IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 4 January 2010 (all times are UTC)
| 01:14 | johnny1 has left #ltsp | |
| 01:14 | johnny has joined #ltsp | |
| 01:15 | johnny is now known as Guest66469 | |
| 01:18 | Sarten-X has quit IRC | |
| 01:20 | Sarten-X has joined #ltsp | |
| 01:30 | Selveste1 has quit IRC | |
| 01:46 | alkisg has joined #ltsp | |
| 02:07 | tstafford_ has quit IRC | |
| 02:13 | Selveste1 has joined #ltsp | |
| 02:21 | Guest66469 has left #ltsp | |
| 02:22 | johnny has joined #ltsp | |
| 02:23 | johnny is now known as Guest66109 | |
| 02:59 | Guest66109 has left #ltsp | |
| 03:04 | johnny has joined #ltsp | |
| 03:04 | johnny is now known as Guest61897 | |
| 03:05 | Selveste1 has quit IRC | |
| 03:05 | vagrantc has quit IRC | |
| 03:09 | Guest61897 has left #ltsp | |
| 03:10 | Guest61897 has joined #ltsp | |
| 03:11 | <Guest61897> hjh
| |
| 03:12 | Selveste1 has joined #ltsp | |
| 03:12 | mic_p has joined #ltsp | |
| 03:13 | <mic_p> Hi all. A question about usb mouse / keyboard that don't want to work on a terminal with X but work on console. Someone?
| |
| 03:13 | Guest61897 has left #ltsp | |
| 03:14 | johnny has joined #ltsp | |
| 03:14 | <Appiah> find anything in the xorg log file mic_p ?
| |
| 03:14 | johnny is now known as Guest56838 | |
| 03:15 | <mic_p> I have remote syslog enabled. on the xorg log of the ltsp client or on the server?
| |
| 03:15 | Guest56838 has quit IRC | |
| 03:15 | <Appiah> client xorg.0.log file
| |
| 03:15 | or xorg.#.log
| |
| 03:15 | johnny1 has joined #ltsp | |
| 03:16 | johnny1 has left #ltsp | |
| 03:17 | <mic_p> Appiah: I don't know, however: http://pastebin.com/d1fd3396d
| |
| 03:18 | and, like now, I have no mouse attached. That work if connected. The keyboard no. on dmesg or remote syslog are both seen like usb devices.
| |
| 03:19 | <alkisg> mic_p: did you run ltsp-update-kernels? (if you're using a distro that uses it, that is...)
| |
| 03:19 | <mic_p> alkisg: I read a lot of posts that said that, but I don't know where that executable is. On the server I have debian etch/lenny
| |
| 03:20 | <alkisg> Just run this command on the server, I think it should be there in Lenny...
| |
| 03:20 | su, and then ltsp-update-kernels
| |
| 03:21 | <mic_p> found on /usr/sbin. I'll try to reboot the client
| |
| 03:23 | done. but nothing. Do I have to modify something before do a ltsp-update-kernels?
| |
| 03:24 | <alkisg> Nope...
| |
| 03:27 | <mic_p> alkisg: so, no keyboard?
| |
| 03:29 | <alkisg> So, I don't know, wait for other opinions... :)
| |
| 03:29 | <mic_p> alkisg: ok, thanks! :)
| |
| 03:32 | <Appiah> does it happen for all clients or just some/one ?
| |
| 03:35 | <mic_p> Appiah: for this type of hardware, yes. for all. I'm using pcengines alix3d3
| |
| 03:35 | <Appiah> O_o never heard of
| |
| 03:36 | small
| |
| 03:36 | <mic_p> pcengines.ch/alix3d3.htm are the specification page, if you need
| |
| 03:37 | I'm using this hardware for... all my projects. Are so cheaper and work so well at first time ( except usb keyboard and X on ltsp ;) )
| |
| 03:38 | <Appiah> "Your xorg.conf file is /tmp/xorg.conf.new"
| |
| 03:38 | can you pastebin that ?
| |
| 03:40 | <mic_p> on ltsp and server: ls: /tmp/xorg.conf.new: No such file or directory
| |
| 03:43 | highvoltage has quit IRC | |
| 03:43 | <Appiah> is ldm still on?
| |
| 03:43 | on the client
| |
| 03:45 | <mic_p> ldm is a daemon? if yes, no. ps ax | grep ldm return nothing
| |
| 03:46 | on the client, of course
| |
| 03:51 | wwx_ has quit IRC | |
| 03:51 | wwx has joined #ltsp | |
| 03:57 | sene has joined #ltsp | |
| 03:59 | vmlintu has joined #ltsp | |
| 04:02 | <mic_p> Appiah: sorry. Like now, no ldm (LDM_DIRECTX) is "false"
| |
| 04:04 | a good news. If I enable ldm, it works!
| |
| 04:04 | <Appiah> keyboard?
| |
| 04:04 | <mic_p> yes!
| |
| 04:04 | <Appiah> works if you set LDM_DIRECTX=True?
| |
| 04:04 | W T F
| |
| 04:06 | should have nothing to do with keyboard/usb
| |
| 04:09 | <mic_p> I think that... my fault: if I used an "old" lts.conf where I modified the screen_02 where I set shell|startx. If I "startx" on screen 02 the keyboard not works, but if set on 07... yes. With ldm active or not.
| |
| 04:10 | <Appiah> :)
| |
| 04:11 | <mic_p> Appiah: first of all, thanks! But it's a know "bug"? If no, I can submit a new one.
| |
| 04:22 | <Appiah> dont know much about touching screen_02
| |
| 04:22 | only know you can set shell there
| |
| 04:22 | and on 07/08 you're supposed to set LDM/RDP
| |
| 04:29 | <mic_p> Appiah: ok. I'll look if need to add a new bug and if yes, I'll submit it. thanks and bye!
| |
| 04:29 | mic_p has quit IRC | |
| 04:35 | Selveste1 has quit IRC | |
| 04:35 | Selveste1 has joined #ltsp | |
| 04:41 | mikkel has joined #ltsp | |
| 05:29 | Selveste1_ has joined #ltsp | |
| 05:30 | highvoltage has joined #ltsp | |
| 05:30 | spectra has joined #ltsp | |
| 05:30 | Selveste1 has quit IRC | |
| 05:45 | Ahmuck-Jr_ has joined #ltsp | |
| 05:45 | Ahmuck has quit IRC | |
| 05:45 | Ahmuck has joined #ltsp | |
| 05:47 | Selveste1__ has joined #ltsp | |
| 05:47 | Selveste1_ has quit IRC | |
| 05:55 | Ahmuck has quit IRC | |
| 05:57 | Ahmuck-Jr__ has joined #ltsp | |
| 05:57 | Ahmuck has joined #ltsp | |
| 05:59 | Ahmuck-Jr has quit IRC | |
| 06:00 | Ahmuck-Jr__ is now known as Ahmuck-Jr | |
| 06:12 | Ahmuck-Jr_ has quit IRC | |
| 06:20 | Ahmuck-Jr_ has joined #ltsp | |
| 06:21 | pmatulis has joined #ltsp | |
| 06:28 | Selveste1__ has quit IRC | |
| 06:28 | Selveste1__ has joined #ltsp | |
| 06:32 | hersonls has joined #ltsp | |
| 06:35 | Ahmuck-Jr has quit IRC | |
| 06:40 | Ahmuck has quit IRC | |
| 06:51 | Selveste1___ has joined #ltsp | |
| 06:52 | Selveste1__ has quit IRC | |
| 07:10 | Ahmuck-Jr has joined #ltsp | |
| 07:11 | gentgeen__ has joined #ltsp | |
| 07:12 | Ahmuck-Jr_ has quit IRC | |
| 07:21 | Selveste1____ has joined #ltsp | |
| 07:21 | Selveste1___ has quit IRC | |
| 07:24 | highvolt1ge has joined #ltsp | |
| 07:25 | scottmaccal has joined #ltsp | |
| 07:26 | mikkel has quit IRC | |
| 07:27 | highvoltage has quit IRC | |
| 07:31 | Ahmuck-Jr_ has joined #ltsp | |
| 07:41 | Ahmuck-Jr has quit IRC | |
| 07:43 | JaixBly has joined #ltsp | |
| 07:45 | <JaixBly> I use Ubuntu, I have just installed ltsp, how do I configure ltsp to work on single network adapter?
