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


Channel log from 18 February 2013   (all times are UTC)

00:03alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.12.47, Read error: Operation timed out)
00:03alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.12.47)
00:34Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 252 seconds)
00:51Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
01:00komunista has left IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk, Read error: Connection reset by peer)
01:00komunista has joined IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk)
01:13
<warren>
knipwim: no. we will not name it based on lsb_release -is
01:14
knipwim: that is an arbitrary string that has different values on our variants of the OS
01:59alexqwesa__ has joined IRC (alexqwesa__!~alex@109.172.12.47)
02:02alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.12.47, Ping timeout: 244 seconds)
03:02komunista has left IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk, Quit: Leaving.)
03:08adrianorg_ has left IRC (adrianorg_!~adrianorg@187.113.247.94, Ping timeout: 256 seconds)
03:26Parker955_Away is now known as Parker955
04:02jammcq has left IRC (jammcq!~jam@c-69-245-75-255.hsd1.mi.comcast.net, Quit: leaving)
04:17alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
04:22wwww has joined IRC (wwww!~chatzilla@60.217.209.97)
04:22
<wwww>
hi
04:22
<alkisg>
Hello
04:23
<wwww>
if can i have a try to look at this product
04:23
a demo site
04:23
<alkisg>
Demo site for netbooting?
04:24
You can download the edubuntu DVD, it includes LTSP so you can test it without installation
04:24
And I think the OpenSUSE DVD also supports directly netbooting with it
04:28
warren, Enslaver: so how many dirs should be there upstream in ltsp-build-client/plugins and in init-ltsp.d/ etc etc for redhat based distros, and how to determine them?
04:29
<wwww>
ltsp thin client only support *nux OS?
04:29
<alkisg>
wwww: it's about netbooting linux, yeah. After it boots, you can e.g. run a windows VM or connect to a windows terminal server or whatever else you can do with a linux client
04:31
<warren>
alkisg: Enslaver is combining them into one directory, just lsb_release is not how it is named
04:31
<alkisg>
warren: so we'd have to special-case it in the code? Make a distro-overridable function for it?
04:32
<Enslaver>
How does ubuntu do it? Dont they have Edbuntu / Zedbuntu and all that?
04:33
<alkisg>
I think they're all Ubuntu's
04:33
<Enslaver>
all of them return Ubuntu on lsb_release -is?
04:33
<alkisg>
If something is so different that it deserves a different DISTRIB_ID, then it should at least get a symlink for the plugins, if not a separate directory...
04:33
<warren>
alkisg: did something change recently? if it relies on lsb_release -is now, it will have totally different results on different RH-based distros here and we have to override it
04:33
<alkisg>
Let me check some of them
04:34
<warren>
Enslaver: I think my old RPM didn't care about it because it only packaged the plugins we actually use.
04:34
<alkisg>
warren: we stopped having distro-specific "ifs" in the common code, and are using distro overridable functions whenever that is needed
04:34
So, how will the new dir be named? RedHat? el6?
04:34
<Enslaver>
warren: I think you checked for rpm.macro file and got the el6 from there, then did a case statment
04:34
<warren>
alkisg: that's fine, choosing which plugins should be at build time, not runtime
04:34
<alkisg>
And will that change with e.g. el7?
04:35
<warren>
alkisg: the dir should be named just Fedora, to cover all current and future Fedora and EL
04:35
alkisg: the output of lsb_release -is is meaningless here
04:35
<alkisg>
warren: ltsp-build-client requires run time, the others are already determined build-time
04:35
OK, so a function, np
04:36
The generic code will have lsb_release -is, and fedora can override it with plain "echo Fedora"
04:36
<Enslaver>
thats what I'm doing currently for check_vendor
04:37
<warren>
Enslaver: whatever they do upstream, just have the .spec package the plugins it actually uses, which may differ between Fedora 11, 18 and EL
04:37
<alkisg>
Enslaver: for the packaging, are you using the new "copy distro specific dirs over common dirs" method?
