IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 19 November 2010 (all times are UTC)
| 00:18 | artista_frustrad has joined #ltsp | |
| 00:22 | artista_frustrad has quit IRC | |
| 00:41 | daya has quit IRC | |
| 00:53 | daya has joined #ltsp | |
| 01:12 | evil_root is now known as zz_evil_root | |
| 01:28 | Uzzi has joined #ltsp | |
| 01:30 | <Uzzi> is it possible ltsp without dhcp?
| |
| 01:32 | <muppis> Uzzi, possbile. Describe more?
| |
| 01:33 | Oh, my typos.. Getting headache from Windows 7..
| |
| 01:33 | <Uzzi> muppis, My network is very strange, I've a lot vlan and geographic wlan and my network admin said that is not simple to have a dchp server in all vlan. Than is it possible to use ltsp without dhcp server
| |
| 01:34 | Can I assign to client an ip and tell its where is the root dir to load the os?
| |
| 01:34 | <Appiah> is there not dhcpserver in the network today?
| |
| 01:34 | s/not/no
| |
| 01:36 | <Uzzi> Yes but it is not found in all vlan
| |
| 01:36 | <Appiah> what a mess..
| |
| 01:37 | <Uzzi> My vlans don't like dhcp (my network admin tell taht!!!!)
| |
| 01:42 | <johnny> yes it's possible to do that
| |
| 01:43 | the hard part is for them to get the initial pxe boot code
| |
| 01:44 | <Uzzi> I'm searchig documetations but now I've not found
| |
| 01:47 | <alkisg> A custom gpxe script from rom-o-matic.net makes wonders
| |
| 01:47 | Uzzi: is there a (different) dhcp server available on each vlan?
| |
| 01:48 | <Uzzi> no, we have not Dhcp servers
| |
| 01:48 | there is only mine!
| |
| 01:48 | <alkisg> OK, then provide all the boot info in a gpxe script
| |
| 01:49 | <Uzzi> ok alkisg now I'm tryng do it
| |
| 01:49 | <alkisg> Uzzi: use google translate on that message: http://users.sch.gr/alkisg/tosteki/index.php?topic=1451.msg30596#msg30596
| |
| 01:50 | I have all the necessary steps + gpxe commands there
| |
| 01:50 | <Uzzi> alkisg, are you live in Grece?
| |
| 01:53 | <alkisg> Yup
| |
| 01:54 | You're in Italy? Close enough :)
| |
| 01:56 | <Uzzi> Yes ;P
| |
| 01:57 | I've sudied Ancient Greek
| |
| 02:05 | dobber has joined #ltsp | |
| 02:12 | gnunux has joined #ltsp | |
| 02:15 | <gnunux> hi
| |
| 02:27 | <Uzzi> alkisg, this is disk error:1005 AX:01212 BX:0000 CX:00001 DX:0000
| |
| 02:49 | artista_frustrad has joined #ltsp | |
| 02:51 | <alkisg> Uzzi: try with another disk and with gpxe 1.0.0 instead of 1.0.1
| |
| 02:52 | <Uzzi> I've used 1.0.0
| |
| 02:53 | artista_frustrad has quit IRC | |
| 02:57 | bobby_C has joined #ltsp | |
| 03:32 | Uzzi has quit IRC | |
| 03:37 | Trixboxer has joined #ltsp | |
| 03:41 | sepski is now known as sep | |
| 03:47 | Uzzi has joined #ltsp | |
| 03:48 | <Uzzi> Now It's ok setting ip, gateway ecc ecc but doesn't find nfs path
| |
| 03:53 | <alkisg> Uzzi: distro/version?