| |
| 07:45 | GodFather has joined #ltsp | |
| 07:45 | <JaixBly> I already have a dhcp server in the network, and my server have dhcp address
| |
| 07:46 | cmm1 has joined #ltsp | |
| 07:47 | <Appiah> just confiugre your dhcpserver to send the client image
| |
| 07:47 | check the dhcpd.conf to see how
| |
| 07:47 | (from ltsp)
| |
| 07:48 | <JaixBly> How do I do that with a win2008 server?
| |
| 07:48 | <Appiah> :)
| |
| 07:48 | well its dhcp like dhcp
| |
| 07:48 | <JaixBly> yep, that's what I thought
| |
| 07:48 | <Appiah> boot option
| |
| 07:49 | <JaixBly> need to enable bootp
| |
| 07:49 | <Appiah> and point to the tftpserver
| |
| 07:49 | and the path
| |
| 07:49 | <JaixBly> in the scope
| |
| 07:50 | <Appiah> well how many other places can you put it ?
| |
| 07:50 | <JaixBly> eh
| |
| 07:50 | none
| |
| 07:50 | <Appiah> Sorry I'm not to familiar with Windows DHCP, havent touched it since ... 8 years ago?!
| |
| 07:50 | <JaixBly> well, I have never thouched linux before, this is totally new to me
| |
| 07:51 | <Appiah> oh
| |
| 07:51 | <JaixBly> hat scope option do I have to set?
| |
| 07:52 | <Appiah> is there any windows dhcp server documentation you can refeer to?
| |
| 07:52 | beacuse telling a dhcp server where the tftpt server and the path is common bootp stuff
| |
| 07:53 | <JaixBly> probably on microsoft.com somewhere
| |
| 07:53 | well, I look through the options to see if I find anything like it
| |
| 07:53 | <Appiah> dont have access to a win2k# server atm =(
| |
| 07:57 | 17 Root Path
| |
| 07:57 | 66 Boot Server Host Name
| |
| 07:57 | 67 Bootfile Name
| |
| 07:58 | <JaixBly> what is the root path?
| |
| 07:59 | <alkisg> /opt/ltsp/i386
| |
| 08:00 | <scottmaccal> Morning all. I hope everyone had a good holiday.
| |
| 08:00 | <alkisg> Good morning
| |
| 08:01 | <JaixBly> eh, just like in dhcp.conf :-)
| |
| 08:01 | * ogra did ... the first winter i dont have to freeze since moving here ... http://ograblog.wordpress.com/2009/12/30/sudo-make-warm/ | |
| 08:02 | gentgeen__ has quit IRC | |
| 08:02 | Selveste1 has joined #ltsp | |
| 08:07 | Selveste1 has quit IRC | |
| 08:07 | Selveste1 has joined #ltsp | |
| 08:08 | Selveste1____ has quit IRC | |
| 08:19 | Egyptian[Home] has quit IRC | |
| 08:21 | Egyptian[Home] has joined #ltsp | |
| 08:23 | jcastro has quit IRC | |
| 08:23 | NeonLicht has quit IRC | |
| 08:23 | panthera has quit IRC | |
| 08:23 | prpplague_afk has quit IRC | |
| 08:23 | panthera_ has joined #ltsp | |
| 08:23 | NeonLicht has joined #LTSP | |
| 08:23 | prpplague_afk has joined #ltsp | |
| 08:24 | prpplague_afk is now known as prpplague | |
| 08:24 | jcastro has joined #ltsp | |
| 08:30 | spectra has quit IRC | |
| 08:31 | spectra has joined #ltsp | |
| 08:33 | alkisg has quit IRC | |
| 08:43 | sbalneav has joined #ltsp | |
| 08:43 | <sbalneav> Morning all.
| |
| 08:47 | <Appiah> heya
| |
| 08:51 | Gadi has joined #ltsp | |
| 08:52 | panthera_ is now known as panthera | |
| 08:53 | CAN-o-SPAM has joined #ltsp | |
| 08:54 | <sbalneav> So, I sat down last night, and mapped out in my mind how to completely re-do ldm, by creating a pam module to handle the ssh.
| |
| 08:54 | I plan to use libssh, and I'm going to have to duplicate the entire command socket thing.
| |
| 08:55 | however, at that point, ldm will switch over to using a pam authentication.
| |
| 08:55 | Which leads to an interesting development.
| |
| 08:56 | <Appiah> What's the advamtage of that?
| |
| 08:56 | <sbalneav> Proper password handling, handling failed passwords, etc.
| |
| 08:56 | extracting out the password screen scraping ugliness.
| |
| 08:57 | plus, we may not even need to use ldm anymore.
| |
| 08:57 | We could, in theory use gdm
| |
| 08:57 | <Appiah> why cant we use gdm today?
| |
| 08:57 | handling failed passwords , that would be nice! :D
| |
| 08:58 | <sbalneav> This has been covered about 1x10^1000 times in the mailing list.
| |
| 08:58 | GIYF
| |
| 08:58 | <Appiah> oh
| |
| 08:58 | I will do some reading then
| |
| 08:58 | <sbalneav> basically, in a nutshell
| |
| 08:59 | gdm = pam, ldm = ssh
| |
| 09:00 | however, if we extrapolate the ssh stuff INTO a pam module...
| |
| 09:00 | then we can EITHER convert ldm to use pam, or patch gdm to do what we want.
| |
| 09:00 | We need more control over the login than gdm gives us.
| |
| 09:01 | shawnp0wers has joined #ltsp | |
| 09:01 | <sbalneav> And we do handle failed passwords now, just not elegantly :)
| |
| 09:02 | Selveste1 has quit IRC | |
| 09:15 | scottmaccal has quit IRC | |
| 09:19 | Selveste1 has joined #ltsp | |
| 09:21 | <Gadi> !s
| |
| 09:21 | <ltspbot> Gadi: "s" :: Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 09:23 | <Gadi> sbalneav: I like ur plan - once you have ur module, whether we switch to gdm or not, it will still be useful
| |
| 09:23 | mushroomb1ue has joined #ltsp | |
| 09:23 | alkisg has joined #ltsp | |
| 09:28 | mushroomtwo has quit IRC | |
| 09:28 | GGD has joined #ltsp | |
| 09:30 | <ogra> !s
| |
| 09:30 | <ltspbot> ogra: "s" :: Scotty!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
| |
| 09:31 | <ogra> Gadi, you pung ?
| |
| 09:31 | alkisg has quit IRC | |
| 09:31 | <Gadi> ogra: did I?
| |
| 09:31 | <ogra> some days ago apparently
| |
| 09:31 | <Gadi> dude - that was so last year
| |
| 09:31 | :)
| |
| 09:31 | <ogra> lol
| |
| 09:31 | last decade as well !
| |
| 09:31 | <Gadi> hehe
| |
| 09:32 | Ive got 2010 problems now
| |
| 09:32 | :)
| |
| 09:32 | <ogra> with typing the date ?
| |
| 09:32 | just say no to whisky :)
| |
| 09:32 | alkisg has joined #ltsp | |
| 09:32 | <Gadi> never!
| |
| 09:33 | <ogra> heh
| |
| 09:36 | <sbalneav> ogra: Happy New Year!
| |
| 09:36 | Gadi: You To!!
| |
| 09:36 | err
| |
| 09:36 | Too
| |
| 09:36 | <ogra> same to you !
| |
| 09:39 | <Gadi> round o' whisky
| |
| 09:39 | <ogra> :)
| |
| 09:39 | <Gadi> on ogra ;)
| |
| 09:39 | <ogra> heh
| |
| 09:48 | highvolt1ge is now known as highvoltage | |
| 09:50 | Ahmuck-Jr_ has quit IRC | |
| 10:05 | Ahmuck has joined #ltsp | |
| 10:07 | spectra has quit IRC | |
| 10:08 | loathsome has quit IRC | |
| 10:10 | scottmaccal has joined #ltsp | |
| 10:11 | Selveste1_ has joined #ltsp | |
| 10:13 | dro has joined #ltsp | |
| 10:20 | Selveste1_ has quit IRC | |
| 10:27 | Selveste1 has quit IRC | |
| 10:28 | mushroomb1ue is now known as mushroomblue | |
| 10:40 | JaixBly has quit IRC | |
| 10:58 | GodFather has quit IRC | |
| 10:59 | alkisg has quit IRC | |
| 11:03 | Barbosa_ has joined #ltsp | |
| 11:07 | gentgeen__ has joined #ltsp | |
| 11:08 | dro_ has joined #ltsp | |
| 11:08 | alkisg has joined #ltsp | |
| 11:09 | Barbosa has quit IRC | |
| 11:10 | dro has quit IRC | |
| 11:33 | <sbalneav> Ahhhh
| |
| 11:33 | looks like libssh2.org may be a winner.