04:37
<Enslaver>
yes
04:37
<alkisg>
Nice
04:37
<Enslaver>
copies common first then distro specific
04:38
<wwww>
hp thin client has its own OS, after boot up ,user can access remote desktop from server?
04:38
<alkisg>
We do that in 3 rounds for ubuntu... common => debian => ubuntu
04:38
wwww: I think that also was linux-based. Yeah the same can be done with ltsp too, there is rdesktop and xfreerdp in linux.
04:39
warren: the idea is for ltsp-build-client --dist=xxx to work on different distros, if at all possible
04:40
I know vagrantc is using it to build ubuntu chroots on debian servers, dunno if that would work on e.g. building fedora chroots on rhel servers etc...
04:41
Of course if there's no chance of it working on some distro, there's no need to include the ltsp-build-client plugin dirs for other distros
04:41
That was arranged that way before I joined ltsp...
04:42
<warren>
Fedora can build EL6 chroots. EL6 probably can't build Fedora chroots because the database format changed
04:42
it's generally backwards but not forward compatible
04:43
Enslaver: there's ugly hacks to bootstrap a chroot with an incompatible rpmdb, but it really isn't worth supporting
04:43
<wwww>
win7 installed in hp client is same with win7 authority?
04:43
<Enslaver>
i was able to build a fc18 chroot using el6
04:43
<alkisg>
wwww: no idea what that is
04:44
<wwww>
i care if all exe application can be installed on hp client
04:44
<Enslaver>
using mock
04:45
doubt yum would do it
04:46
<alkisg>
wwww: we don't support win7 here. LTSP is ***Linux*** Terminal Server Project.
04:47
<wwww>
hp client can use win7 as OS
04:47
<alkisg>
wwww: nice, please ask HP and Microsoft for support.
04:47
This IRC channel is for using Linux as OS.
04:51
<warren>
Enslaver: oh right, the rpmdb change happened between EL5 and EL6
04:53alexqwesa__ has left IRC (alexqwesa__!~alex@109.172.12.47, Ping timeout: 276 seconds)
04:55* alkisg found the redhat family tree that he was looking for... :D http://upload.wikimedia.org/wikipedia/commons/a/a3/Redhat_family_tree_11-06.png
04:55sha_ has joined IRC (sha_!~sha@e177118215.adsl.alicedsl.de)
04:56sha has left IRC (sha!~sha@e177162159.adsl.alicedsl.de, Read error: Operation timed out)
05:03Parker955 is now known as Parker955_Away
05:05vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc)
05:10
<wwww>
it seems ltsp is similar with QVD
05:12
<warren>
alkisg: most of those distros are dead or niche-purpose
05:12
<alkisg>
warren: basically I never looked into rh / rhel / el* / centos / fedora / etc etc naming and history
05:12
So I wanted to organize those in my mind a bit
05:14
wwww: QVD says that it's about Linux VDI, while LTSP is about netbooting Linux. After Linux netboots, you can use whatever Linux can use, rdesktop to connect to windows, virtualbox for VMs, whatever
05:14
They are different technologies
05:14
<wwww>
i see
05:15
<alkisg>
In short, LTSP == booting linux from local network without using a hard disk
05:15
After it boots, you can do whatever you like with it
05:15
So the "client" might only see a windows interface
05:16
<wwww>
after load kernel, it will mount a nfs diretory as root /?
05:16
<alkisg>
NFS or NBD
05:20alexqwesa has joined IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net)
05:23alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
05:27Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 245 seconds)
05:47Parker955_Away is now known as Parker955
06:25vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi)
06:29work_alkisg is now known as alkisg
06:33khildin has joined IRC (khildin!~khildin@ip-80-236-222-143.dsl.scarlet.be)
06:38alkisg has left IRC (alkisg!~alkisg@plinet.ioa.sch.gr, Quit: Leaving.)
