windows ;) There were some virtual sound drivers, but I haven't looked for one lately
| |
14:35 | <Lns> alkisg: nice to see you're moving (at least partly) to Linux/ltsp development =)
| |
14:36 | <alkisg> Lns, I've moved to Linux completely since Spring, and I'm not going back, except if specific programming-related jobs come up. About development, I'm still not sure which way to go (qt/gtk/wxwidgets/mono/web-only interface etc)
| |
14:37 | <Lns> alkisg: what kind of development will you be doing?
| |
14:37 | <alkisg> I think I'll start with X11 (not kidding) to get the base stuff
| |
14:37 | Gadi_eeepc has left #ltsp | |
14:37 | <Lns> alkisg: my vote would go for gtk :)
| |
14:38 | but just because i'm a pretty big gnome user/admin
| |
14:38 | <alkisg> Well, I could start with something easy, like an lts.conf manager, or something deep, like a multicast screen projecting application in X only (like a small italc that takes only a little bandwidth)
| |
14:39 | bobby_ has joined #ltsp | |
14:40 | <Lns> alkisg: don't know what it's worth but you might look at easy-ltsp, they made an attempt at an lts.conf gui, but it doesn't look much more like some text-boxes and [ OK ] buttons. :p
| |
14:40 | bobby_ is now known as Guest9677 | |
14:40 | staffencasa has quit IRC | |
14:42 | <johnny> ogra had something alkisg
| |
14:42 | that he hasn't finished
| |
14:42 | you could ask him
| |
14:42 | krishna has quit IRC | |
14:43 | <alkisg> johnny, I know, I wonder why noone maintains it, it looked like it could help newcomers to ltsp...
| |
14:44 | !seen ogra
| |
14:44 | <ltspbot> alkisg: ogra was last seen in #ltsp 1 day, 19 hours, 21 minutes, and 37 seconds ago: <ogra> anyway ... bed ...
| |
14:45 | <alkisg> He's probably exhausted after intrepid... :)
| |
14:45 | <Lns> or hung over ;)
| |
14:45 | <alkisg> hh
| |
14:45 | heh
| |
14:45 | <johnny> i really don't think you can cover over alot of the issues with ltsp
| |
14:45 | since they realted to modifying dhcp servers
| |
14:45 | that exist on the network elsewhere
| |
14:46 | as an example
| |
14:46 | <alkisg> johnny, so? Modifying scripts and running services it not a big deal, the difficult thing is to make it work on multiple distros... :)
| |
14:46 | <johnny> sure.. but it might be some cisco server
| |
14:46 | or some other non linux thing altogether
| |
14:46 | <alkisg> If the dhcp server is elsewhere, it's not a part of the program to modify it, but it may offer alternatives to the user
| |
14:46 | <Lns> johnny: what's wrong with a checkbox option "DHCP server elsewhere" (that grays out DHCP options in the gui) ?
| |
14:47 | <johnny> it shouldn't be ltsp specific
| |
14:47 | that's a general admin thing
| |
14:47 | <Lns> an ltsp manager shouldn't be ltsp specific ?
| |
14:47 | <alkisg> E.g. "according to your selections, you should specify these things in your router: .... / or in your windows dhcp server: ..."
| |
14:48 | <johnny> i'm saying very few of the options are ltsp specific
| |
14:48 | almost none
| |
14:48 | <alkisg> But even if it could take care of the more usual setups, it would make it easier for a lot of newcomers...
| |
14:48 | <johnny> the only thing ltsp specific is chroot generation
| |
14:48 | and lts.conf mods
| |
14:48 | <Lns> johnny: well most of the tie into the fact that you're running ltsp...so...
| |
14:48 | s/the/them
| |
14:48 | <johnny> they already exist tho
| |
14:48 | that aren't ltsp specific
| |
14:48 | <Lns> so?
| |
14:48 | <johnny> so.. use em :)
| |
14:49 | <Lns> if you're running them BECAUSE you're running LTSP...
| |
14:49 | oic, GUIs for dhcp/etc
| |
14:49 | <johnny> yes
| |
14:49 | <Lns> well..let there be two
| |
14:49 | <johnny> no
| |
14:49 | please don't
| |
14:49 | <Lns> that's the beauty of not vendor lockin ;)
| |
14:50 | * Lns has seen many horrendous attempts at DHCP GUI/WUI/etc | |
14:50 | <johnny> cuz then you'll have two crappy solutions.. instead of one good one
| |
14:50 | better to contribute to a set of generic tools that work well
| |
14:50 | <Lns> or one crappy one and one good one.. ?
| |
14:50 | <johnny> if you believe it..
| |
14:50 | lol
| |
14:51 | <Lns> what generic tools are there?
| |
14:51 | Webmin sucks
| |
14:51 | linuxconf sucks and is non-existant in most distros now a days
| |
14:51 | yast is distro specific, as is most other ones that are semi decent
| |
14:52 | <alkisg> A dhcpd.conf GUI tool should embed the dhcpd.conf parser to be proper... I don't suppose webmin or other similar tools do that
| |
14:53 | Guest9677 has quit IRC | |
14:53 | <Lns> I dunno..i still say screw GUIs all together for things like this, it's better to learn how it works anyway from the commandline. I don't see anything wrong with people learning the commandline these days.