| |
| 11:36 | Eghie has joined #ltsp | |
| 11:46 | alkisg has quit IRC | |
| 11:51 | alkisg has joined #ltsp | |
| 12:05 | mushroomtwo has joined #ltsp | |
| 12:08 | mushroomblue has quit IRC | |
| 12:13 | alkisg has quit IRC | |
| 12:22 | litlebuda has joined #ltsp | |
| 12:25 | dro_ has quit IRC | |
| 12:26 | dro_ has joined #ltsp | |
| 12:28 | alkisg has joined #ltsp | |
| 12:33 | Ahmuck-Sr has joined #ltsp | |
| 12:35 | vagrantc has joined #ltsp | |
| 12:35 | mushroomb1ue has joined #ltsp | |
| 12:39 | Lns has joined #ltsp | |
| 12:48 | <Ahmuck-Sr> ltsp clients can use usb keyboard/mouse upon bootup?
| |
| 12:48 | mushroomtwo has quit IRC | |
| 12:56 | <alkisg> Ahmuck-Sr: Like all other PCs... When exactly do you mean?
| |
| 12:57 | <Ahmuck-Sr> i was looking at a usb only solution from aopen
| |
| 12:58 | * vagrantc would be surprised if any current distros don't support usb keyboad & mouse | |
| 13:01 | <Ahmuck-Sr> from startup
| |
| 13:01 | some windows machines don't
| |
| 13:01 | <vagrantc> i mean, some bios's don't support usb keyboards
| |
| 13:02 | <sbalneav> I've got several thin clients with usb only keyboard and mouse.
| |
| 13:03 | So, if the bios supports it, you're golden.
| |
| 13:03 | * alkisg thinks that every PC after 2000 supports it... | |
| 13:04 | <Ahmuck-Sr> sbalneav: ok, thx
| |
| 13:05 | <vagrantc> alkisg: oh, i've definitely seen *lots* of PCs newer than 2000 that don't support usb keyboards with bios
| |
| 13:05 | but once the OS loads, no problem.
| |
| 13:05 | * vagrantc is 3 feet from one such machine right now | |
| 13:05 | <alkisg> Really? I've seen more than a hundend different PC models, and all of them supported it...
| |
| 13:06 | <vagrantc> alkisg: really.
| |
| 13:06 | <alkisg> (except for the older ones, < 1998-99...
| |
| 13:06 | ok :)
| |
| 13:06 | <vagrantc> an AMD 1700 right here doesn't, for example.
| |
| 13:07 | at freegeek, we've been honing in on the fact that if it has PS/2 keyboard ports, don't assume it supports USB keyboards in bios.
| |
| 13:08 | <alkisg> what bios is that?
| |
| 13:08 | <vagrantc> hardware that doesn't have PS/2 keyboard ports is a whole different story.
| |
| 13:08 | alkisg: it's not a particular bios...
| |
| 13:08 | <highvoltage> alkisg: I've also seen many old computers (from >2000) that don't support usb keyboards in the bios
| |
| 13:08 | <vagrantc> alkisg: we go through hundreds of different bios's a day.
| |
| 13:14 | <ogra> vagrantc, hey
| |
| 13:15 | cliebow has joined #ltsp | |
| 13:15 | <ogra> vagrantc, i'd like to have a tour for the canonical ubuntu dev team at freegeek, do you do such stuff after hours too ? (we usually work until 6pm at sprints, not sure i can convince the managers to give us 1-2h during the day)
| |
| 13:15 | <sbalneav> Latest wrinkle is now mobos that don't have ps/2 ports, but DO allow you to turn off usb keyboard support in the bios. :)
| |
| 13:16 | If you do that accidentally, you have to yank the bios battery
| |
| 13:16 | :)
| |
| 13:16 | <ogra> cool
| |
| 13:16 | innovative !
| |
| 13:16 | <cliebow> security
| |
| 13:17 | <ogra> vagrantc, i'll mail you about it ...
| |
| 13:17 | * ogra has a bad tooth and needs to lay down again ... | |
| 13:17 | <ogra> ... another 12h to the dentist apptmnt.
| |
| 13:17 | <sbalneav> ouch
| |
| 13:17 | lost a filling?
| |
| 13:17 | NM, go lay down
| |
| 13:18 | <vagrantc> ogra: we could probably work something out, though it really would be good to catch the splendor of the organization with all the gears turning and such :)
| |
| 13:20 | <ogra> sbalneav, nope, root infection ... i'll likely lose the tooth
| |
| 13:20 | <sbalneav> eugh
| |
| 13:20 | no fun
| |
| 13:20 | hope you feel better soon
| |
| 13:21 | <alkisg> gute Besserung :(
| |
| 13:21 | <ogra> got painkillers (i have it since new years ... ) but the body gets used to them so they dont kill it completely
| |
| 13:21 | dentist appointment is tomorrow morning, just one more night
| |
| 13:23 | vagrantc, well, i'll try to get some midday hours, probanly i can get mark intrested, that would make it easier ...
| |
| 13:23 | <vagrantc> ogra: it is a shop that rolls out probably 10 ubuntu machines a week to new computer users...
| |
| 13:23 | <ogra> i just dont know when he will attend he usually only drops in for a day oor two during sprints
| |
| 13:23 | <vagrantc> ogra: and that's a low estimate
| |
| 13:24 | <ogra> i know :) i mailed that to the canonical list already
| |
| 13:24 | and a good bunch of people wants to see it
| |
| 13:25 | i guess people are still to busy with first day of work stuff atm but i expect to at least get 20 ppl out of the 40 that will be at the sprint
| |
| 13:25 | <vagrantc> ogra: early feb?
| |
| 13:25 | <ogra> first week
| |
| 13:25 | * ogra just booked his flights | |
| 13:25 | <vagrantc> look forward to it!
| |
| 13:25 | <ogra> me too :)
| |
| 13:25 | i love portland ... though i'm a bit scared about winter :)
| |
| 13:26 | <Ahmuck-Sr> http://www.newegg.com/Product/Product.aspx?Item=N82E16856140035&cm_re=aopen_barebone-_-56-140-035-_-Product
| |
| 13:26 | http://usa.aopen.com/products_detail.aspx?Auno=2946#
| |
| 13:26 | <ogra> anyway, i need to lay down again ... sitting upright starts to hurt ... see you all
| |
| 13:26 | <vagrantc> ogra: probably really grey and rainy, with a chance of a sunny dry spell.
| |
| 13:26 | <Ahmuck-Sr> no ps2 ports
| |
| 13:27 | <alkisg> Ahmuck-Sr: Max LAN Speed 10/100Mbps ?
| |
| 13:27 | <Ahmuck-Sr> :(
| |
| 13:31 | Ahmuck-Sr has quit IRC | |
| 13:54 | dro_ has quit IRC | |
| 13:54 | Ahmuck-Sr has joined #ltsp | |
| 13:56 | <Ahmuck-Sr> alkisg: u pointed me to something about group permission the other day. that bookmark is on another machine, any chance you have that bookmark handy again?
| |
| 13:57 | Selveste1_ has joined #ltsp | |
| 14:05 | spectra has joined #ltsp | |
| 14:07 | Ahmuck-Sr has quit IRC | |
| 14:08 | <alkisg> Ahmuck-Sr: sure :D
| |
| 14:09 | shawnp0wers has quit IRC | |
| 14:10 | <Ahmuck> i goofed
| |
| 14:10 | badly
| |
| 14:10 | <alkisg> Ahmuck: http://en.wikipedia.org/wiki/Setuid#setgid_on_directories
| |
| 14:11 | <sbalneav> Ahmuck: One problem you're likely to run into:
| |
| 14:12 | muzzio has joined #ltsp | |
| 14:12 | <sbalneav> Files *created* in the directory, will work correctly
| |
| 14:12 | but if one mv's a file in, either through a file manager, or the command line, it WON'T inherit the perms.