06:41alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.12.47)
06:41alexqwesa has left IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net, Read error: Connection reset by peer)
07:15DIoX|DaZ has left IRC (DIoX|DaZ!~KaKa@server.civicclub.lt)
07:41work_alkisg has joined IRC (work_alkisg!~alkisg@plinet.ioa.sch.gr)
07:45work_alkisg is now known as alkisg
07:54Parker955 is now known as Parker955_Away
07:56vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving)
08:06dievel has joined IRC (dievel!~dievel@2-229-104-66.ip196.fastwebnet.it)
08:21Gremble has joined IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com)
08:29kyriazis has joined IRC (kyriazis!51ba2f97@gateway/web/freenode/ip.81.186.47.151)
08:30
<kyriazis>
Καλημέρα. ΠÏŽς καθαρίζω τη λίστα με τις εντολές που έχω δÏŽσει στους επÏŒπτες;
08:33
<knipwim>
omg, it's the greeks :)
08:33
<kyriazis>
Should I write in english?
08:34
<muppis>
On this channel yes. If you prefer greece, you shoudl go to #ts.sch.gr
08:34
*should
08:34
<kyriazis>
How do I clear the history of commands I have executed in the epoptes application?
08:35
<alkisg>
kyriazis: hello
08:36
kyriazis: type /j #ts.sch.gr
08:36
This one is the english channel, that's the greek one
08:38
!learn greek as Στο παρÏŒν κανάλι μιλάνε μÏŒνο Αγγλικά, για υποστήριξη στα Ελληνικά απÏŒ την υπηρεσία Τεχνικής Στήριξης ΣΕΠΕΗΥ διαβάστε το http://ts.sch.gr/wiki/IRC και στη συνέχεια πληκτρολογήστε /j #ts.sch.gr
08:38
<ltsp>
The operation succeeded.
08:39meamy has joined IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de)
08:41orestis12 has joined IRC (orestis12!51ba4741@gateway/web/freenode/ip.81.186.71.65)
08:41orestis12 has left IRC (orestis12!51ba4741@gateway/web/freenode/ip.81.186.71.65)
08:45
<ogra_>
looks greek to me ...
08:46kyriazis has left IRC (kyriazis!51ba2f97@gateway/web/freenode/ip.81.186.47.151)
08:47JesseCW has joined IRC (JesseCW!~JesseCWor@wsip-98-175-20-126.br.br.cox.net)
08:47JesseC is now known as Guest6582
08:51orestis12 has joined IRC (orestis12!51ba4741@gateway/web/freenode/ip.81.186.71.65)
08:54vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 256 seconds)
08:56
<orestis12>
alkisg:hi
08:56
<alkisg>
!greek
08:56
<ltsp>
greek: Στο παρÏŒν κανάλι μιλάνε μÏŒνο Αγγλικά, για υποστήριξη στα Ελληνικά απÏŒ την υπηρεσία Τεχνικής Στήριξης ΣΕΠΕΗΥ διαβάστε το http://ts.sch.gr/wiki/IRC και στη συνέχεια πληκτρολογήστε /j #ts.sch.gr
09:00orestis12 has left IRC (orestis12!51ba4741@gateway/web/freenode/ip.81.186.71.65, Ping timeout: 245 seconds)
09:20JesseCW has left IRC (JesseCW!~JesseCWor@wsip-98-175-20-126.br.br.cox.net, Ping timeout: 245 seconds)
09:21JesseCW has joined IRC (JesseCW!~JesseCWor@wsip-98-175-20-126.br.br.cox.net)
09:23Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
09:37alkisg has left IRC (alkisg!~alkisg@plinet.ioa.sch.gr, Quit: Leaving.)