| |
14:53 | Just because it used to be used before GUIs doesn't mean it's arcane or obsolete
| |
14:53 | i think we can all agree on that
| |
14:54 | <alkisg> Lns, I agree. A GUI in this case will only be useful for the most common cases, where the "admin" will never see the console
| |
14:55 | E.g. there are 2-3 possible setups for schools, the admin can select the one it fits him via a wizard, and never find out about dhcpd.conf or lts.conf etc
| |
14:55 | But if something different is needed, then console is a must
| |
14:56 | <Lns> alkisg: yep. I haven't run into a case yet that didn't require at least a LITTLE fiddling on the console to get set up and running with all requirements
| |
14:56 | A GUI can easily get crowded with every possible option (see Windows for that)
| |
14:56 | johnny has left #ltsp | |
14:57 | krishna__ has joined #ltsp | |
14:57 | <Lns> And it turns into a case of either "Which buttons do i click" or "what do I type", the latter usually taking much less time overall
| |
14:57 | krishna__ has quit IRC | |
14:58 | warren has joined #ltsp | |
14:58 | <alkisg> I was just thinking of something to let school "admins" (=computer teachers) switch to ltsp more easily... After a year of usage, they'll dump the GUI tool and get to the console anyway
| |
14:58 | pmatulis has quit IRC | |
15:00 | chrisinajar has quit IRC | |
15:00 | <stgraber> sbalneav: NumberOfPasswordPrompts
| |
15:01 | <Lns> alkisg: you know what i've heard from almost all of my teachers that would be nice to have a gui for
| |
15:01 | CAN-o-SPAM has quit IRC | |
15:01 | <Lns> is not the setup of LTSP but the day to day administration
| |
15:01 | <alkisg> Lns, that's a *big* project ;)
| |
15:02 | <Lns> well not if you look at LTSP cases
| |
15:02 | alkisg: what about a user interface for modifying user-level app config
| |
15:02 | like OOo
| |
15:02 | (removing their configuration to start over from scratch, say)
| |
15:02 | <alkisg> OK, that covers about 1/1000 of the day to day administration
| |
15:02 | bobby_C has quit IRC | |
15:02 | <alkisg> :P
| |
15:03 | bobby_C has joined #ltsp | |
15:03 | <Lns> lol..ok, and killing user processes
| |
15:03 | sanely, that is, with g-s-m (which is currently broken)
| |
15:03 | maybe simple integration of other GUIs in an LTSP centric gui
| |
15:03 | [ System Monitor ] [ User Configuration Editor ] [ Profile Manager ]
| |
15:04 | that would be pretty slick IMHO
| |
15:04 | and easy as pie
| |
15:04 | <alkisg> Slick, I agree, easy... hmm... I don't know, I'll have to look at the code!
| |
15:04 | <Lns> well if it's just a GUI that launches other GUIs...
| |
15:05 | <alkisg> OK, but I think this calls for a plugin system, scripting and more
| |
15:05 | <Lns> well sure
| |
15:05 | was just suggesting something easy to start with :)
| |
15:05 | that would help LTSP admins
| |
15:06 | <alkisg> Lns, I'll have some free time in the summer, I'll start something then. I'll have more experience also, to see what would be needed more. Thanks thoug!
| |
15:06 | though
| |
15:07 | <Lns> no, do it now! NOW! =p
| |
15:09 | <alkisg> Heh! This is the nice thing about developing apps without paying customers above your head... noone can tell you that! :)
| |
15:09 | <Lns> yep =)
| |
15:09 | but still..do it now. please? :p
| |
15:09 | j/k
| |
15:10 | <alkisg> Sure... if you send me a 10000+ lines patch for it :P
| |
15:12 | <Lns> sure
| |
15:12 | I'll just call my windows programmer :p
| |
15:15 | chrisinajar has joined #ltsp | |
15:17 | krishna_ has quit IRC | |
15:26 | johnny has joined #ltsp | |
15:30 | DawnLight has joined #ltsp | |
15:30 | jc2it has joined #ltsp | |
15:31 | <DawnLight> hello. in kubuntu intrepid i've gnome installed as well as kde and i'd like to have "ubuntu" dislayed on the login manager of ltsp clients instead of "kubuntu" and i'd like gnome to be the default desktop
| |
15:31 | is this something to do when building the client chroot?
| |
15:33 | <jc2it> Using Ubuntu 8.04 + LTSP 5. I would like to install wyse60 terminal emulation client to the thin client image. What step must I take. I am hitting a brick wall with sudo chroot /opt/ltsp/i386 and then using apt-get. apt-get works fine outside of chroot.
| |
15:35 | Basti_dash has quit IRC | |
15:36 | six2one has quit IRC | |
15:36 | <DawnLight> jc2it: can you paste your terminal session please?
| |
15:38 | <jc2it> sorry I am using a different computer to connect to IRC.
| |
15:38 | I think the repositories are not set up correctly in chroot
| |
15:39 | I can ping from chroot accross my network
| |
15:39 | i do not know how to test if I can connect to repositories.
| |
15:42 | <cliebow_> jc2it, almost seems like one nedds to mount a couple things..
| |
15:44 | <jc2it> I may have found my problem i am reading about it here... https://help.ubuntu.com/community/UbuntuLTSP/UpdatingChroot
| |
15:45 | <cliebow_> yeah that proc piece?
| |
15:48 | <alkisg> DawnLight, chroot /opt/ltsp/<arch> and update-alternatives --config ldm-theme
| |
15:48 | <DawnLight> thanks alkisg
| |
15:49 | <alkisg> DawnLight, or, while building the client, ltsp-build-client --dist ubuntu (I haven't tried this one)
| |
15:49 | <DawnLight> i thought about this
| |
15:57 | chrisinajar has quit IRC | |
15:59 | alkisg has left #ltsp | |
15:59 | chrisinajar has joined #ltsp | |
16:17 | jimcooncat has quit IRC | |
16:32 | <jc2it> I forgot apt-get update to build the repository database
| |
16:35 | <vagrantc> "ltsp-build-client --dist" is for selecting the release dist, not the distribution ... i.e. hardy,intrepid,lenny,sid vs. ubuntu,debian
| |
16:38 | cliebow has quit IRC | |
16:38 | cliebow has joined #ltsp | |
16:43 | cliebow has left #ltsp | |
16:50 | etyack has quit IRC | |
17:08 | <DawnLight> hey. is there some way to use intrepid's guest account thingy?
| |
17:16 | <vagrantc> LDM_GUESTLOGIN=true in lts.conf
| |
17:17 | davidj has joined #ltsp | |
17:17 | <davidj> sbalneav: Got a sec?
| |
17:17 | <Lns> vagrantc: does that function differently than the 8.04 Ubuntu lts.conf guest feature?
| |
17:17 | <johnny> no
| |
17:18 | there is a timeout
| |
17:18 | tho
| |
17:18 | i need to set that up :)
| |
17:18 | oh wait.. that didn't make intrepid did it?
| |
17:18 | might be nice to have a ppa for up to date ltsp stuff
| |
17:18 | <vagrantc> Lns: i didn't think it made it into hardy.
| |
17:18 | <Lns> I was thinking the other day it would be nice to have a single central "guest" account for any client (not a separate account for each client) with a volatile homedir (say in /tmp or something) that got wiped upon logout and can be used with simultaneous logins
| |
17:18 | oh maybe it is different
| |
17:18 | <johnny> don't think so..