| |
| 14:12 | for that, you'll need acl's
| |
| 14:13 | <muzzio> hi guys... I need to upgrade the kernel from a fedora LTSP server, I can do it by hand? Or there is a fedora way to do this?
| |
| 14:14 | <sbalneav> https://help.ubuntu.com/community/UbuntuLTSP/ACLSupport
| |
| 14:14 | muzzio: upgrade the server's kernel? or the kernel in the chroot?
| |
| 14:15 | <muzzio> sbalneav: Sorry, the kernel in chroot
| |
| 14:16 | <sbalneav> You should be able to update it by:
| |
| 14:16 | chroot /opt/ltsp/i386
| |
| 14:16 | your yum update commands
| |
| 14:16 | exit
| |
| 14:16 | ltsp-update-image && ltsp update-kernels
| |
| 14:16 | AFAIK
| |
| 14:16 | <muzzio> sbalneav: Ok... but I think I will need to update this by hand
| |
| 14:17 | <sbalneav> what, as in manually fiddle with the kernel? i.e. add a driver to it etc?
| |
| 14:17 | <muzzio> Is fedora 10, and the kernel is not very well with the thin-client hardware
| |
| 14:17 | sbalneav: the thin-client doesnt complete the boot
| |
| 14:18 | sbalneav: its freeze.
| |
| 14:18 | so, I think this can be a kernel problem
| |
| 14:18 | <sbalneav> have you tried just updating the chroot first?
| |
| 14:18 | <muzzio> this steps?
| |
| 14:19 | yum update
| |
| 14:19 | <sbalneav> I'd start there with the procedure I outlined before starting to manually fiddle with the kernel
| |
| 14:19 | <muzzio> ltsp-update-image && ltsp update-kernels
| |
| 14:19 | ?
| |
| 14:20 | <sbalneav> I'm not familiar with Fedora, but whatever yum commands you'd do to update your packages.
| |
| 14:20 | but don't forget the chroot
| |
| 14:20 | <muzzio> ok. I will try this!
| |
| 14:20 | <sbalneav> the docs should explain the concepts
| |
| 14:20 | <muzzio> Thanks a lot for your help!
| |
| 14:20 | <sbalneav> !docs
| |
| 14:20 | <ltspbot> sbalneav: "docs" :: For the most current documentation, see https://sourceforge.net/apps/mediawiki/ltsp/index.php?title=Ltsp_LtspDocumentationUpstream
| |
| 14:20 | <sbalneav> ^^^
| |
| 14:28 | <Ahmuck> i just chown /home for user:user -R *
| |
| 14:37 | <muzzio> sbalneav: But, its possible to use a kernel from kernel.org in chroot?
| |
| 14:38 | <Lns> muzzio: it's possible just like it would be with any other system
| |
| 14:42 | the only diff is you're using a chroot environment so you have to chroot to it to update/compile your stuff
| |
| 14:42 | <muzzio> Lns: Ok, I did this on LTSP 4.2...
| |
| 14:42 | <Lns> same deal iirc
| |
| 14:43 | LTSP5 is just more integrated into each distribution's packaging system instead of providing all the bits itself
| |
| 14:43 | <muzzio> Lns: Ok... I will try this, if the chroot update doenst solve my problem
| |
| 14:43 | Lns: Yes.. I know, 5 is better!
| |
| 14:43 | <Lns> =)
| |
| 14:46 | mushroomb1ue is now known as mushroomblue | |
| 14:53 | <alkisg> Lns: I don't think that's true, at least in Ubuntu the kernel is customized
| |
| 14:54 | scottmaccal has quit IRC | |
| 14:54 | dx9s_work has joined #ltsp | |
| 14:55 | <Ahmuck> so, now that i've really mussed up, do i go back and do a chown on every user directory back to thier own user:user permissions - (not a ltsp question ...)
| |
| 14:56 | <alkisg> Ahmuck: did you only run chown? Or also chmod?
| |
| 14:58 | <Ahmuck> only chown. however things have quit working so to speak
| |
| 14:58 | <alkisg> (cd /home; for u in *; do echo sudo chown -R $u:$u /home/$u; done)
| |
| 14:58 | <Ahmuck> from the /home directory "sudo chown -R user:user *"
| |
| 14:58 | <alkisg> ^^ This should give you the necessary commands, so that you can then copy/paste them...
| |
| 14:58 | <Ahmuck> whew, thanks
| |
| 14:58 | <dx9s_work> I am being stupid.. can (via login prompt) login and then (via .profile) startx -> ~/.xsession .. but when using gdm to log in locally, .xsession is ignored and I get xterm instead of what I am starting in .xsession (which works from startx, but not gdm)... what am I missing.. assume that gdm does read the options in /etc/X11/Xsession.options ( allow-user-xsession) or something..
| |
| 14:59 | <Ahmuck> i normally don't mess things up that bad, but ... i guess that's why they named me Ahmuck, cause sometimes things go amuck
| |
| 14:59 | <cliebow> dont feel alone 8~)
| |
| 15:00 | <alkisg> dx9s_work: is this ltsp-related? do you mean ldm instead of gdm?
| |
| 15:00 | <dx9s_work> I am using ltsp for not-so-thin clients
| |
| 15:01 | and ldm is for networked accounts instead of local
| |
| 15:02 | knipwim_ has joined #ltsp | |
| 15:02 | muzzio has quit IRC | |
| 15:03 | <dx9s_work> was hoping somebody had a " .. oh.. I know what you are talking about and here.. read this..."
| |
| 15:03 | * dx9s_work is now reading http://lists.debian.org/debian-user/2002/05/msg00413.html | |
| 15:04 | hersonls has quit IRC | |
| 15:07 | <sbalneav> alkisg: I'm going to begin work shortly on a ssh-based pam module.
| |
| 15:07 | Gadi has quit IRC | |
| 15:07 | <sbalneav> using libssh2
| |
| 15:07 | <alkisg> sbalneav: wow!!! really really nice!
| |
| 15:07 | <sbalneav> I'm going to map out a wiki page tonight
| |
| 15:08 | If you're interested, we can do it together
| |
| 15:08 | <alkisg> I am, but I don't know how much I'll be able to help
| |
| 15:08 | <sbalneav> after that, I'll write the pam module to tie the dbus's together. (session part)
| |
| 15:08 | <alkisg> I've never programmed any glib app yet :(
| |
| 15:09 | <sbalneav> alkisg: Call it a learning experience then. :)
| |
| 15:09 | Gadi has joined #ltsp | |
| 15:09 | <sbalneav> I'll rough out the design tonight. I spent a lot of time last night thinking about it.
| |
| 15:09 | I'll post something to the -dev list as well.
| |
| 15:09 | * vagrantc looks for some unlearning experiences | |
| 15:10 | <alkisg> sbalneav: will that be usuable outside of ltsp?
| |
| 15:10 | <vagrantc> sbalneav: this is to hopefully end up with an ldm that has a cleaner ssh authentication/communication thingamawhatzit?
| |
| 15:10 | <sbalneav> that's the plan.
| |
| 15:11 | vagrantc: this is to have an ldm that has NO ssh authentication
| |
| 15:11 | <vagrantc> sbalneav: oh...
| |
| 15:11 | <sbalneav> it'll do pam authentication.
| |
| 15:11 | <vagrantc> i.e. ldm just passes the buck to pam?
| |
| 15:11 | nice.
| |
| 15:11 | <sbalneav> bingo
| |
| 15:11 | that way, we'll be able to support:
| |
| 15:11 | <vagrantc> will that make local and remote auth... the same?
| |
| 15:12 | * sbalneav grins | |
| 15:12 | <sbalneav> why yes
| |
| 15:12 | * vagrantc waits patiently :) | |
| 15:12 | <sbalneav> yes it will
| |
| 15:13 | <vagrantc> nice, nice!
| |
| 15:13 | <sbalneav> a libnss-ssh would also be in the offing
| |
| 15:13 | as a compantion to the libpam-ssh
| |
| 15:13 | * alkisg was just about to ask that | |
| 15:13 | <vagrantc> whoah
| |
| 15:13 | <sbalneav> If I'm gonna solve this problem...
| |
| 15:13 | I'm going to freaking *SOLVE THE LIVING DAYLIGHTS* out of it.