09:39work_alkisg has joined IRC (work_alkisg!~alkisg@plinet.ioa.sch.gr)
09:48bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
09:57vmlintu has joined IRC (vmlintu!~vmlintu@cs180030.pp.htv.fi)
10:14Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 248 seconds)
10:14khildin_ has joined IRC (khildin_!~khildin@ip-80-236-222-115.dsl.scarlet.be)
10:17alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
10:18khildin has left IRC (khildin!~khildin@ip-80-236-222-143.dsl.scarlet.be, Ping timeout: 248 seconds)
10:21alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Client Quit)
10:26ltspuser_28 has joined IRC (ltspuser_28!c31772ca@gateway/web/freenode/ip.195.23.114.202)
10:26
<ltspuser_28>
Hi
10:27alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
10:27
<ltspuser_28>
I want to have a thin client that opens only a web page...
10:27
it's possible?
10:31
<Gremble>
yep
10:32Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
10:32
<Gremble>
and there's many ways to do that ltspuser_28
10:34
<ltspuser_28>
how?
10:35
<Gremble>
have a look at the kiosk modes avaialable in various desktop enviroments
10:35
and wait for more responses here too, there where plans for something kiosky in LTSP itself IIRC
10:36
essentially you'd have auto-login to a locked down account
10:37
<alkisg>
There's a kiosk screen script in ltsp which you can enable, yeah
10:37
SCREEN_07=kiosk etc etc
10:38alexqwesa_ has left IRC (alexqwesa_!~alex@109.172.12.47, Read error: Operation timed out)
10:39alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.12.47)
10:42
<ltspuser_28>
I have tried kiosk but the terminal didn't boot...
10:43
I don't know if I'm it's a bad configuration or if the chroot need something
10:48
<alkisg>
!release
10:48
<ltsp>
release: please mention the linux distro and release you're using :)
10:48
<ltspuser_28>
ubuntu 12.04
10:48
the chroot is from ubuntu 12.04
10:49
<alkisg>
Probably misconfiguration then
10:49
It's recent enough, I think the kiosk fixes would be included there
11:02dobber_ has joined IRC (dobber_!~dobber@89.190.199.210)
11:15ltspuser_55 has joined IRC (ltspuser_55!1fc792fb@gateway/web/freenode/ip.31.199.146.251)
11:16ltspuser_25 has joined IRC (ltspuser_25!1fc792fb@gateway/web/freenode/ip.31.199.146.251)
11:19
<ltspuser_25>
hi
11:20ltspuser_28 has left IRC (ltspuser_28!c31772ca@gateway/web/freenode/ip.195.23.114.202, Ping timeout: 245 seconds)
11:23ltspuser_25 has left IRC (ltspuser_25!1fc792fb@gateway/web/freenode/ip.31.199.146.251, Ping timeout: 245 seconds)
11:39ltspuser_19 has joined IRC (ltspuser_19!c31772ca@gateway/web/freenode/ip.195.23.114.202)
11:41
<ltspuser_19>
is there any procedure to put kiosk working in a ubuntu 12.04 chroot?
12:00adrianorg has joined IRC (adrianorg!~adrianorg@177.204.76.244.dynamic.adsl.gvt.net.br)
12:01
<ltspuser_19>
I want to know what I need to install
12:01
on the chroot
12:04komunista has joined IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk)
12:17
<alkisg>
ltspuser_19: ltsp-build-client --extra-help
12:17
--kiosk a simple webkiosk mode
12:18
So you need the --kiosk parameter when building the chroot
12:18
And then, SCREEN_07=kiosk in lts.conf
12:31
<ltspuser_19>
ok
12:31
thank you alkisg
12:31ltspuser_19 has left IRC (ltspuser_19!c31772ca@gateway/web/freenode/ip.195.23.114.202, Quit: Page closed)
12:52Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
12:53komunista has left IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk, Quit: Leaving.)