| |
17:19 | it did make it into hardy vagrantc , just not into gutsy
| |
17:19 | <vagrantc> Lns: that's very different.
| |
17:20 | Lns: this is just basically a variation on LDM_AUTOLOGIN, except the greeter comes up and gives people with real accounts a chance to log in.
| |
17:20 | <Lns> vagrantc: i imagine so
| |
17:20 | vagrantc: gotcha
| |
17:20 | <johnny> LDM_ALLOW_GUEST it was
| |
17:20 | <Lns> yeah, those 2 features are in hardy
| |
17:20 | autologin and allow_guest
| |
17:20 | <johnny> it is not LDM_GUESTLOGIN tho
| |
17:20 | now*
| |
17:20 | <vagrantc> but the auto-wiped homedir sounds like a nice feature.
| |
17:20 | <Lns> vagrantc: and secure!
| |
17:21 | <johnny> why bother with just one account?
| |
17:21 | why not just create accounts on the fly?
| |
17:21 | <vagrantc> yeah.
| |
17:21 | <Lns> johnny: even better
| |
17:21 | <vagrantc> though this isn't really ltsp at that point.
| |
17:21 | would be useful to ltsp, but isn't something i'd think we could implement as part of ltsp.
| |
17:22 | <Lns> vagrantc: could or should?
| |
17:22 | <vagrantc> Lns: both, really. :)
| |
17:22 | <Lns> hehe
| |
17:22 | <DawnLight> its not working for me, the LDM_GUESTLOGIN
| |
17:22 | this is intrepid
| |
17:22 | it isn't doing anything different than when it is not set
| |
17:22 | <Lns> DawnLight: do you have the username/pass in there too?
| |
17:23 | <DawnLight> Lns: ahm... no. i didn't know i needed those. are there docs about this?
| |
17:23 | <Lns> DawnLight: LDM_USERNAME and LDM_PASSWORD
| |
17:23 | (hardy anyway, dunno if ibex is different)
| |
17:23 | <vagrantc> it shouldn't require setting LDM_USERNAME And LDM_PASSWORD
| |
17:23 | <Lns> oh ok, nevermind
| |
17:23 | <vagrantc> should default to the hostname of the thin client
| |
17:24 | <Lns> but you need the user acct in there regardless
| |
17:24 | <johnny> you still need accounts based on the hostname daduke_
| |
17:24 | err
| |
17:24 | DawnLight,
| |
17:24 | <vagrantc> you need the accounts on the server
| |
17:24 | <johnny> i use ssh key logins for that
| |
17:24 | * Lns has seen many people ask about the guest feature, not knowing about the actual acct creation part | |
17:24 | <DawnLight> what accounts? a 'guest account' that i've devised?
| |
17:24 | <johnny> all accounts
| |
17:24 | for each computer
| |
17:25 | <Lns> DawnLight: yes
| |
17:25 | <johnny> multiple guest accounts
| |
17:25 | <DawnLight> ah... shoot. is there a script for this?
| |
17:25 | <Lns> see in a 200+ TC network...having 200 guest accounts is kinda sloppy ;)
| |
17:25 | and 200 entries in lts.conf
| |
17:25 | <johnny> Lns, you can define a script in dhcp to automatically create them
| |
17:25 | skip the 200 entries lns
| |
17:25 | <Lns> johnny: hey that's true
| |
17:26 | <johnny> i create a user per computer
| |
17:26 | <warren> Ryan52: what is the function to switch focus to the login box? it seems broke again
| |
17:26 | <johnny> and put them all in a "terminals" group
| |
17:26 | and then iterate over that group, empty their home directories
| |
17:26 | put in the ssh keys
| |
17:26 | kill all proces that are hanging around
| |
17:26 | <Lns> jeez that sounds trivial..
| |
17:26 | <johnny> unmount .gvfs
| |
17:26 | <Ryan52> warren: hrm...in the fedora ldm or in normal ldm?
| |
17:26 | <johnny> it's only 20 lines
| |
17:26 | <warren> Ryan52: fedora
| |
17:27 | <johnny> Lns, you could more easily.. create users on the fly based on dhcp hosts
| |
17:27 | i know dnsmasq lets you run a script on lease creation
| |
17:27 | <Lns> johnny: that's what i'm thinking
| |
17:27 | <warren> Ryan52: could you adapt your patch as a conditional that can be merged into ldm, so I don't have to maintain a separate patch?
| |
17:28 | <Lns> well even if-up.d/
| |
17:28 | <johnny> i don't even do if up
| |
17:28 | <Ryan52> warren: ogra started a thingy so that it reads the layout from a file.
| |
17:28 | warren: dunno if he ever plans to finish that, though..
| |
17:28 | <johnny> i just keep what happens in the initrd
| |
17:28 | <DawnLight> does anyone have examples for guest accounts? and... how about using the guest account thing in the intrepid gdm?
| |
17:29 | <johnny> DawnLight, just create accounts for each computer based on thier hostname, and give them the hostname as the password
| |
17:29 | that's it
| |
17:29 | <Lns> johnny: and with SSH keys there's no passing of username/passwd to ltsp, should autologin.. ?
| |
17:29 | <johnny> yes
| |
17:29 | <Lns> as long as hostname var is used
| |
17:29 | <johnny> so just generate a nonsense password
| |
17:29 | <Ryan52> warren: the gtk_widget_grab_focus function is there..
| |
17:29 | <warren> Ryan52: I know, thus I could really use a conditional until then.
| |
17:29 | <Lns> BRILLIANT!
| |
17:29 | <johnny> ranpwd
| |
17:29 | <warren> Ryan52: http://cvs.fedoraproject.org/viewvc/devel/ldm/ldm-rearrange-stuff.patch?view=log
| |
17:29 | Ryan52: I might have screwed something up
| |
17:29 | * vagrantc uses disabled passwords | |
17:29 | <johnny> or that :)
| |
17:29 | <Lns> to log in?
| |
17:29 | oh w/ssh keys
| |
17:30 | <vagrantc> for accounts that should only login using ssh keys
| |
17:30 | <johnny> vagrantc, all password logins?