| |
| 15:14 | * vagrantc chuckles | |
| 15:14 | <alkisg> And I was about to start writing script to automate ldap installation in schools... with this it won't be needed anymore :D
| |
| 15:15 | <sbalneav> 'prolly take me a couple months to write both.
| |
| 15:15 | <alkisg> sbalneav: will you start off the same code base that you sent me some months ago? Or will it be a total rewrite?
| |
| 15:15 | <vagrantc> with a libpam-ssh and linss-ssh ... that would pretty much be a viable alternative to avoid having to configure LDAP?
| |
| 15:16 | * alkisg needs to start reading up... | |
| 15:16 | <sbalneav> total rewirte using libssh2
| |
| 15:16 | libssh2.org
| |
| 15:16 | pmatulis has quit IRC | |
| 15:16 | <sbalneav> vagrantc: correct.
| |
| 15:16 | BUT
| |
| 15:16 | <alkisg> Wow... cool! OK, I'm sold :)
| |
| 15:16 | <sbalneav> If we move to pam, then someone who DOES want to use ldap's in.
| |
| 15:17 | knipwim has quit IRC | |
| 15:17 | <sbalneav> auth via ldap, then just insert the libpam-ssh with USE_AUTHTOK
| |
| 15:17 | since libpam-ssh will have to create the socket we need.
| |
| 15:17 | we can also insert kerberos into the mix that way as well.
| |
| 15:17 | and
| |
| 15:18 | people who want ltsp with smartcard auth?
| |
| 15:18 | fixed.
| |
| 15:18 | via the libpam-smartcard module.
| |
| 15:18 | so, in short, a win-win-win all 'round.
| |
| 15:18 | Now all I gotta do is:
| |
| 15:18 | * vagrantc cheers! | |
| 15:18 | <sbalneav> design it (tonight)
| |
| 15:18 | then write it.
| |
| 15:19 | <vagrantc> and then someone's going to have to get it into the various distros...
| |
| 15:19 | petre has joined #ltsp | |
| 15:19 | * sbalneav looks at vagrant for debian | |
| 15:19 | <petre> greetings friends
| |
| 15:19 | <sbalneav> One of my design elements is "distro friendly"
| |
| 15:19 | <vagrantc> looks like debian has libssh-2 and libssh2-1
| |
| 15:20 | that makes things *obvious*
| |
| 15:20 | <sbalneav> i.e. having a config.d directory where you can drop config files in.
| |
| 15:20 | <Gadi> nss will be difficult
| |
| 15:20 | knipwim_ is now known as knipwim | |
| 15:20 | <Gadi> not having proper privileges
| |
| 15:20 | <sbalneav> Gadi: I'm thinking we only need /etc/passwd and /etc/group
| |
| 15:21 | and don't need to worry about /etc/shadow.
| |
| 15:21 | <Gadi> even so
| |
| 15:21 | ssh is primarily a user-domain thing
| |
| 15:21 | nss is a system-domain thing
| |
| 15:22 | its gonna be tricky
| |
| 15:22 | but, I don't doubt you'll solve it
| |
| 15:22 | :)
| |
| 15:22 | * dx9s_work thinks it is a bug in /etc/gdm/Xsession | |
| 15:22 | <dx9s_work> (out of context.. sorry)
| |
| 15:22 | <sbalneav> right, but if we use the command socket to send "getent passwd <blah> " commands over an established socket.
| |
| 15:23 | <Gadi> sbalneav: oh, like our current cheat
| |
| 15:23 | :)
| |
| 15:23 | <sbalneav> right
| |
| 15:23 | <Gadi> works for processes run as user
| |
| 15:23 | <sbalneav> right
| |
| 15:23 | and that's all we're trying to solve, atm
| |
| 15:23 | <Gadi> local procs run under sudo would fail
| |
| 15:24 | <sbalneav> yep.
| |
| 15:24 | for now.
| |
| 15:24 | here was my thought:
| |
| 15:24 | * alkisg votes sbalneav for president | |
| 15:24 | <sbalneav> libpam-ssh sets a session environment variable
| |
| 15:24 | <Gadi> and ur nss mod would have to find the right socket
| |
| 15:25 | <sbalneav> LIBNSS_SESSION_CONTROL_SOCKET or the like
| |
| 15:25 | libnss-ssh then, when it gets the chain, checks for the env var
| |
| 15:25 | <petre> sbalneav is Canadian; they don't have a president
| |
| 15:25 | <vmlintu> How about just loading user and group information for the user when logging in (pam-ssh) and store it in /etc/passwd + /etc/group ?
| |
| 15:26 | <sbalneav> We're doing that now, more or less
| |
| 15:26 | however, if someone mods group access during your session, you don't pick it up until you log in again.
| |
| 15:27 | but basically, all I'm proposing is, "do what we're doing now, but move it up into the libnss heirarchy"
| |
| 15:27 | It accomplishes 2 things:
| |
| 15:28 | 1) it lightens the LTSP code iself, since all that hackery moves out of the ltsp code. We simply assume our groups, etc. will be "good"
| |
| 15:28 | 2) it moves it all into a separate project, which puts all the "ugly" in one place, and in a way which MAY be usefil for other people/projects
| |
| 15:29 | i.e. having a network lab of standalone machines, but with a simple network auth scheme like alkisg wants.
| |
| 15:30 | <Gadi> hmm... actually you might be good to go with all apps
| |
| 15:30 | not just user processes
| |
| 15:30 | <sbalneav> Gadi: we should, so long as we pass through the env var.
| |
| 15:30 | <Gadi> because the nss lookup will be done in system space
| |
| 15:31 | <sbalneav> sudo does some env-var cleaning
| |
| 15:31 | <vmlintu> Are there applications that would need the groups to be updated in realtime?
| |
| 15:31 | <sbalneav> but I think there's a way to tell it NOT to clean some bars.
| |
| 15:31 | vars
| |
| 15:31 | * alkisg guesses that if that libpam-ssh provides some hooks, it'll even be possible to modify some of the ltsp bits so that they are ran from gdm... resulting in standalone PCs being able to function as LTSP clients ;) | |
| 15:31 | <Gadi> sudo shouldnt affect the env
| |
| 15:31 | you should be ok
| |
| 15:31 | etyack has joined #ltsp | |
| 15:31 | <sbalneav> Oh, yeah, sudo mucks with your env big time.
| |
| 15:32 | <Gadi> ssh does
| |
| 15:32 | <sbalneav> The -E (preserve environment) option will override the env_reset option in sudoers(5)).
| |
| 15:32 | <Gadi> ah
| |
| 15:32 | eh, ok
| |
| 15:32 | you can also do it with a file in /var/run rather than an env var
| |
| 15:32 | <sbalneav> Well, that's another issue.
| |
| 15:33 | potentially, you could use gdm instead of ldm
| |
| 15:33 | problem is, gdm doesn't have the fine grained control over "when to run things" that we've added
| |
| 15:33 | and even if you FIXED gdm to do that
| |
| 15:33 | <Gadi> right - but, then we need an xdg-compliant way of porting all the .d/ scripts :)
| |
| 15:33 | <sbalneav> then what about kdm, etc.
| |
| 15:33 | right.
| |
| 15:34 | so, I'd say we want to KEEP ldm as a "desktop agnostic" way to log in
| |
| 15:35 | Gadi: Only problem with a file in /var/run is: running two ldm's to two different servers.
| |
| 15:36 | <Gadi> true
| |
| 15:36 | but, we dont support that right now, either
| |
| 15:36 | :)
| |
| 15:36 | <sbalneav> which one does a libnss-ssh pick to send it's commands over? That's why I think we need a env var, but I'll write it up in the wiki tonight, and then we can bang it around in -dev
| |
| 15:36 | We don't?
| |
| 15:36 | <Gadi> I think some of the localapps stuff breaks if you do that
| |
| 15:36 | not 100% sure
| |
| 15:36 | never tried
| |
| 15:36 | <sbalneav> AFAIK, when I rewrote ldm, I specifically wrote it with the idea of running 2 or more
| |
| 15:36 | <Gadi> we certainly dont test it
| |
| 15:36 | :)
| |
| 15:37 | oh, ldm is fine
| |
| 15:37 | <sbalneav> Well, it *should* be a use case.