13:23gvy has joined IRC (gvy!~mike@altlinux/developer/mike)
13:40adrianorg_ has joined IRC (adrianorg_!~adrianorg@187.113.245.66)
13:42adrianorg has left IRC (adrianorg!~adrianorg@177.204.76.244.dynamic.adsl.gvt.net.br, Ping timeout: 248 seconds)
13:48Gremble has left IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com, Quit: I Leave)
14:05
<meamy>
arg for pressh scripts the user credentials are stored in $USER an $PASS correct?
14:12jammcq has joined IRC (jammcq!~jam@c-69-245-75-255.hsd1.mi.comcast.net)
14:20wwww has left IRC (wwww!~chatzilla@60.217.209.97, Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130201065344])
14:28eghie has joined IRC (eghie!~michiel@ip4da2ead1.direct-adsl.nl)
14:31JesseCW is now known as JesseC
14:32meamy has left IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de, Ping timeout: 245 seconds)
14:36dead_inside has joined IRC (dead_inside!~dead_insi@76.75.3.174)
14:38igorek_RU has joined IRC (igorek_RU!5f34e021@gateway/web/freenode/ip.95.52.224.33)
14:42meamy has joined IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de)
14:46anunnaki has left IRC (anunnaki!~anunnaki@c-174-54-115-236.hsd1.pa.comcast.net, Ping timeout: 248 seconds)
14:47komunista has joined IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk)
14:50
<igorek_RU>
hi guys!
14:51
maybe anybody can help me...LTSP5 (NFS), Debian Wheezy. i need to connect usb printer to thin stations. My lts.conf before update works perfectly. Now, however, if you include the option PRINTER_0_DEVICE=/dev/usb/lp0 thin client will not boot. if left ostavlnye option enabled, the download is correct. If I loaded with disabled option PRINTER_0_DEVITSE, and then do "jetpipe /dev/usb/lp0 9100", the printer will work.
14:52
<alkisg>
igorek_RU:
14:52
https://bugs.launchpad.net/ltsp/+bug/996533
14:54
<jammcq>
alkisg: you fixed that bug back in August. if he's on wheezy, shouldn't he have to fixed version?
14:55
<alkisg>
I think wheezy has been frozen before that, but lets see...
14:55
$ grep jetpipe /opt/ltsp/wheezy/usr/share/ltsp/init-ltsp.d/50-jetpipe
14:55
/usr/sbin/jetpipe ${JETPIPE_ARGS} ${PRINTER_DEVICE} ${PORT}
14:55
No & there
14:56
<igorek_RU>
I downloaded the latest build wheezy, apparently this bug is not fixed in it! Thanks a lot! I will correct in itself!
14:57
<alkisg>
Btw let's also add a note in the hackathon to replace jetpipe with nc or socat, plus a small shell script
14:57
I think we can completely remove C from ltsp... :P :D
14:58
<jammcq>
hmm, doesn't jetpipe also take care of serial printers?
14:58
and I don't understand the desire to get rid of C
14:59
<alkisg>
There's no point in duplicating code
14:59
<igorek_RU>
I do not have lpt printers, in order to check this
14:59
<alkisg>
jetpipe is just an nc reimplementation
14:59
ldm => lightdm, cdpinger => not needed,
14:59
ltspfs => we need that
14:59
The smaller LTSP is, the less bugs it has and the more easy it is to maintain
15:18
<meamy>
Is it possible to get the user credentials to use them for example mount a network drive. I guess the way to go is a script in /usr/share/ldm/rc.d but i dont know in witch variable the credentials are stored in
15:18
?