| |
| 03:54 | FGXR6 is now known as F-GT | |
| 03:54 | <Uzzi> debian
| |
| 03:54 | <alkisg> Set the nfs root path in gpxe
| |
| 03:56 | Uzzi: also the kernel command line is different on debian
| |
| 03:56 | That post was for Ubuntu
| |
| 03:56 | See your server's pxelinux.cfg/default for an example kernel command line for debian
| |
| 03:57 | <Uzzi> DEFAULT vmlinuz initrd=initrd.img nfsroot=172.16.1.131:/opt/ltsp/i386 boot=nfs ro quiet 3
| |
| 03:58 | <alkisg> Right, put those on the gpxe embedded script
| |
| 03:58 | <Uzzi> kernel /ltsp/i386/vmlinuz ro initrd=initrd.img nfsroot=172.16.1.131:/opt/ltsp/i386 ip=${ip}:${next-server}:${gateway}:${netmask}:${hostname}:eth0:none nbd_proxy=false nocompcache
| |
| 03:58 | now i'm tryng with this
| |
| 04:00 | <alkisg> Uzzi: also there's no need to specify "nbd_proxy=false nocompcache", these aren't used on debian anyway
| |
| 04:00 | <Uzzi> k
| |
| 04:05 | litlebuda has joined #ltsp | |
| 04:14 | <Uzzi> alkisg, now rootserver: ok rootpah: and filename: are emppy
| |
| 04:14 | empty*
| |
| 04:17 | <alkisg> Uzzi: what script are you currently using? If you set a filename, it woulnd't be empty...
| |
| 04:19 | <Uzzi> /ltsp/i386/vmlinuz there isn't in my server
| |
| 04:20 | <alkisg> AFAIK it should be in /var/lib/tftpboot/ltsp/i386/vmlinuz
| |
| 04:20 | I.e. that's relative to your tftp root path
| |
| 04:21 | (or /srv/tftp, I think squeeze switched to that)
| |
| 04:54 | daya has quit IRC | |
| 04:55 | daya has joined #ltsp | |
| 05:05 | gorkhaan has quit IRC | |
| 05:14 | gorkhaan has joined #ltsp | |
| 05:24 | daya has quit IRC | |
| 05:28 | ogra_ac has joined #ltsp | |
| 05:33 | artista_frustrad has joined #ltsp | |
| 05:47 | zz_evil_root is now known as evil_root | |
| 05:53 | evil_root is now known as zz_evil_root | |
| 06:03 | [GuS] has joined #ltsp | |
| 06:06 | ogra_ac has quit IRC | |
| 06:21 | ogra_ac has joined #ltsp | |
| 06:28 | bobby_C has quit IRC | |
| 06:35 | bobby_C has joined #ltsp | |
| 06:57 | <Trixboxer> alkisg: hi
| |
| 06:57 | hi all
| |
| 06:57 | <alkisg> Hello
| |
| 07:30 | <elias_a> alkisg: I am writing to fsfe people about the FLOSS school map thing.
| |
| 07:31 | <alkisg> Nice, I hope they gather all the maps in a page with links for each country
| |
| 07:32 | <elias_a> Please give me the URL. What do you think about the importance of having the information in both local language and english (or esperanto if english is not acceptable for some countries)?
| |
| 07:32 | <alkisg> http://goo.gl/maps/nOoQ
| |
| 07:33 | <elias_a> alkisg: Thank you!
| |
| 07:33 | <alkisg> Hmmm not sure about having the information in 2 languages...
| |
| 07:34 | At least our map is not maintained by anyone, and I'm not sure if local teachers would take the time to submit their data in 2 langs
| |
| 07:34 | So I think that either in english or in local lang it would be ok, but I'm not sure about having both of them
| |
| 07:35 | <elias_a> alkisg: OK. In the case of Finland I may bear the responsability of translating the stuff to english. My handicap is that I do not read greek :(
| |
| 07:36 | <alkisg> I can ask someone if he's willing to do the greek translation
| |
| 07:36 | <Hyperbyte> I've never ever used Dutch resources on the internet, unless it's about a subject where there's only information available in Dutch. :|
| |
| 07:36 | <alkisg> Where would the translation show up? in the same map?
| |
| 07:37 | <elias_a> Maybe the best approach would be to just combine the 2 national maps we have and toss the ball further to see what happens...
| |
| 07:37 | <alkisg> elias_a: right, that's what I thought too. And if many more emerge, then we can come up with a better plan
| |
| 07:37 | <elias_a> Hyperbyte: Dutch legislation would be a good example on that... :)
| |
| 07:37 | <alkisg> The data will be obsolete every few years anyway and will need to be updated ;)
| |
| 07:42 | <elias_a> Umm... how does one change the language of google maps UI? I would like to export the Greek map in order to combine it with the Finnish map.