| |
| 15:37 | <Gadi> I think its just the localapps stuff
| |
| 15:37 | <sbalneav> so now's the chance to design that it by default
| |
| 15:37 | <Gadi> because it mucks with /etc/passwd et al
| |
| 15:38 | brb
| |
| 15:44 | <petre> when a client is booting, once the kernel is loaded, how does it determine which network driver module to use?
| |
| 15:44 | udev?
| |
| 15:45 | <sbalneav> petre: yeah
| |
| 15:46 | <petre> is udev contained within an initrd?
| |
| 15:48 | <sbalneav> I think it is...
| |
| 15:48 | Hmmmm
| |
| 15:48 | could be just a bulk-module-load.
| |
| 15:48 | * sbalneav can't remember | |
| 15:49 | <sbalneav> We generally don't care anymore, since that's the distro's kernel's problem :)
| |
| 15:50 | <vagrantc> stgraber: i was wondering if you minded me changing the ltsp-{shutdown|restart}.desktop files to use alkisg's new method
| |
| 15:51 | <petre> but udev is in the initrd, or rather, initramfs, isn't it?
| |
| 15:51 | artista_frustrad has joined #ltsp | |
| 15:51 | <vagrantc> alkisg: and i was wondering if we could change it to LTSP_LOGOUT_ACTION ?
| |
| 15:51 | <petre> And is the initramfs just bundled in with the kernel file?
| |
| 15:51 | Or is that another "depends on the distro" situation?
| |
| 15:52 | <vagrantc> alkisg: if it's intended to be general purpose, i think we should change the name to LTSP_LOGOUT_ACTION and keep it in ltsp sources ...
| |
| 15:52 | alkisg: if it's just going to be an LDM feature, we should move it all to ldm sources and keep the variable name.
| |
| 15:52 | <sbalneav> petre: I think so.
| |
| 15:52 | * alkisg is thinking of possible non-ldm uses... | |
| 15:52 | <alkisg> *use cases...
| |
| 15:53 | <petre> sbalneav, ok, tx, at least that suggests I'm going in the right direction
| |
| 15:53 | alkisg, tx for your help last week on my netboot problem; I did get it working finally.
| |
| 15:53 | <alkisg> nice :)
| |
| 15:54 | vagrantc: ok, but please tell me when you rename it to LTSP_LOGOUT_ACTION so that I change my proposed patch to ubuntu's gnome-session... and thanks :)
| |
| 15:55 | <vagrantc> alkisg: i think i'll change it right now, unless stgraber objects
| |
| 15:57 | adrienn has joined #ltsp | |
| 15:57 | <alkisg> vagrantc: could you also make a small change in client/screen-session.d/XS00-halt-reboot ?
| |
| 15:57 | case "$(cat /var/run/ldm-logout-action)" in <== either check if the file exists before the cat, or put a 2>/dev/null...
| |
| 15:58 | <vagrantc> alkisg: stgraber already did that
| |
| 15:58 | <alkisg> Ah, good :)
| |
| 16:00 | <adrienn> Hello, can you help me find an answer to my problem? When I put a cd-rom on the server it shows on every thin clients and then I'm unable to unmount it, I've got the error : only user "user" can unmount ...
| |
| 16:01 | * alkisg had this code after ldm was stopped, initially, so the file was always there... | |
| 16:02 | <vagrantc> and now i need to consider if i want to move rc.d/X99-zzz-logout-action into the ltsp sources...
| |
| 16:03 | i think it's technically an LTSP feature that ldm is plugging in to...
| |
| 16:03 | so it makes sense to leave it in ldm.
| |
| 16:04 | though you run into versioned dependency issues
| |
| 16:05 | but it's also a pain to keep interoperability code in multiple sources...
| |
| 16:06 | <alkisg> Well technically the xprop can be set even if ldm wasn't started at all, and the user invoked e.g. rdesktop or xterm...
| |
| 16:06 | <vagrantc> right
| |
| 16:06 | <alkisg> So maybe it should all go in the ltsp tree
| |
| 16:06 | <vagrantc> that parts clearly in ltsp.
| |
| 16:06 | the question is if the ldm-specific hook should go into ltsp.
| |
| 16:07 | we could actually have both, i guess.
| |
| 16:07 | <alkisg> Why is rc.d/X99-zzz-logout-action an ldm-specific hook?
| |
| 16:07 | * alkisg just put it there because the file already existed... | |
| 16:07 | <vagrantc> because it's installed in LDM's hooks.
| |
| 16:08 | <alkisg> OK, but if you move it, it won't be anymore :D
| |
| 16:08 | <vagrantc> there's probably a way we could hook it into the screen-session.d scripts...
| |
| 16:09 | <alkisg> That runs after X is started, right? So there will be flickering...
| |
| 16:09 | * vagrantc wonders if client/screen-session.d/XK* happen while X is still running. | |
| 16:09 | <vagrantc> well, i don't want to move XS00-halt-reboot. that should stay where it is.
| |
| 16:09 | <alkisg> I think I tried both on start and on stop, and both had problems
| |
| 16:09 | Either flickering, or ghost processes
| |
| 16:10 | <vagrantc> alkisg: i'll leave that alone. just the part that sets the xprop is what i'm looking at moving.
| |
| 16:10 | <alkisg> The .desktop files?
| |
| 16:11 | <vagrantc> well, that too.
| |
| 16:11 | <alkisg> In my labs, gnome-session sets the xprop...
| |
| 16:11 | <vagrantc> er, rather, *checks* the xprop
| |
| 16:11 | <alkisg> Ah, ok
| |
| 16:11 | <vagrantc> currently in ldm-trunk/rc.d/X99-zzz-logout-action
| |
| 16:12 | i think that should be moved to something in ltsp-trunk/client/screen-session.d/XK*
| |
| 16:12 | <alkisg> Maybe you should remove the reboot/shutdown that stgraber has there, as it leaves ghost processes
| |
| 16:12 | <vagrantc> but it looks like the XK scripts are unimplemented.
| |
| 16:12 | artista-frustrad has joined #ltsp | |
| 16:12 | <Gadi> nope
| |
| 16:12 | <vagrantc> yeah, ldm-trunk/rc.d/X99-zzz-logout-action will go away completely.
| |
| 16:13 | <Gadi> screen-session.d is when X is not around
| |
| 16:13 | <vagrantc> ah. for shame.
| |
| 16:13 | <Gadi> you would need to extend xinitrc
| |
| 16:14 | but Xatom setting is an ldm-pnly thing atm anyway
| |
| 16:14 | <vagrantc> sure, but it's easy enough to leave available to other things.
| |
| 16:15 | <Gadi> hard to set an xatom in an rdp session ;)
| |
| 16:15 | <alkisg> Gadi: why not? sshd on the client, and a script on the windows server ;)
| |
| 16:15 | <Gadi> painful
| |
| 16:16 | <alkisg> Right, it's easier if one just creates the resulting file in /var/run :)
| |
| 16:16 | artista_frustrad has quit IRC | |
| 16:16 | <Gadi> bbiab
| |
| 16:16 | <vagrantc> Gadi: based on the error codes from rdesktop (or some other hypothetical session) you might be able to set xatoms
| |
| 16:17 | big *might*
| |
| 16:17 | <Gadi> at that point you can simply act
| |
| 16:17 | Ahmuck-Jr has joined #ltsp | |
| 16:18 | <vagrantc> and potentially implement an inconsistant shutdown policy
| |
| 16:20 | <sbalneav> Okiedokie, heading home for the day.
| |
| 16:20 | be on later tonight.
| |
| 16:20 | <vagrantc> we'd have to not exec the xinitrc scripts for it to work...
| |
| 16:20 | <alkisg> bye sbalneav
| |
| 16:20 | <vagrantc> i think i'll leave it as an ldm hook for now.
| |
| 16:26 | we can have both, at least for a while.
| |
| 16:27 | etyack has quit IRC | |
| 16:31 | CAN-o-SPAM has quit IRC | |
| 16:32 | artista_frustrad has joined #ltsp | |
| 16:33 | * alkisg would love an "update-pxelinux-cfg" program, similar to "update-inetd"... :-/ | |
| 16:34 | adrienn has quit IRC | |
| 16:41 | artista-frustrad has quit IRC | |
| 16:45 | <vagrantc> ok, pushing my latest *-logout-action commits.
| |
| 16:46 | doesn't require any changes to ldm, and keeps all the code within ltsp-trunk mostly reasonably...