| |
17:30 | oh
| |
17:30 | * Ryan52 looks at this cvs thing in confusion | |
17:30 | <johnny> yeah.. probably an option in adduser or something
| |
17:30 | Lns, that is probably better
| |
17:30 | you do need keys in /root/.ssh in the chroot tho
| |
17:31 | <DawnLight> johnny: but they should be purged... i'm not sure how to do that
| |
17:31 | <vagrantc> that prevents the user from changing the accounts password
| |
17:31 | <johnny> purged when?
| |
17:31 | DawnLight, i purge mine nightly
| |
17:31 | <vagrantc> we purge accounts daily at freegeek.
| |
17:31 | <warren> Ryan52: this is the patch I apply against ldm-2.0.16
| |
17:31 | <Lns> johnny: you like purging your whole server at night don't you =p
| |
17:31 | <johnny> does anybody wanna see the nonsense script i have?
| |
17:32 | <Lns> sure
| |
17:32 | <DawnLight> good night and thank you!
| |
17:32 | DawnLight has quit IRC | |
17:32 | <johnny> http://localmomentum.net/~johnny/prep_terminals.sh
| |
17:34 | this line could be better
| |
17:34 | rm -rf /home/${i}/* /home/${i}/.*
| |
17:34 | i had another way.. but i lost it
| |
17:34 | something with find
| |
17:34 | bobby_C has quit IRC | |
17:35 | * Lns keeps forgetting about pkill | |
17:35 | <Ryan52> warren: I have a call to gtk_widget_grab_focus, and yours doesn't...I think.
| |
17:36 | <johnny> it was a quick and dirty thing
| |
17:36 | * Ryan52 wonders why interdiff is not working on these two diffs | |
17:36 | <johnny> when i realized that people were logging into my server remotely with bots
| |
17:36 | * Lns wonders if the infamous gnome-watchdog package looks for .gvfs mounts | |
17:36 | <johnny> i should check out gnome-watchdog, but i was hoping for something to either be fixed in gnome.. or something to happen in ssh
| |
17:36 | err
| |
17:36 | <warren> Ryan52: I might have screwed up and removed it by accident
| |
17:37 | <johnny> something to happen in ltsp directly
| |
17:37 | <vagrantc> building from ldm-trunk failed for me ... :(
| |
17:37 | <warren> Ryan52: where do I insert it?
| |
17:37 | <Ryan52> vagrantc: pfft. what do you expect when a bunch of hackers get together and write code all day :)
| |
17:38 | <ltsppbot> "vagrantc" pasted "ldm-trunk build error" (13 lines) at http://pastebot.ltsp.org/96
| |
17:38 | <Ryan52> warren: i have "gtk_widget_grab_focus(GTK_WIDGET(entry));" right after the gdk_window_set_cursor call.
| |
17:38 | <Lns> johnny: well it's nice because you can't guarantee that all apps will exit cleanly 100% of the time anyway
| |
17:38 | <vagrantc> was anything added that would require more build dependencies?
| |
17:38 | <Ryan52> warren: well...uhhhh...ummm...after the call that was taken out. so after the gtk_widget_show_all call.
| |
17:38 | <Lns> even if the real problem procs (bonobo-activation-whatever, gconfd-2) get fixed upon ltsp logout
| |
17:38 | <johnny> Lns, the problem is that ltsp doesn't traditionally handle such things due to screen type apps
| |
17:39 | <warren> Ryan52: i'm a little confusede
| |
17:39 | <johnny> but.. maybe there should be an option in lts.conf like killall_on_exit or whatever
| |
17:39 | to satsify 99% of users
| |
17:39 | anyways.. i gotta go for a bit Lns
| |
17:39 | bbl
| |
17:39 | <Lns> johnny: i agree.. gnome-watchdog has been great to me so far, 8 sites, no complaints, just praise that people stopped locking up upon subsequent logins
| |
17:39 | johnny: take care
| |
17:40 | <Ryan52> warren: one second.
| |
17:40 | warren: curl http://slexy.org/raw/s25uVSRvra | patch -p 1
| |
17:40 | <warren> i'll add that curl line to the .spec file for package builds
| |
17:41 | * Ryan52 really hopes warren was joking | |
17:41 | <warren> of course
| |
17:41 | <Ryan52> (sarcasm over the internet just doesn't work right :()
| |
17:42 | <vagrantc> who broke sarcasm!?
| |
17:44 | * Lns slowly walks away | |
17:45 | jacd has joined #LTSP | |
17:49 | <vagrantc> fgiraldeau, stgraber: did the tab patches require adding a build-dependency for you?
| |
17:49 | <warren> Ryan52: ok, scottie agreed to get this alt layout into as a conditional at least until oliver merges the new layout thing
| |
17:50 | Ryan52: what kind of conditional do you think is best? runtime or ifdef?
| |
17:50 | <Ryan52> messy messy messy! /me just knows ogra is gonna be pissed about that.
| |
17:50 | ifdef, please.
| |
17:52 | <stgraber> vagrantc: shouldn't no, I'll ask fgiraldeau when he goes back
| |
17:52 | *comes
| |
17:52 | <sbalneav> hello
| |
17:52 | sbalneav has quit IRC | |
17:53 | <vagrantc> i can build from ldm-trunk minus the tab patches
| |
17:53 | * Ryan52 wonders if there is a script that will merge diffs into code with #ifdefs around them...that would be soooooooo cool. then everybody could make such a big mess! :) | |
17:54 | <Ryan52> vagrantc: "grep -R name_of_thing_that_is_erroring /usr/include"...then just follow the logical chain of include backwards until you figure out which header should be included.
| |
17:54 | Q-FUNK has quit IRC | |
17:54 | <Ryan52> or google it...but googling isn't as fun :P
| |
17:55 | <stgraber> vagrantc: ok, having a look at it now.
| |
17:56 | <vagrantc> Ryan52: i don't think it's in my /usr/include/
| |
17:56 | Ryan52: turns up nothing ...
| |
17:56 | egrep -r GDK_TAB /usr/include/
| |
17:56 | <stgraber> vagrantc: hmm, ok I know what's broken, fgiraldeau told me to remove a "useless" include but I think it was actually quite useful
| |
17:57 | just let me fix that
| |
17:57 | * vagrantc suspects gdk-imlib11-dev | |
17:58 | <stgraber> ok, I have a fix, let me upload it
| |
17:58 | <Ryan52> vagrantc: it's GDK_Tab
| |
17:59 | <stgraber> +#include <gdk/gdkkeysyms.h>
| |
17:59 | pushed
| |
17:59 | * Ryan52 notes that that's what the grep -R turned up to do too :) | |
17:59 | <vagrantc> what package is gdkkeysyms.h from?