15:19
<alkisg>
I don't think they are
15:20Gremble has joined IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com)
15:22
<meamy>
alkisg: so no way to grep the credentials while the startup (unfortunately i cant use ldap in this case because the setup dont support it)
15:29staffencasa has joined IRC (staffencasa!~staffenca@8-220.ptpg.oregonstate.edu)
15:34yalu has left IRC (yalu!~yalu@109.134.141.1, Ping timeout: 252 seconds)
15:36
<igorek_RU>
if I add "&" at the end of the line "/usr/sbin/jetpipe ${JETPIPE_ARGS} ${PRINTER_DEVICE} ${PORT}", the thin client boots fine and the printer works. But jetpipe then stops working (after a single print job) that is port 9100 is no longer available
15:36yalu has joined IRC (yalu!~yalu@201.128-66-87.adsl-dyn.isp.belgacom.be)
15:40eghie has left IRC (eghie!~michiel@ip4da2ead1.direct-adsl.nl, Quit: Ex-Chat)
15:41vmlintu has left IRC (vmlintu!~vmlintu@cs180030.pp.htv.fi, Ping timeout: 264 seconds)
15:43yalu has left IRC (yalu!~yalu@201.128-66-87.adsl-dyn.isp.belgacom.be, Read error: Connection reset by peer)
15:44yalu has joined IRC (yalu!~yalu@109.134.174.29)
15:46meamy has left IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de, Ping timeout: 264 seconds)
15:46meamy has joined IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de)
15:51Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
16:00highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage)
16:03
<knipwim>
Enslaver: i encounter this when booting with nfs from dracut
16:03
http://article.gmane.org/gmane.linux.kernel.initramfs/2553
16:03
did you solve this error?
16:04
<alkisg>
meamy: not unless you patch ldm (ssh.c, I think) to cache/reuse the credentials
16:04
igorek_RU: and if you start it after the client boots, it works without problems?
16:06
igorek_RU: if so, try `ps aux` on the client to see the jetpipe command line, and then completely remove the jetpipe call from init-ltsp.d, and then use this:
16:06
RCFILE_1="jetpipe params"
16:06
in lts.conf
16:06
So that we see if the problem is that jetpipe starts too early, or if it's buggy
16:06
<igorek_RU>
ok, i try this
16:06
thanks
16:10
very strange if I run jetpipe hands, it works, even after I sent to print multiple jobs. I'll try it always works. Thank you again
16:11
*hands -> manually (sorry for my english )) )
16:12
<knipwim>
alkisg: changed the date call in the upcoming meeting and added the GSoC topic
16:12
thx
16:12
<Enslaver>
knipwin: haven't seen that yet
16:12
<alkisg>
knipwim: cool - can we also make a page for the hackathon?
16:12
To program some things for then?
16:12
E.g. drop jetpipe, cdpinger...
16:13
<Enslaver>
i turn off host only and include nfs in the conf
16:14
There is one issue i have with it, I'm not sure if its directly related or not but my mouse acts up in X, i see HAL handling the mouse and I'm trying to figure out how to turn HAL off
16:16yalu has left IRC (yalu!~yalu@109.134.174.29, Read error: Operation timed out)
16:17
<knipwim>
Enslaver: this is an example of the command i use: dracut --force -m "kernel-modules nbd nfs network base" --filesystems "squashfs" /boot/initramfs-dracut-x86_64-3.3.8-gentoo.img 3.3.8-gentoo
16:18
alkisg: will do, i already have a page for ltsp-lightdm
16:18
<alkisg>
Nice, thanks
16:18
<knipwim>
i'll make the dev page a bit more visible also
16:19
<Enslaver>
knipwim: -m specifies dracut modules, do you specify kernel ones as well?
16:19
<knipwim>
nope, they are include with the kernel-modules
16:20yalu has joined IRC (yalu!~yalu@183.133-66-87.adsl-dyn.isp.belgacom.be)
16:20
<alkisg>
jammcq: ah I forgot, also jetpipe is in python while nc is in C (and we're already using it anyway) so it'll need much less memory
16:20
<knipwim>
hostonly was optional, so i didn't need to include it
16:21
hmm
16:21
maybe it's included through the dracut.conf
16:22
<Enslaver>
knipwim: http://pastebin.com/RDCxwShx
16:22
<knipwim>
bbiab
16:29
Enslaver: what are those for?