| |
| 07:45 | <Hyperbyte> Don't you just change the domain name?
| |
| 07:45 | I.e. http://www.google.nl/search?q=ltsp or http://www.google.fi/search?q=ltsp ?
| |
| 07:48 | Oh, no, apparently not. :\
| |
| 07:50 | Ah, you add &hl=en
| |
| 07:50 | <elias_a> Hyperbyte: Cheers - did not even think about that :)
| |
| 08:12 | ogra_ac has quit IRC | |
| 08:23 | ogra_ac has joined #ltsp | |
| 08:23 | alkisg has quit IRC | |
| 08:31 | ogra has quit IRC | |
| 08:32 | ogra has joined #ltsp | |
| 08:38 | vagrantc has joined #ltsp | |
| 08:44 | alkisg has joined #ltsp | |
| 08:44 | Gadi has joined #ltsp | |
| 08:45 | Uzzi has quit IRC | |
| 08:48 | mgariepy has joined #ltsp | |
| 08:51 | shawnp0wers has joined #ltsp | |
| 08:52 | alkisg has quit IRC | |
| 08:54 | ogra_ac has quit IRC | |
| 08:56 | Da-Geek has joined #ltsp | |
| 09:10 | mgariepy has quit IRC | |
| 09:24 | cyberorg has joined #ltsp | |
| 09:32 | Pweg has joined #ltsp | |
| 09:43 | <sbalneav> Morning all
| |
| 09:54 | Da-Geek has quit IRC | |
| 09:55 | <Hyperbyte> Hi Scotty. :-)
| |
| 10:06 | <chupacabra> hidy
| |
| 10:08 | gnunux has quit IRC | |
| 10:19 | mgariepy has joined #ltsp | |
| 10:24 | komunista has joined #ltsp | |
| 10:34 | * vagrantc waves | |
| 10:36 | * ogra shores | |
| 10:43 | * vagrantc erodes | |
| 10:52 | dobber has quit IRC | |
| 10:55 | staffencasa has joined #ltsp | |
| 11:01 | alkisg has joined #ltsp | |
| 11:07 | chupacabra has quit IRC | |
| 11:09 | chupacabra has joined #ltsp | |
| 11:14 | <sbalneav> So, as of rev. 25 of pam_sshauth, I've now got looping for incorrect passwords, and messages spitting back via the pam mechanism for incorrect passwords. So no more restarting the session entirely on a failed password.
| |
| 11:14 | <abeehc> awesome stuff thank you
| |
| 11:15 | <sbalneav> At this point, even though the pam module's not fully complete, there's enough there to actually start playing around with a different display manager.
| |
| 11:16 | vagrantc, stgraber, highvoltage, alkisg, et al., do any of you have time this weekend for a "mini hack session" to see if we can come up with a branch of LTSP that substitutes a different DM + the pam module for ldm?
| |
| 11:17 | I doubt we'll get everything working, but it'd be neat to see if we could at least get it logging in.
| |
| 11:18 | Gadi as well.
| |
| 11:19 | <alkisg> sbalneav: since I missed the BTS, I'm in, anytime except for sunday morning (utc+2)
| |
| 11:20 | <sbalneav> Here's a paste of the pamtest:
| |
| 11:20 | <alkisg> Getting the pam module working would be a major progress ;)
| |
| 11:21 | <sbalneav> sbalneav@feniks:~/Source/bzr/libpam-sshauth/src$ ./pamtest flarp sbalneav
| |
| 11:21 | Trying to authenticate <sbalneav> for service <flarp>.
| |
| 11:21 | Password:
| |
| 11:21 | Received SSH_MSG_DISCONNECT: Too many authentication failures for sbalneav
| |
| 11:21 | Password:
| |
| 11:21 | Authentication successful.
| |
| 11:21 | open session successful.
| |
| 11:21 | Env PAM_SSHAUTH_SOCKET = /tmp/pam_sshauth_28039_oin.ath.cx
| |
| 11:21 | Press any key to continue
| |
| 11:21 | Exit request sent.