| |
18:00 | <Ryan52> vagrantc: libgtk2.0-dev
| |
18:00 | vagrantc: full path is /usr/include/gtk-2.0/gdk/gdkkeysyms.h
| |
18:00 | <stgraber> doh, Ryan52 was faster
| |
18:01 | vagrantc: it's already a build-dep
| |
18:01 | <vagrantc> yup, just a missing #include, then
| |
18:01 | stgraber: although ubuntu and debian have really strayed in the various debian dirs.
| |
18:02 | stgraber: i don't think we've seen full sync of ldm or ltsp since pre-gutsy
| |
18:02 | <stgraber> right, I hope we can get most of the packaging synced between both, there isn't that much distro-specifc stuff in there
| |
18:02 | <vagrantc> yeah.
| |
18:03 | something that would go a long way to making it easier would be to split out all the ubuntu themes into a separate package
| |
18:03 | i've been meaning to work on a debian theme package that you could use as an example, but artwork isn't my strong point and it requires some tweaking of the logos
| |
18:05 | <stgraber> right, then it's just a matter of a depend or even a distro-specific script in ltsp-build-clients for the artwork
| |
18:05 | other distro-specific stuff include that udev thing but you got that fix IIRC ?
| |
18:05 | <Ryan52> vagrantc: just use gdm's theme. ldm needs to be exactly like gdm in every single way! :)
| |
18:06 | davidj has quit IRC | |
18:06 | <vagrantc> stgraber: yeah, you could sync 0.5.5-1 from debian experimental into jaunty
| |
18:06 | jacd has left #LTSP | |
18:06 | jacd has joined #LTSP | |
18:07 | <vagrantc> stgraber: ltspfs, that is
| |
18:10 | <stgraber> yeah, this one will be a simple sync, others will take a bit longer to make sure we don't have some unknown ubuntu-specific changes forgoten in the process
| |
18:10 | <vagrantc> stgraber: what's your timeline for syncing with debian?
| |
18:11 | <stgraber> automatically sync deadline is end of december, then I can still manually sync
| |
18:11 | until the feature freeze
| |
18:13 | Q-FUNK has joined #ltsp | |
18:21 | <vagrantc> the /dev/null trick didn't fix the problem with "kill -1 $PPID"
| |
18:29 | * vagrantc is very happy to see ldm started from xinit | |
18:29 | brendan0powers has quit IRC | |
18:31 | <Lns> vagrantc: woah, does that mean we'll have better X integration w/ldm? like numlockx stuff working better in ldm? or am i missing something
| |
18:35 | <vagrantc> and ldm logging to syslog is the best!
| |
18:35 | Lns: we already got that
| |
18:35 | Lns: with the ldm/rc.d/I* scripts
| |
18:36 | <Lns> vagrantc: woo on syslog! how about rebooting of the server? Is there a possibility of clients/ldm/etc handling that better (i.e. not having to reboot themselves after server reboot) ?
| |
18:36 | (at least in nbd style setups anyway)
| |
18:37 | <vagrantc> nothing changed there, really.
| |
18:38 | <stgraber> Lns: that may already work in Ubuntu with nbd using -persist, as long as you keep the same nbd root and the server never rejects the request (DROP is fine, REJECT makes nbd not to try again)
| |
18:40 | <Lns> stgraber: well requests would probably be DROPed and not REJECTed since it's rebooting i'd assume
| |
18:40 | I'll have to try the -persist thing..it's a pain having to reboot a bunch of TCs whenever bouncing the server
| |
18:41 | specially with all these ubu kernel update storms =p
| |
18:41 | <Ryan52> Lns likes iptables a little bit too much :)
| |
18:41 | cyberorg has quit IRC | |
18:41 | <Lns> Ryan52: guess i had the wrong context eh?
| |
18:42 | <Ryan52> oh, stgraber started it.
| |
18:42 | anyway, it would be REJECTED, I guess. connection refused before the server thing starts.
| |
18:43 | before the nbd server starts.
| |
18:43 | but after the server gets it's ip and set up networking.
| |
18:44 | * vagrantc wonders if xatomwait is holding open the X connection | |
18:44 | laprag has joined #ltsp | |
18:44 | <vagrantc> i seem to get an indefinitely stalled logout on my second and subsequent connections, but the first one logs out properly.
| |
18:45 | <Ryan52> oh -_-
| |
18:45 | so, then if it's a bug in ldm/ltsp/whatever I completely destroyed my test environment for no reason.
| |
18:46 | <stgraber> bah, just pull. Scotty did quite a lot of bugfix in LDM :)
| |
18:47 | <vagrantc> stgraber: i pulled the fixes, and they're what broke it :P
| |
18:47 | oh, i see some more
| |
18:51 | <stgraber> yeah, the hackfest is probably not the best time to plan a new stable update :)
| |
18:52 | * vagrantc hasn't done a stable update in years | |
18:52 | <vagrantc> i'm just trying to keep up :)
| |
18:52 | sbalneav has joined #ltsp | |
18:53 | <sbalneav> Hey hey
| |
18:53 | * vagrantc waves to sbalneav | |
18:53 | <sbalneav> hey vagrant
| |
18:53 | * vagrantc re-builds ldm with sbalneav's latest | |
18:53 | <sbalneav> so, We've now got a framework for signal handling
| |
18:55 | So, we can now handle ssh dying into the middle of the chat
| |
18:57 | <vagrantc> sbalneav: i think the fix that removed "kill -1 $PPID" doesn't work in all cases
| |
18:57 | it seems to fail for me on the second and subsequent logins ... just hangs
| |
18:57 | works fine with newer openssh, but with older versions it breaks.
| |
18:58 | <sbalneav> I'm running it at work with hardy
| |
18:58 | <Lns> have a good weekend all
| |
18:58 | * Lns waves | |
18:58 | Lns has quit IRC | |
18:58 | <sbalneav> Is that an "old" or a "new" ssh?
| |
19:00 | vagrantc: So, it's hanging on logout for you?