16:31
<Enslaver>
Thats the patch to dracut for the error you linked
16:31
looks like a cheap workaround tho
16:31
<knipwim>
not the systemd i prosume, i don't use it
16:32
<Enslaver>
no
16:34
it looks like it can be applied to anything
16:35igorek_RU has left IRC (igorek_RU!5f34e021@gateway/web/freenode/ip.95.52.224.33, Quit: Page closed)
16:35
<Enslaver>
You ever play with buildroot?
16:35
<knipwim>
nope
16:37
<Enslaver>
http://buildroot.uclibc.org/
16:38highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Quit: bbl)
16:40
<knipwim>
for building chroots it looks interesting
16:40
especially the cross build options
16:42ltspuser_37 has joined IRC (ltspuser_37!c31772ca@gateway/web/freenode/ip.195.23.114.202)
16:42
<Enslaver>
supports roots overlay's now too
16:42
<ltspuser_37>
Hi
16:43
<alkisg>
Hi
16:44
<ltspuser_37>
I'm using kiosk and I'm configuring lts.conf like this : [Default] SERVER = "192.168.20.1" DNS_SERVER = "192.168.20.2" SEARCH_DOMAIN = "domain.com" SCREEN_07 = "kiosk" KIOSK_OPTIONS="http://test.domain.com/test.php"
16:44
but firefox didn't open the page...
16:44
and appears one message that firefox is free....
16:45
I need to disable that message on firefox settings too but I don't find where I can change that
16:45dobber_ has left IRC (dobber_!~dobber@89.190.199.210, Remote host closed the connection)
16:46vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi)
16:46meamy has left IRC (meamy!~hannes@pd95cdee4.dip0.t-ipconnect.de, Quit: meamy)
16:50
<knipwim>
alkisg: http://wiki.ltsp.org/wiki/Dev:Hackathon_2013.03
16:51
<alkisg>
knipwim: cool, ty
16:51
<Enslaver>
Well thats good dates at least :) I am away on vacation march 5-10
16:52
and my bday is 17th
16:52
<knipwim>
nice, then you'll be all rested when we start it :)
16:52
<Enslaver>
if I'm done by then :)
16:54
knipwim: http://bazaar.launchpad.net/~enslaver-l/ltsp/ltsp-redhat/revision/2462
16:56
<knipwim>
that won't fix my nfs issue :(
16:57
<Enslaver>
heh
16:57
it fixed mine
16:58
<knipwim>
when using the overlay
16:58
<Enslaver>
yah
16:58ltspuser_37 has left IRC (ltspuser_37!c31772ca@gateway/web/freenode/ip.195.23.114.202, Quit: Page closed)
16:58
<Enslaver>
are you using my 00-overlay?
16:58Gremble has left IRC (Gremble!~Ben@cpc35-aztw23-2-0-cust207.18-1.cable.virginmedia.com, Quit: I Leave)
16:59
<knipwim>
not yet
16:59
i'm just about to build a test chroot
16:59
<Enslaver>
give it a try, i think i made it pretty compatible with initramfs-tools and even to be run from init-ltsp.d
17:00
<knipwim>
i'm also confident in the dracut overlay module you made
17:00
<Enslaver>
well that is the dracut overlay module
17:00
all my overlay module does is call that 00-overlay into the initramfs
17:01
<knipwim>
hehe
17:01
if it works, we should make it more generic, so it can be included upstream
17:02
<Enslaver>
its as generic as i could get, i have support for it being run as a dracut module, through initramfs-tools, and even if you don't use an initrd
17:12andygraybeal_ has joined IRC (andygraybeal_!~andy@h154.214.22.98.dynamic.ip.windstream.net)
17:14highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage)
17:26ltspuser_52 has joined IRC (ltspuser_52!c31772ca@gateway/web/freenode/ip.195.23.114.202)
17:26
<ltspuser_52>
Hi
17:27
I need to have a scenario like this
17:27yalu has left IRC (yalu!~yalu@183.133-66-87.adsl-dyn.isp.belgacom.be, Ping timeout: 256 seconds)
17:27
<ltspuser_52>
I have a card, or an usb pen that has the user information
17:27
and I need to have a screen asking for this info
17:27
when I put the pen or the card I run a script and I get the rdp server
17:28
and I start the session on a display
17:28yalu has joined IRC (yalu!~yalu@59.132-66-87.adsl-dyn.isp.belgacom.be)
17:28
<ltspuser_52>
which is the best way to have this scenario?