| |
| 16:46 | though it puts the ldm hook in the localapps tree, which isn't technically correct.
| |
| 16:47 | but then it requires no changes to packaging :)
| |
| 16:47 | cmm1 has quit IRC | |
| 17:10 | prpplague is now known as prpplague_afk | |
| 17:19 | <dx9s_work> welp discovered that /etc/gdm/Xsession has issues (like not setting USERXSESSIONRC and other things so /etc/X11/Xsession.d/* is happy).. work around... mv /etc/gdm/Xsession /etc/gdm/Xsession.save; ln -s /etc/X11/Xession /etc/gdm/Xsession ...
| |
| 17:20 | then one can put things into $HOME and get things to work more like you expect
| |
| 17:22 | <vagrantc> alkisg: so, with XS00-halt-reboot it doesn't flicker or have issues with lingering processes, right?
| |
| 17:23 | <alkisg> vagrantc: no, it works fine. It does show a 2 sec delay (after logoff) until that point is reached, but it isn't noticable.
| |
| 17:24 | <vagrantc> yeah, once boot-up times are perfected, we'll get people complaining about how long it takes to power down.
| |
| 17:25 | <alkisg> :P
| |
| 17:25 | <vagrantc> we'll burn that bridge while we're still on it
| |
| 17:25 | * alkisg wonders why there are lingering processes when you logoff and kill the X, though... | |
| 17:27 | <vagrantc> at some point, we should be able to remove X99-zzz-logout-action entirely. for the moment, the ldm variation and the ltsp variation can co-exist.
| |
| 17:27 | alkisg: using xprop for the shutdown/reboot stuff was brilliant.
| |
| 17:27 | <alkisg> It was Gadi's idea...
| |
| 17:28 | I just kept nagging until it was included :D
| |
| 17:28 | <vagrantc> Gadi: using xprop for the shutdown/reboot stuff was brilliant!
| |
| 17:28 | * Gadi doesnt recall exactly - but if its rube-goldberg you're after ... :) | |
| 17:29 | jelly-bean has joined #ltsp | |
| 17:29 | grey-monkey has joined #ltsp | |
| 17:31 | <Gadi> alkisg: do you remember which procs lingered?
| |
| 17:31 | <alkisg> Gadi: no, they were many, though... 10 of them or so
| |
| 17:31 | (for one user)
| |
| 17:31 | mushroomblue has quit IRC | |
| 17:32 | <alkisg> I think that "fclose" in ldm, that fixed the ldm-hanging-on-logout bug in June, leaves the processes running...
| |
| 17:32 | ...and fortunately they're killed when X stops
| |
| 17:32 | <vagrantc> Gadi: would you be up for a hackfest in NY late july or mid august?
| |
| 17:33 | <Gadi> but, if they are killed when X stops, why not when the client shuts down?
| |
| 17:33 | vagrantc: what a question!
| |
| 17:33 | :)
| |
| 17:33 | <vagrantc> Gadi: i'll be in the neighborhood for the debian conference, and i think jammcq and maybe sbalneav expressed interest at some point
| |
| 17:34 | <Gadi> awesome
| |
| 17:34 | is the debcon at javitz?
| |
| 17:34 | <jelly-bean> with latest ubuntu karmic alternative install ltsp when the thin clients shut down they get a bunch of endless SQUASHFS errors scrolling down the page and the machine never turns off
| |
| 17:34 | someone showed me a patch and i was able to just update my apt sources to get it
| |
| 17:34 | <Gadi> jelly-bean: thats a different bug
| |
| 17:34 | <vagrantc> Gadi: http://debconf10.debconf.org/
| |
| 17:35 | <alkisg> Gadi: "Putting the script in rc.d/X* or rc.d/K* or in ldm.c leaves user processes running, so it's *not* OK there." ==> at that point, X hasn't properly stopped, so a reboot -fp there just kills X without giving it the time to stop the processes, I guess...
| |
| 17:35 | <Gadi> thats the poweroff -fp bug
| |
| 17:35 | <jelly-bean> does anyone know where i can find that again
| |
| 17:35 | <vagrantc> Gadi: http://debconf10.debconf.org/venue.xhtml columbia university
| |
| 17:35 | <jelly-bean> Gadi: ;link ?
| |
| 17:35 | <vagrantc> Gadi: basically the conference is july 25th through august 7th ...
| |
| 17:35 | <Gadi> jelly-bean: dunno where the patch is, but I can tell you what to fix
| |
| 17:35 | :)
| |
| 17:36 | vagrantc: excellent
| |
| 17:36 | I love local venues
| |
| 17:36 | :)
| |
| 17:37 | <alkisg> jelly-bean: http://bazaar.launchpad.net/~ltsp-upstream/ltsp/ldm-trunk/revision/1222
| |
| 17:37 | For just getting newer debs, you can use this:
| |
| 17:37 | <jelly-bean> Gadi: what
| |
| 17:37 | <alkisg> !stgraber-ppa
| |
| 17:37 | <ltspbot> alkisg: "stgraber-ppa" :: https://launchpad.net/~stgraber/+archive/ppa
| |
| 17:38 | <Gadi> ah, alkisg to the rescue!
| |
| 17:38 | <vagrantc> that patch causes some thin clients to fail to properly shut down.
| |
| 17:39 | at least on debian ... while the without -fp it worked fine.
| |
| 17:39 | <jelly-bean> vagrantc: what do u suggest
| |
| 17:39 | <vagrantc> i suppose i should just patch it in debian.
| |
| 17:39 | * Gadi slaps head about the lingering procs - of course X hasn't stopped properly yet! | |
| 17:40 | <alkisg> Gadi: well, the user has logged off - there shouldn't be any processes running even if X wasn't killed :-/
| |
| 17:40 | vagrantc: wasn't that -fp there for some years, before stgraber changed it to -f ?
| |
| 17:40 | <Gadi> alkisg: well, the session has ended, but you still have the ssh socket
| |
| 17:40 | <vagrantc> alkisg: yeah, and i always complained about it being broken on debian.
| |
| 17:40 | <alkisg> Heh :)
| |
| 17:41 | <vagrantc> i was so happy when it got fixed, and then unhappy to see it brokened.
| |
| 17:41 | <Gadi> what kind of client didnt like it?
| |
| 17:41 | <jelly-bean> how do i add that stgraber
| |
| 17:42 | just paste that url as a source?
| |
| 17:42 | <Gadi> it should be violent enough for all clients
| |
| 17:42 | <jelly-bean> do i need a gpg key?
| |
| 17:42 | <alkisg> jelly-bean: what ubuntu version do you have?
| |
| 17:42 | karmic, sorry
| |
| 17:42 | <Gadi> otherwise, we should start using: echo o>/proc/sysrq-trigger
| |
| 17:42 | <jelly-bean> karmic ya
| |
| 17:42 | <Gadi> ;)
| |
| 17:42 | <alkisg> sudo add-apt-repository ppa:stgraber
| |
| 17:42 | <vagrantc> Gadi: stupid compaq desktops, i think. maybe qemu virtual thin clients ... i forget exactly
| |
| 17:43 | Gadi: but they wouldn't fully power down with "-fp"
| |
| 17:43 | <Gadi> sounds like a bios thing
| |
| 17:43 | <vagrantc> the OS looked like it stopped, but the machine would still be sitting there, sucking electrons
| |
| 17:43 | * vagrantc prefers a "proper" shutdown process anyways | |
| 17:44 | <Gadi> must be a windows user ;)
| |
| 17:44 | <vagrantc> i know the nbd-client hooks for nbd root properly recognize to not be bringing down networking when the root filesystem is network based....
| |
| 17:44 | <alkisg> "You may now poweroff your computer properly"
| |
| 17:44 | <vagrantc> no, not *that* kind of proper
| |
| 17:45 | just go through the motions of shutting down all processes and whatnot.
| |
| 17:45 | runlevel 0 or 6, like it was meant to be.
| |
| 17:45 | loathing has joined #ltsp | |
| 17:45 | <Gadi> its always the whatnot that gets us ;)
| |
| 17:45 | * vagrantc seems to handle whatnot fine | |
| 17:46 | * vagrantc fires up the grumpitude engine | |
| 17:46 | * vagrantc is tempted to do a new ltsp version.... | |
| 17:47 | <vagrantc> i've mostly been letting stgraber do all the version tagging and snagging the tarballs, lately...