| |
19:01 | <vagrantc> sbalneav: yes, with a 4.7 version of openssh-server
| |
19:01 | sbalneav: works fine with 5.1
| |
19:02 | <sbalneav> I've got 4.7 here on Hardy, and it's working
| |
19:02 | <vagrantc> using icewm here
| |
19:02 | <sbalneav> What happens if you use gnome?
| |
19:02 | <vagrantc> ubuntu might have backported some patches to 4.7, too
| |
19:03 | sbalneav: good question!
| |
19:03 | * vagrantc tries | |
19:04 | <vagrantc> gueslogin seems broken with a changed session ...
| |
19:04 | <sbalneav> According to the ssh manpage, redirecting sdtin & stdout is supposed to cure the problem that we were having, and certainly at least on Ubuntu, it does fix it.
| |
19:04 | <vagrantc> guestlogin is just broken...
| |
19:05 | <sbalneav> What do we need to add to lts.conf for guest login?
| |
19:06 | <vagrantc> LDM_GUESTLOGIN=true
| |
19:06 | * vagrantc logs in manually with gnome | |
19:06 | <vagrantc> gnome is hanging also.
| |
19:07 | <sbalneav> Hm.
| |
19:08 | Q-FUNK has quit IRC | |
19:08 | <sbalneav> ok, we'll have a look at the guestlogin function.
| |
19:08 | We can see it not working here.
| |
19:09 | <vagrantc> that it logs to syslog makes it so much easier to debug!
| |
19:12 | <stgraber> vagrantc: so if we have: LDM_SERVER, LDM_USERNAME, LDM_PASSWORD and LDM_GUESTLOGIN set it should work right ?
| |
19:12 | <vagrantc> stgraber: you shouldn't need anything but LDM_GUESTLOGIN=true
| |
19:13 | <Ryan52> stgraber: it defaults to the hostname of the client for username and password.
| |
19:13 | <vagrantc> and an account on the server named the same as the hostname
| |
19:13 | <sbalneav> vagrantc: fixing guestlogin
| |
19:14 | GodFather has quit IRC | |
19:15 | <sbalneav> vagrantc: pushed fixed, try guestlogin now
| |
19:20 | <vagrantc> sbalneav: haven't tested, but i think you'll need to special-case get_userid as well ..
| |
19:20 | oh, maybe not.
| |
19:20 | <sbalneav> We got logged in with that pusg.
| |
19:21 | push
| |
19:21 | So, I'm interested as to why the < /dev/null > /dev/null doesn't work.
| |
19:23 | <jammcq> Scotty !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
19:26 | <sbalneav> jammcq!!!!!!
| |
19:29 | gm04030276 has quit IRC | |
19:32 | <vagrantc> sbalneav: guestlogin fix worked.
| |
19:34 | <sbalneav> ok, well, THAT's good :)
| |
19:34 | We need to figure out the kill -1 thing though
| |
19:40 | <vagrantc> sbalneav: i think it's the same problem with the last time we tried it ... something keeps the ssh session going such that it never calls "ssh -O exit"
| |
19:40 | <stgraber> sbalneav: http://ubuntu.pastebin.com/f6132bd03
| |
19:40 | <vagrantc> seems like LDM_LOGIN_TIMEOUT is also broken
| |
19:42 | LDM_AUTOLOGIN works
| |
19:45 | maybe i'm wrong on LDM_LOGIN_TIMEOUT ... seems to be working with LDM_GUESTLOGIN=true
| |
19:46 | i so frequently have LDM_GUESTLOGIN enabled when testing, i don't know if that's a change in behavior
| |
19:47 | but i *thought* it used to work without guestlogin
| |
19:47 | nevermind. it's working just fine.
| |
19:47 | oh, it is broken...
| |
19:48 | the timeout counts down, but does nothing.
| |
19:48 | <Ryan52> I'm pretty sure it previously worked on it's own...I do most of my testing with everything off but the feature I'm testing.
| |
19:49 | laprag has quit IRC | |
19:50 | <vagrantc> although, the use case where you'd want it to log in after a certain time but not provide them a button to handle the login seems ... rare.
| |
19:51 | <Ryan52> true
| |
19:52 | mccann has quit IRC | |
19:53 | * vagrantc never understood how the timeout worked ... | |
19:53 | <Ryan52> how it worked from a users point of view or a coders point of view?
| |
19:54 | <vagrantc> the code
| |
19:55 | i see the update_timeout function, and where it's called ... but i don't really understand what g_timeout_add does ...
| |
19:56 | * vagrantc wonders why scopy won't die. | |
19:57 | <Ryan52> so at the main part of ldm we hand over contrl to glib, and then everything is event based. g_timeout_add makes glib run the function you give it every x milliseconds (I think it is).
| |
19:58 | <sbalneav> ldm? or gtkgreet?
| |
19:58 | <vagrantc> gtkgreet
| |
19:58 | <sbalneav> right.
| |
19:59 | So, when logout's hung, have you run an strace on the session, to see what it's stuck on?
| |
20:00 | <vagrantc> sbalneav: where would i run it from?
| |
20:00 | <sbalneav> Prolly on the server
| |
20:00 | Here's a test
| |
20:00 | one sec
| |
20:01 | <vagrantc> the last time we tried it, killing the hanging dbus-launch or dbus-daemon would cause it to continue ...
| |
20:01 | but with this patch, even killing all the user's processes still leaves it hung
| |
20:03 | <sbalneav> vagrantc: Does your icewm session script redirect stderr to .xsession-errors?
| |
20:06 | <vagrantc> sbalneav: no idea ...
| |
20:08 | <sbalneav> ummm, welll......
| |
20:08 | could you...
| |
20:08 | maybe....
| |
20:08 | check?
| |
20:08 | On the gnome one, it redirects stderr to .xsession-errors
| |
20:08 | I'm wondering if stderr's hanging you up
| |
20:09 | <vagrantc> i'm looking, it's just ... looking for a needle in a needle pit
| |
20:09 | <Ryan52> I don't think it's the window manager that does it.
| |
20:09 | I think it's the thing that calls the window manager.