| |
| 11:21 | close session successful.
| |
| 11:21 | Press any key to continue
| |
| 11:21 | sbalneav@feniks:~/Source/bzr/libpam-sshauth/src$
| |
| 11:21 | <alkisg> Too many authentication failures for sbalneav ?
| |
| 11:21 | <sbalneav> So, in the pam client (i.e. gdm) in the message area, when you typed in an incorrect password, you'd get the "too many..." message.
| |
| 11:21 | that was me mistyping my password.
| |
| 11:22 | that's the ssh message out of sshlib
| |
| 11:22 | <alkisg> Ah, ok (more weird messages :D)
| |
| 11:22 | <sbalneav> later on, we can put in something more friendly, like "password incorrect"
| |
| 11:22 | but for now, I'm returning the ssh error messages directly, for debugging purposes.
| |
| 11:23 | <alkisg> How's the env handled? The pam module outputs the socket var to the environment? Or it takes that as input?
| |
| 11:24 | <sbalneav> exports. One sec, I'll show you...
| |
| 11:27 | So, it sets the PAM_SSHAUTH_SOCKET environment variable, and, from ps:
| |
| 11:27 | sbalneav 28123 1 0 11:25 pts/10 00:00:00 ssh -Y -t -N -qq -M -S /tmp/pam_sshauth_28118_oin.ath.cx -l sbalneav -p 22 oin.ath.cx
| |
| 11:27 | sbalneav@feniks:~$ ssh -S /tmp/pam_sshauth_28118_oin.ath.cx oin.ath.cx date
| |
| 11:27 | Fri Nov 19 11:27:32 CST 2010
| |
| 11:27 | So the socket's functioning.
| |
| 11:28 | <alkisg> So LDM can either use the socket to execute "echo $PAM_SSHAUTH_SOCKET", or check `ps`
| |
| 11:28 | Sorry silly
| |
| 11:28 | LDM can't use the socket to find out the socket name :D
| |
| 11:28 | <sbalneav> yeah, well, what we'd probably want to is just set LDM_SOCKET=${PAM_SSHAUTH_SOCKET"
| |
| 11:28 | then most of our scripts would remain the same.
| |
| 11:29 | <alkisg> And will you do an initial nss implementation this weekend, or later on? I mean, can ldm work without nss?
| |
| 11:30 | Hmm the localapps user copying code would need to be changed once this is implemented...
| |
| 11:30 | <sbalneav> Well, the NSS implementation's a bit of work, but I should be able to bang something out *fairly* quickly.
| |
| 11:31 | Well, you wouldn't need much in the way of user-copying code.
| |
| 11:31 | <vagrantc> sbalneav: LTSPROCKS!
| |
| 11:31 | <alkisg> With NSS, user copying wouldn't be needed at all, right?
| |
| 11:31 | * alkisg votes to change that to SBALNEAVROCKS :D | |
| 11:33 | <alkisg> If that pam module could somehow be used by gdm, then I imagine many standalone labs wouldn't need LDAP any longer...
| |
| 11:34 | It'd be much easier to implement the "school server" package that was mentioned in the edubuntu ML
| |
| 11:36 | <vagrantc> sbalneav: this is probably the most exciting development in LTSP in the last year :)
| |
| 11:37 | sbalneav: i can try it some today, but tomorrow i'll be working and such.
| |
| 11:37 | sbalneav: might have some time sunday
| |
| 11:38 | i haven't even tried the newer ldm-trunk with all the crazy new changes
| |
| 11:38 | i really ought to try and at least get that to compile
| |
| 11:40 | <sbalneav> alkisg: No, shouldn't be. Once the socket's been established, user geometry should "just work"
| |
| 11:43 | Well, it'd be interesting to sit down for 3 or 4 hours and actually sub in that new TinyDM or whatever it was called that "everyone" is going to switch over too.
| |
| 11:43 | <Appiah> oops
| |
| 11:43 | 7 mails from ltsp-discuss
| |
| 11:43 | <sbalneav> once we get a "full" pam & nss implementation, then things get much nicer.