| |
| 17:47 | <Gadi> bbiab
| |
| 17:53 | <Lns> I think I was one of the ones that was talking about compaq desktops/crappy BIOS not shutting down w/ -fp..not certain though.. it would shut down but not wake up via WoL so my startup scripts wouldn't work in the mornings
| |
| 17:53 | I'm sure there's no way to detect which options to use..?
| |
| 18:00 | Ahmuck-Jr has quit IRC | |
| 18:05 | <stgraber> vagrantc: please wait, I'm waiting for the fat client script from highvoltage
| |
| 18:05 | vagrantc: it seems to work well so I'll just push it before you or I release ;)
| |
| 18:05 | <vagrantc> oops.
| |
| 18:06 | <stgraber> or did you already ? ;)
| |
| 18:06 | <vagrantc> stgraber: i pushed, but didn't tag,.
| |
| 18:06 | <alkisg> stgraber: any info on how to use that fat client script?
| |
| 18:06 | <vagrantc> stgraber: i pushed the changes with LDM_LOGOUT_* to LTSP_LOGOUT_*
| |
| 18:06 | <stgraber> vagrantc: oh, mind if I do some bzr magic ? and add that before the actual tag ?
| |
| 18:06 | <vagrantc> stgraber: no tag
| |
| 18:07 | stgraber: i misinterpreted what you meant to wait on
| |
| 18:07 | <stgraber> oh, great, then I can still add that plugin :)
| |
| 18:07 | <vagrantc> stgraber: i'll gladly leave the tagging/release to you :)
| |
| 18:07 | <stgraber> alkisg: ltsp-build-client --arch i386 --fat-client
| |
| 18:07 | alkisg: then FAT_CLIENT=True and you're done
| |
| 18:07 | <alkisg> wow
| |
| 18:07 | <vagrantc> stgraber: but i did mess with the logout-action code.
| |
| 18:08 | * alkisg will try this tomorrow, sounds very promising... | |
| 18:09 | <stgraber> vagrantc: yep, I saw these, I'll quickly make sure I don't have weird things in my packaging using that (I don't remember if I copy these examples or if I have my own inside the package branch)
| |
| 18:09 | <vagrantc> stgraber: i think, the way i did it, the old stuff will continue to work until we yank it from ldm
| |
| 18:10 | i.e. it should work both ways.
| |
| 18:10 | \
| |
| 18:11 | * stgraber quickly checks his LTSP todo list for other things to include | |
| 18:12 | <vagrantc> sbalneav: i just realized another nice feature of your libssh2 master plans ... you can run screensavers as a localapp, and when it locks, they'll be able to auth with their password.
| |
| 18:12 | i keep wanting to run screensavers as a local app, but that's prevented me from doing so.
| |
| 18:14 | <alkisg> Hey, libssh2 supports the "none" cipher?!
| |
| 18:14 | <stgraber> ok, all the things I had waiting to be commited now are in, so I'm ready to tag a new LTSP
| |
| 18:15 | are we still ok with 5.1.100 being 5.2.0 ? (I suspect 5.1.98 to be slightly buggy so that will let us time to fix it and have something working correctly for 5.2.0)
| |
| 18:16 | tagged + pushed
| |
| 18:16 | <vagrantc> stgraber: yeah, i'd like to see 5.2.0 then :)
| |
| 18:16 | but i guess we can do it whenever
| |
| 18:22 | alkisg has quit IRC | |
| 18:23 | artista_frustrad has quit IRC | |
| 18:27 | Egyptian[Home] has quit IRC | |
| 18:27 | * stgraber quickly setups ltsp-cluster at home to test 5.1.98 ;) | |
| 18:29 | Egyptian[Home] has joined #ltsp | |
| 18:37 | * vagrantc wonders if the ltsp-logout-action patches are feasible to get into upstream gnome-session | |
| 18:39 | mushroomblue has joined #ltsp | |
| 18:40 | staffencasa has quit IRC | |
| 18:46 | dx9s_work has quit IRC | |
| 18:58 | <Lns> vagrantc: sbalneav is an upstream gnome guy now, right?
| |
| 19:08 | mushroomtwo has joined #ltsp | |
| 19:08 | mushroomblue has quit IRC | |
| 19:08 | spectra has quit IRC | |
| 19:08 | cliebow has quit IRC | |
| 19:08 | vagrantc has quit IRC | |
| 19:08 | klausade has quit IRC | |
| 19:08 | japerry has quit IRC | |
| 19:08 | tstafford has quit IRC | |
| 19:08 | ltsplogbot has quit IRC | |
| 19:08 | petre has quit IRC | |
| 19:08 | vmlintu has quit IRC | |
| 19:08 | sene has quit IRC | |
| 19:08 | wwx has quit IRC | |
| 19:08 | Sarten-X has quit IRC | |
| 19:08 | ogra has quit IRC | |
| 19:08 | Faithful has quit IRC | |
| 19:08 | alexqwesa has quit IRC | |
| 19:08 | moldy has quit IRC | |
| 19:08 | mushroomtwo is now known as mushroomblue | |
| 19:10 | moldy has joined #ltsp | |
| 19:12 | hersonls has joined #ltsp | |
| 19:17 | ltsplogbot has joined #ltsp | |
| 19:17 | tstafford has joined #ltsp | |
| 19:17 | japerry has joined #ltsp | |
| 19:17 | klausade has joined #ltsp | |
| 19:17 | vagrantc has joined #ltsp | |
| 19:17 | cliebow has joined #ltsp | |
| 19:17 | spectra has joined #ltsp | |
| 19:17 | alexqwesa has joined #ltsp | |
| 19:17 | petre has joined #ltsp | |
| 19:17 | vmlintu has joined #ltsp | |
| 19:17 | sene has joined #ltsp | |
| 19:17 | wwx has joined #ltsp | |
| 19:17 | Sarten-X has joined #ltsp | |
| 19:17 | ogra has joined #ltsp | |
| 19:17 | Faithful has joined #ltsp | |
| 19:24 | Faithful has joined #ltsp | |
| 19:25 | Ahmuck-Jr has joined #ltsp | |
| 19:25 | Lns has quit IRC | |
| 19:26 | try2free has joined #ltsp | |
| 19:30 | try2free has left #ltsp | |
| 19:47 | ogra has quit IRC | |
| 19:48 | ogra has joined #ltsp | |
| 20:06 | etyack has joined #ltsp | |
| 20:15 | hersonls_ has joined #ltsp | |
| 20:29 | litlebuda has quit IRC | |
| 20:32 | hersonls has quit IRC | |
| 20:45 | GGD has quit IRC | |
| 20:50 | etyack has quit IRC | |
| 20:53 | japerry has quit IRC | |
| 20:55 | <sbalneav> Evening all
| |
| 21:05 | artista_frustrad has joined #ltsp | |
| 21:09 | * vagrantc waves | |
| 21:11 | Barbosa has joined #ltsp | |
| 21:21 | hersonls_ has quit IRC | |
| 21:23 | artista-frustrad has joined #ltsp | |
| 21:23 | artista_frustrad has quit IRC | |
| 21:26 | Ahmuck-Jr has quit IRC | |
| 21:31 | petre has quit IRC | |
| 21:41 | artista-frustrad has quit IRC | |
| 22:32 | japerry has joined #ltsp | |
| 22:58 | <sbalneav> aaaaaaaaand mail sent to ltsp-developer
| |
| 22:59 | For anyone interested in a giggle:
| |
| 22:59 | https://blueprints.edge.launchpad.net/ltsp/+spec/libpam-sshauth
| |
| 23:41 | <vagrantc> sbalneav: ltsp6, eh?
| |
| 23:45 | stgraber: we'd better get ltsp5.2 out before it's too late!
| |
| 23:49 | <Ryan52> I thought we didn't want to use libssh2?
| |
| 23:55 | spectra has quit IRC | |
| 23:55 | <vagrantc> sbalneav's gonna fix it up real good.
| |
| 23:56 | 13:14 < sbalneav> I'm going to freaking *SOLVE THE LIVING DAYLIGHTS* out of it.
| |
| 23:56 | my favorite quote of the year so far.
| |
| 23:56 | spectra has joined #ltsp | |
| 23:58 | <Ryan52> heh
| |