| |
20:09 | * vagrantc suspects dbus-lauch --exit-with-session x-session-manager | |
20:12 | * vagrantc needs to eat dinner | |
20:12 | CAN-o-SPAM has joined #ltsp | |
20:12 | sbalneav has quit IRC | |
20:12 | sbalneav has joined #ltsp | |
20:12 | <sbalneav> ergh
| |
20:12 | sorry
| |
20:13 | vagrantc: can I get you to try a local patch?
| |
20:14 | at about line 580
| |
20:15 | > /dev/null < /dev/null
| |
20:15 | sorry < >
| |
20:15 | can you do < /dev/null > /dev/null 2> /dev/null
| |
20:15 | that should send stderr to /dev/null as well
| |
20:18 | Am I here?
| |
20:18 | Is there anybody... out there
| |
20:18 | <jammcq> not me
| |
20:18 | <sbalneav> ok
| |
20:18 | You cannot reach me now
| |
20:18 | no matter how you try
| |
20:19 | goodbye cruel world it's over
| |
20:19 | walk on by
| |
20:19 | Sitting in my bunker
| |
20:19 | here inside my wall
| |
20:19 | waiting for the worms
| |
20:19 | to come to call
| |
20:20 | <stgraber> sbalneav: gtk_misc_set_alignment((GtkMisc *) entry, 1, 0.5);
| |
20:22 | gm04030276 has joined #ltsp | |
20:23 | gm04030276 has quit IRC | |
20:23 | <sbalneav> vagrantc: Warren confirms that exiting works for him, but I don't know his ssh version.
| |
20:24 | vagrantc: here's where I got the idear from:
| |
20:24 | http://www.openssh.org/faq.html#3.10
| |
20:34 | <Ahmuck> if a mobo does not have sound, ie server mobo, but the client has sound, will the client still have sound?
| |
20:34 | <johnny> yes
| |
20:35 | unless it's windows ..
| |
20:35 | then the server must have sound
| |
20:36 | <Ahmuck> *confusion*. windows programs or windows in a virtual machine?
| |
20:36 | isn't ltsp a linux server, there isn't a ltsp windows server is there?
| |
20:36 | <Ryan52> Ahmuck: that was just a windows sucks comment, I think :)
| |
20:37 | <Ahmuck> ok. lol
| |
20:37 | <johnny> Ahmuck, some people run a windows machine in combination with an ltsp machine
| |
20:37 | so they can use rdesktop to get linux thin clients that just boot into a windows server
| |
20:37 | <Ahmuck> next question. servers are configured with non-ecc ecc and fully buffered ram. i assume that fully buffered ram is best
| |
20:37 | <Ryan52> Ahmuck: ya, you can use rdesktop from thinclients.
| |
20:38 | <Ahmuck> rdesktop to boot into a windows server?!
| |
20:38 | <johnny> yeah.. some people do that
| |
20:38 | <Ahmuck> windows server --> rdesktop --> linux client --> windows interface ?
| |
20:38 | <johnny> yes
| |
20:38 | <Ahmuck> how odd
| |
20:38 | <johnny> well windows doesn't have netbootable thin clients
| |
20:39 | <Ryan52> Ahmuck: it's just a linux command...ltsp can run anything on bootup.
| |
20:39 | <Ahmuck> one could put a basic sound card into a windows server via pci i suppose and it would work
| |
20:39 | <johnny> yes
| |
20:39 | but as Ryan52 said.. windows sucks
| |
20:39 | <Ahmuck> i doubt you'd get full capable fiber sound via ltsp anyway
| |
20:39 | <johnny> you can't get sound through rdesktop without a soundcard in the windows server
| |
20:39 | <Ahmuck> ok, well, that's not my problem, as i'm ubuntu
| |
20:40 | <johnny> as far as ram goes..
| |
20:40 | people just get the best they can afford
| |
20:40 | <Ahmuck> i've noticed a lot of sysadmin menu entries in kde, and gnome. i assume there is a way to remove them
| |
20:40 | i can afford fully buffered ecc ram.
| |
20:40 | <johnny> i'm using a standard desktop relatively
| |
20:40 | <Ahmuck> which is why i asked
| |
20:40 | <johnny> with 2 GB ram
| |
20:41 | whatever i got cheap off newegg :)
| |
20:41 | hehe
| |
20:41 | as my server
| |
20:41 | warren has quit IRC | |
20:41 | <Ahmuck> :) the cost difference is about 100.00
| |
20:41 | <johnny> well.. whatever you like
| |
20:41 | <vagrantc> sbalneav: will try the 2> /dev/null patch
| |
20:41 | <Ahmuck> anywho, so, can one build seperate login screens and desktops for different users and cut out the admin meny options
| |
20:42 | <vagrantc> sbalneav: but then i'll turn in for the night.
| |
20:46 | CAN-o-SPAM has quit IRC | |
20:46 | johnny has left #ltsp | |
20:46 | <vagrantc> sbalneav: i've had it working ever since switching to openssh-server 5.1 ... but when i downgrade to a 4.7 version it fails.
| |
20:47 | sbalneav: also, does your patch give a chance for rc.d/K* scripts to trigger?
| |
20:49 | johnny has joined #ltsp | |
20:54 | johnny has left #ltsp | |
20:55 | <vagrantc> sbalneav: no luck on redirecting 2> /dev/null
| |
20:55 | johnny has joined #ltsp | |
20:55 | <Ahmuck> i guess the question i was asking, is there a performance hit for using fully buffered ram?
| |
20:56 | johnny is now known as Guest27817 | |
20:56 | <Guest27817> no idea
| |
20:56 | never had the opportuntity to even thinking about paying so much
| |
20:57 | check the benchmarks in a non partisan web site
| |
20:57 | this is not ltsp specifc
| |
20:57 | so any ole benchmarks.. even for win should be telling
| |
20:58 | jammcq has quit IRC | |
20:59 | <loather-work> some systems require FBDs
| |
20:59 | for instance, every current xeon system
| |
20:59 | <Ahmuck> Guest27817: k, i found out, that non-ecc is faster. is ltsp more stable using ecc ram vs non-ecc. i don't want a server that is going to be crashing on non-ecc
| |
21:00 | loather-work: thx. that's what i'm building
| |
21:00 | that pretty much makes the choice for me
| |
21:01 | Guest27817 is now known as johnny | |
21:02 | <johnny> Ahmuck, people rarely talk about such issues here
| |
21:02 | <loather-work> the performance difference is going to be minimal at best anyways.