| |
| 11:43 | <Appiah> just now
| |
| 11:43 | 8
| |
| 11:44 | <sbalneav> I'll post somethig to -develop
| |
| 11:44 | <Appiah> wonder if they all got stopped somewhere
| |
| 11:44 | and sent at the same time..
| |
| 11:53 | <sbalneav> ogra: ping
| |
| 11:54 | What was that DM again? tinydm?
| |
| 11:54 | <alkisg> Appiah: there's html messages, they're help back until a list admin approves them
| |
| 11:55 | bah that didn't make sense
| |
| 11:55 | *they are html messages, they are held back... etc
| |
| 11:55 | <Appiah> so all these mails were sent in html? :)
| |
| 11:55 | <alkisg> Yup, usually from outlook or something similar
| |
| 11:55 | <vagrantc> sbalneav: slim?
| |
| 11:56 | <sbalneav> No, it was like "tinydm" or something like that
| |
| 11:57 | <alkisg> Why is a DM needed? E.g. is gdm using one?
| |
| 11:58 | erm dm or wm? :D
| |
| 11:58 | <sbalneav> gdm is a display manager
| |
| 11:58 | We could use gdm, too.
| |
| 11:58 | <vagrantc> GNOME Display Manager
| |
| 11:58 | <sbalneav> Basically, any DM which is pam aware, should work.
| |
| 11:58 | <alkisg> Ah, for testing, until we implement that in ldm... got it
| |
| 11:59 | <vagrantc> sbalneav: so are you proposing to hack on this without ldm at all?
| |
| 11:59 | well, for starters, it looks like slim is pam-aware, and it's already in debian
| |
| 12:00 | (it at least depends on libpam0g)
| |
| 12:00 | if that's all it would take
| |
| 12:04 | gentgeen__ has joined #ltsp | |
| 12:06 | * alkisg got mixed up from an older talk about ldm using a minimal wm... | |
| 12:12 | gentgeen__ has quit IRC | |
| 12:44 | alkisg has quit IRC | |
| 12:55 | <vagrantc> i'm scared of all these LDM changes...
| |
| 12:55 | they sure sound nice...
| |
| 12:57 | <Gadi> lightdm?
| |
| 13:11 | <sbalneav> lightdm
| |
| 13:11 | that was it.
| |
| 13:11 | Just sent a message to -develop
| |
| 13:11 | vagrantc: right.
| |
| 13:12 | Well, depending on how things go, and what we all think is the best route for us, we could *potentially* drop ldm alltogether.
| |
| 13:12 | just make sure we get the right bits and bobs in place to use an existing dm, and reduce our LOC significantly.
| |
| 13:13 | I dunno. We need to get everyone talking about this, Hopefully my email will spark some discussion
| |
| 13:14 | alkisg has joined #ltsp | |
| 13:17 | * Gadi thinks sbalneav always manages to spark discussion :) | |
| 13:19 | <vagrantc> it uses webkit for theming?
| |
| 13:19 | for a display manager?
| |
| 13:19 | that's weird.
| |
| 13:20 | * Gadi thinks vagrantc's an old foagie | |
| 13:20 | <sbalneav> From my point of view, I think getting our stuff in a format that it can be used with any dm would be a good move.
| |
| 13:21 | Gnome die-hards will want to use gdm. The KDM people have been grumbling at us for years. LXDE people will want their own thing, etc.
| |
| 13:21 | Trixboxer has quit IRC | |
| 13:21 | <sbalneav> by going this route, we can keep 'em all happy.
| |
| 13:21 | <vagrantc> hear hear
| |
| 13:21 | <sbalneav> And, then we don't have to worry about a dm and providing theming, etc.
| |
| 13:22 | <Gadi> which keeps us happy
| |
| 13:22 | :D
| |
| 13:22 | <vagrantc> tell it like it should be!
| |
| 13:22 | <alkisg> So why is ldm currently being rewritten? :D
| |
| 13:22 | <vagrantc> because in the meantime...
| |
| 13:22 | * Gadi loves how an LTSP dev's entire focus is to lessen his workload | |
| 13:22 | <Gadi> :D
| |
| 13:22 | <sbalneav> Well, I was afk for quite some time, with my work problems.
| |
| 13:23 | and during that time, pam_sshauth was stalled.