| |
21:02 | FBDs are very, very fast.
| |
21:03 | <Ahmuck> loather-work: thx
| |
21:05 | johnny: well, i was asking because i'm looking at the fastes ltsp server i can get. i'm concerned about speed. gigabyte ethernet, etc. i did have a client (450mhz, 256mb ram) that would not boot. since i have many of these, i've been concerned i'm going to have to find better client machines but won't know till i do some more testing
| |
21:07 | rslrdx has joined #ltsp | |
21:07 | <Ahmuck> johnny: btw, i'm in ##hardware asking similar questions. the question i asked here i felt was more specific to the performance of ltsp on hardware
| |
21:08 | warren has joined #ltsp | |
21:17 | <johnny> that's plenty of ram
| |
21:17 | people run with less proc..
| |
21:17 | if they didn't boot.. it is another problem
| |
21:23 | <Ahmuck> ah, ok. i'm going to tackle the boot problem later.
| |
21:28 | <sbalneav> Well, THAT was a first-rate amount of hacking today.
| |
21:29 | <vagrantc> the xinit and syslog stuff is excellent! :)
| |
21:29 | <Ahmuck> in a ltsp environment, i assume one can "limit" drive space for each user
| |
21:30 | how much drive space does a user normally use?
| |
21:30 | <vagrantc> Ahmuck: not in an ltsp environment any more than a "normal" system.
| |
21:30 | <Ahmuck> k, guess it varies on the user
| |
21:31 | it's not uncommon for users to "backup" cd's or purchase mp3's for their mp3 player. i assume that usb ports work on the client and allow the transfer of data between devices?
| |
21:32 | <johnny> usb works.. but not cd writers
| |
21:32 | <Ahmuck> johnny: really?
| |
21:32 | would they work on a "fat" client?
| |
21:32 | <johnny> yes
| |
21:32 | i think so
| |
21:32 | maybe you can make the cd burning app run local
| |
21:41 | cliebow_ has quit IRC | |
21:45 | petre has quit IRC | |
21:57 | helpme has joined #ltsp | |
22:02 | CaScAdE^FarAway has joined #ltsp | |
22:03 | <helpme> may i join?
| |
22:03 | <vagrantc> !question
| |
22:03 | <ltspbot> vagrantc: "question" is if you have a question about ltsp, please go ahead and ask it, and people will respond if they can. please also mention the linux distro and release you're using. :)
| |
22:04 | cyberorg has joined #ltsp | |
22:04 | <helpme> ok, i have problem with ubuntu 8.04 and ltsp 5. ACPI: DMI BIOS year==0, assuming ACPI-capable machine
| |
22:05 | seems like my kernel needs modification in dhcp.conf itry using option option-128 code 128 = string;
| |
22:05 | option option-129 code 129 = text;
| |
22:05 | option option-128 e4:45:74:68:00:00;
| |
22:05 | option option-129 "acpi=off";
| |
22:05 | but not reading or work at all, any suggestion??
| |
22:06 | CaScAdE^1arAway has quit IRC | |
22:19 | alekibango has joined #ltsp | |
22:22 | <johnny> no
| |
22:23 | you should set it in /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default
| |
22:23 | i don't know anybody who has ever done it via option 129
| |
22:24 | or specifically /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/macprefix
| |
22:24 | without the colons
| |
22:25 | <helpme> because i am using etherboot, i am using this way http://www.ltsp.org/twiki/bin/view/Ltsp/BootingFromHarddrive
| |
22:26 | if i correct your way is for PXE or booting from lancard ony??
| |
22:26 | <johnny> i don't know anything about etherboot
| |
22:26 | yes.. pxe
| |
22:26 | that's all i use
| |
22:26 | i use floppy from rom-o-matic on machines that can't netboot
| |
22:27 | none of my machines have hard drives
| |
22:27 | <helpme> i think may be same with me, just different media. me from harddisk
| |
22:27 | <johnny> yeah.. well then set it to use pc instead
| |
22:27 | err pxe instead
| |
22:27 | <helpme> are you using ltsp5 and ubuntu 8.04?
| |
22:27 | <johnny> uggh
| |
22:27 | maybe i'm just confusing myself.. too tired
| |
22:27 | 8.10 now
| |
22:28 | but 8.04 before..
| |
22:28 | <helpme> how you handle acpi?
| |
22:28 | <johnny> i leave acpi on
| |
22:28 | acpi is good
| |
22:28 | all my machines work fine with acpi
| |
22:28 | <helpme> i mean, i have old machine problem with acpi, so i must turn off acpi
| |
22:28 | <johnny> :(
| |
22:29 | <helpme> are you using boot filename "/ltsp/i386/nbi.img" ??
| |
22:29 | or filename "/ltsp/i386/pxelinux.0"
| |
22:30 | cmiiw /var/lib/tftpboot/ltsp/i386/pxelinux.cfg/default is for "PXEClient"??
| |
22:33 | chrisinajar has quit IRC | |
22:35 | wpgmb has joined #ltsp | |
22:37 | <wpgmb> Can anyone help me with Bug #269188? (Extreme slowness and FF is already running)? From what I can tell, a patch is available, but I have no idea how to implement it?
| |
22:38 | Egyptian[Home]1 has joined #ltsp | |
22:46 | Egyptian[Home] has quit IRC | |
22:47 | <helpme> anybody using etherboot?
| |
22:48 | <wpgmb> helpme: yes. What's the problem?
| |
22:52 | <helpme> the problem with acpi on old machine
| |
22:52 | i want to disable it using dhcp option option-129 "acpi=off"; but it never read that. how?
| |
22:55 | wpgmb: how you set your dhcp.conf? i mean for making option option-129 readable?
| |
22:56 | <wpgmb> helpme: I don't have any issues with it, so don't bother with it. Are you booting from a floppy and you're having issues?
| |
22:57 | <helpme> i am booting from hard drive
| |
22:58 | i have that issue because my client is old machine, which not support acpi
| |
23:00 | <wpgmb> helpme: I'm missing something... you're booting from a harddrive, using etherboot?
| |
23:05 | <johnny> what are the specs of this old client?
| |
23:13 | wpgmb has quit IRC | |
23:44 | Ahmuck has quit IRC | |
23:49 | <helpme> tq
| |
23:49 | bye
| |
23:50 | helpme has left #ltsp | |