| |
| 13:23 | <vagrantc> but there's a glimmer of hope now
| |
| 13:23 | <Gadi> alkisg: ldm's rewrite still has merit
| |
| 13:23 | <sbalneav> And there's nothing to say we can't ALSO keep ldm
| |
| 13:23 | <Gadi> most of it has been backend rewrite
| |
| 13:23 | another route (other than using gdm, et al) is to use lightdm as the FRONT end
| |
| 13:24 | and LDM 3.0's "plugin" system as the back end
| |
| 13:24 | <vagrantc> the key will be getting other display managers to have all the hooks we want
| |
| 13:24 | <sbalneav> right, that was what we talked about at the hackfest.
| |
| 13:24 | <Gadi> which would make LDM the display manager of choice for integration with ltsp-cluster
| |
| 13:24 | <vagrantc> so $ANYDM logs into LDM and thus we have all of LDM's hooks?
| |
| 13:25 | <Gadi> not exactly
| |
| 13:25 | its not about the hooks
| |
| 13:25 | the rev. linux guys have been writing a plugin system for ldm
| |
| 13:25 | <vagrantc> right
| |
| 13:25 | <Gadi> so ldm would be the frontend not just to an X-over-ssh session
| |
| 13:25 | but also the front end to rdp, Citrix, NX, etc
| |
| 13:26 | which allows them to collect credentials and make decisions before launching the session
| |
| 13:26 | and by make decisions, I mean a control center making decisions
| |
| 13:26 | like, hey - go to this least-loaded server
| |
| 13:26 | <vagrantc> sounds tasty
| |
| 13:27 | <Gadi> or, hey you're "bob"? you can only get an RDP session with these restrictions
| |
| 13:27 | <alkisg> And if sbalneav was here at that point, he'd implement all those with pam modules so that they could be used from any DM :P
| |
| 13:27 | <Gadi> alkisg: exactly
| |
| 13:28 | so, all of this work may be in different directions, but can be leveraged
| |
| 13:30 | * vagrantc eagrely awaits instructions | |
| 13:31 | <Gadi> vagrantc: scotty already sent his email
| |
| 13:32 | I guess the next step is to start to butcher a chroot
| |
| 13:32 | :)
| |
| 13:33 | * vagrantc grabs a butcher's knife | |
| 13:34 | abeehc has quit IRC | |
| 13:35 | * vagrantc reads | |
| 13:37 | <mistik1> did I see something about workload?
| |
| 13:38 | Gadi: what else should a programmer do?
| |
| 13:39 | <stgraber> sbalneav: I'm still against getting rid of ALL of ldm. I still believe the backend is very useful especially as there's no plan to support citrix/NX/RDP/... in a regular DM
| |
| 13:39 | <sbalneav> stgraber: sure.
| |
| 13:39 | <stgraber> sbalneav: though I'd love to get rid of the greeter and have lightdm's greeter be used in place
| |
| 13:39 | sbalneav: so our backend would support rdp, citrix, nx, ... and PAM. With PAM being used with whatever backend we want (by default, your ssh implementation)
| |
| 13:40 | that way we keep some of our cool stuff (rc.d, autologin, load-balancing, ...) and get rid of what's currently sucky in ldm (as in, the greeter)
| |
| 13:41 | vagrantc: the current ldm-trunk is "clean" by "my" standards. I tried building from lucid to natty and fixed all issues I saw. GCC doesn't report any error and I'd only expect issues with quite old system (python < 2.4 and old gtk)
| |
| 13:42 | vagrantc: if it fails to build on Debian, just le me know and send me a build log so I can try to fix that
| |
| 13:44 | Gadi: you mean like LDM_FORCE_BACKEND=rdesktop ? that already exists today ;)
| |
| 13:44 | <Gadi> ah, you mean if you are bob?
| |
| 13:44 | :)
| |
| 13:44 | poor bob
| |
| 13:44 | forced to use windows
| |
| 13:45 | <stgraber> not yet but you can update the initialization code of the RDP plugin to reject bob ;)
| |
| 13:45 | <Gadi> now bob's rejected?
| |
| 13:45 | poor poor bob
| |
| 13:46 | <stgraber> highvoltage will be working on the python daemon (ltsp-agent) next week with the goal of having it get config from either lts.conf (over tftp) or from our control center (over https) and offer a unified interface to the data
| |
| 13:46 | my goal (as in, next 6 months or so) is to get rid of getltscfg and getltscfg-cluster. Replacing that by the daemon and plugins
| |
| 13:46 | with the added benefit of caching and auto-update of the config (no more reboot)
| |
| 13:47 | then, if ldm is properly integrated, it'll be super easy to restrict session type, cluster groups, ... based on the username (or any other criteria)
| |
| 13:49 | that also means, no more different codepath when using ltsp-cluster. LTSP-Cluster will eventually just be upstream LTSP with a web GUI instead of a text file (lts.conf) ;)
| |
| 13:49 | after that, we'll really need to release ltsp 6 and ldm 3 ;)
| |
| 13:57 | [GuS] has quit IRC | |
| 13:57 | <vagrantc> stgraber: any new build-deps?
| |
| 13:57 | stgraber: last time i tried to build it seemed like something was missing
| |
| 13:57 | <alkisg> A web-served lts.conf would be handy for non-cluster setups too ;)
| |
| 13:57 | shawnp0wers has quit IRC | |
| 13:59 | <vagrantc> alkisg: that's what i hear...
| |
| 14:01 | <stgraber> alkisg: that'd be extremely easy to add to the daemon. Initially I just want to merge our current two use cases (regular ltsp with tftped lts.conf and ltsp-cluster with it's php-rendered lts.conf)
| |
| 14:01 | alkisg: but I'll make sure we can add other ways of fetching it as well
| |
| 14:02 | alkisg: so you could for example have a "usb" module where a lts.conf on a usb stick would have override the one on the tftp server ;)
| |
| 14:02 | vagrantc: I just checked my source package and I don't see any new build-dep
| |
| 14:03 | vagrantc: I'm building with: debhelper (>= 7), flex, bison, libpopt-dev, libgtk2.0-dev, dpatch, intltool, python-lxml, iso-codes, locales
| |
| 14:05 | Gadi has left #ltsp | |
| 14:12 | chupacabra has quit IRC | |
| 14:16 | <vagrantc> stgraber: i've got: debhelper (>= 7), flex, bison, libpopt-dev, libgtk2.0-dev, intltool, iso-codes, locales, python
| |
| 14:17 | <stgraber> should work fine, dpatch isn't used at the moment in Ubuntu and I don't think python-lxml is actually that useful (other than bringing python ;))
| |
| 14:17 | <vagrantc> switch to 3.0 quilt format instead of dpatch ... so that pretty much leaves python-lxml
| |
| 14:17 | <stgraber> yeah, I still need to merge your packaging one day ;)
| |
| 14:18 | I've started using 3.0 quilt for a few packages (all the new ltsp-cluster packages)
| |
| 14:19 | * vagrantc is happy with it | |
| 14:19 | <vagrantc> there are some things that are just infinitely easier
| |
| 14:19 | <stgraber> yeah, packaging python with it is way too easy now ;) and auto generated patches are quite cool
| |
| 14:19 | <vagrantc> i've switched all of the ltsp related packages to it ... and most any others i get my hands on
| |
| 14:19 | <stgraber> not to mention that binary diffs are now possible ;)
| |
| 14:20 | will be useful for a few packages where I have to ship icons in debian/
| |
| 14:23 | * alkisg hasn't yet managed to properly mix python packaging (setup.py) with debhelper 7 packaging... | |
| 14:26 | mistik1 has quit IRC | |
| 14:31 | mistik1 has joined #ltsp | |
| 14:33 | alkisg has quit IRC | |
| 14:37 | <sbalneav> So, an nss module looks like it'll be ffffffaaaairly simple.
| |
| 14:38 | Use the TAILQ macros for building linked lists.
| |
| 14:39 | a _setdbent_r() call will do a "getent <db>" call over the ssh tunnel, read in the list of results, and creat a linked list of all the resultant lines.
| |
| 14:40 | a call to _getent_r() will pop the head item off the queue parse the text command line, and break the fields up into the structure (ie. pwent or grent)
| |
| 14:40 | a _enddbent_r() will clear out the linked list.
| |
| 14:41 | and a _getdbnam_r() will simply do a getent db nam over the ssh link, and feed that into the structure.
| |
| 14:41 | duplicate for the passwd, group, and (possibly) shadow files, and voila
| |
| 14:41 | <vagrantc> stgraber: ok, it built sucessfully
| |
| 14:42 | <stgraber> vagrantc: cool
| |
| 14:42 | sbalneav: seems easy
| |
| 14:42 | <vagrantc> think it's probably time we drop the /usr/lib -> /usr/share symlinks?
| |
| 14:43 | <stgraber> yeah, I guess we've shipped these for long enough now ;)
| |
| 14:43 | people should have transitioned their scripts now
| |
| 15:02 | AJMC has joined #ltsp | |
| 15:07 | <AJMC> OK, Friday afternoon / night is probably not the best time to be asking questions here...
| |
| 15:08 | <vagrantc> but still a good idea to try :)
| |
| 15:08 | <AJMC> OK :) can anyone give me tips on the best way to add launchers to all users gnome panels?
| |
| 15:08 | I mean, is there a nice easy graphical way, or do I have to wade through the config files?
| |
| 15:09 | I've done things that way for XFCE, but hoped there'd be something a little more straightforward for gnome
| |
| 15:14 | johnny has left #ltsp | |
| 15:14 | johnny has joined #ltsp | |
| 15:18 | <vagrantc> stgraber: without much hackery, it appears to be working
| |
| 15:20 | at least the basic functionality
| |
| 15:27 | and even connects to xrdp simply :)
| |
| 15:31 | gorkhaan has quit IRC | |
| 15:41 | mgariepy has quit IRC | |
| 15:44 | <Pweg> would anybody be able to point me in the right direction for figuring out why when I try to use vncviewer to remote my thin clients I get: ConnectToTcpAddr: connect: Connection refused?
| |
| 15:54 | <vagrantc> Pweg: are you using the server or thin client's ip address?
| |
| 15:55 | <Pweg> vagrantc: client's ip address, so 192.168.0.22, 254 being the server
| |
| 15:57 | <vagrantc> Pweg: where is vncviewer started from?
| |
| 15:57 | Pweg: er, rather ... what are you using to start a vnc server? and where is that started from?
| |
| 15:58 | <Pweg> vagrantc: I'm trying to do it from my LTSP server, or at least try to get the Thin Client Manager to be able to use the Share Screen option (sorry, I'm really new at this)
| |
| 16:00 | * vagrantc wonders if thin client manager is still really buggy | |
| 16:00 | <Pweg> That's what I've been reading. =\ I'm using Karmic, all the other services seem to work, just the one doesn't.
| |
| 16:05 | AJMC has left #ltsp | |
| 16:36 | <Pweg> Ah, figured it out, had to use the proper port for that client. Thanks for your help vagrantc
| |
| 16:42 | gorkhaan has joined #ltsp | |
| 16:59 | shawnp0wers1 has joined #ltsp | |
| 17:13 | johnny has left #ltsp | |
| 17:27 | BWMerlin has joined #ltsp | |
| 17:29 | Pweg has left #ltsp | |
| 17:50 | komunista has quit IRC | |
| 18:02 | staffencasa has quit IRC | |
| 18:07 | Selveste1 has quit IRC | |
| 18:34 | Selveste1 has joined #ltsp | |
| 18:50 | vagrantc has quit IRC | |
| 19:10 | BWMerlin has quit IRC | |
| 19:44 | chupacabra has joined #ltsp | |
| 20:01 | abeehc has joined #ltsp | |
| 20:03 | johnny has joined #ltsp | |
| 20:18 | BWMerlin has joined #ltsp | |
| 20:23 | bobby_C has quit IRC | |
| 21:09 | litlebuda has joined #ltsp | |
| 21:13 | litlebuda has quit IRC | |
| 22:51 | sweetpi has joined #ltsp | |
| 22:53 | alkisg has joined #ltsp | |
| 23:00 | spectra has quit IRC | |
| 23:03 | spectra has joined #ltsp | |