17:30
it's possible?
17:31dberkholz has left IRC (dberkholz!~dberkholz@gentoo/developer/dberkholz, *.net *.split)
17:31hays has left IRC (hays!~quassel@unaffiliated/hays, *.net *.split)
17:31lmds_ has left IRC (lmds_!~lmds@213.41.240.245, *.net *.split)
17:34highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Ping timeout: 252 seconds)
17:36dberkholz has joined IRC (dberkholz!~dberkholz@gentoo/developer/dberkholz)
17:36hays has joined IRC (hays!~quassel@unaffiliated/hays)
17:36lmds_ has joined IRC (lmds_!~lmds@213.41.240.245)
17:51highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage)
17:56highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Read error: Operation timed out)
18:03mikkel has joined IRC (mikkel!~mikkel@80-71-132-15.u.parknet.dk)
18:07
<sbalneav>
ltspuser_52: Are you asking for smartcard authentication, or the like?
18:11
<ltspuser_52>
yes
18:11
it's possible
18:11
?
18:12
<sbalneav>
Not currently, no.
18:12
We're going to be having a hackathon in March, to work in getting LTSP to use a regular display manager and PAM
18:13
Once LTSP uses a pam stack, then at that point, you may be able to use one of the pam smartcard modules.
18:13
<vmlintu>
where is the hackathon?
18:13
<sbalneav>
http://www.gooze.eu/howto/gnu-linux-smartcard-logon-using-pam-pkcs11
18:14
<jammcq>
vmlintu: in irc
18:14
<sbalneav>
vmlintu: here :)
18:14
<jammcq>
sbalneav: Scotty !!!!!!!!!!!!!!!!!!!!!!!
18:14
<vmlintu>
oh, ok :)
18:14
<sbalneav>
Morning jammcq
18:14
or afternoon
18:14
after 12 I suppose.
18:14
<vmlintu>
lightdm does work nicely with ltsp, but it does require a bit more memory than ldm
18:15
<sbalneav>
Theoretically, any DM would work.
18:15
So something like slim may be good.
18:16
<awilliams>
Hello ltsp'rs
18:16
<jammcq>
hey awilliams
18:17
<awilliams>
How you guys doing
18:17
<jammcq>
dong well, how's things down in Jamaica?
18:17vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc)
18:17
<awilliams>
jammcq: I may have a use for ltsp comming up shortly
18:17
jammcq: Its raining something fierce
18:18
<vmlintu>
In some cases lightdm is quite slow to start compared to ldm. Maybe I have to test how slim would work instead..
18:20
<jammcq>
my fear at this point is ltsp is getting bloated
18:21
<awilliams>
jammcq: I have someone here who will be answering the phone for support calls and such, I need to eliminate the fans and other computer noise in the room
18:21
I can't think of a better way to handle it
18:21
<jammcq>
sounds like a great fit
18:21
<vmlintu>
jammcq: compared to 4.2 it's quite heavy
18:22
but then again fat clients weren't really an option then
18:24
<awilliams>
sbalneav: How you been man, in case you don't know its me mistik1 :)
18:24andygraybeal_ has left IRC (andygraybeal_!~andy@h154.214.22.98.dynamic.ip.windstream.net, Read error: Connection reset by peer)
18:25
<sbalneav>
awilliams: Hey there! How are you?
18:27
<vagrantc>
it's not so much LTSP is getting heavy, as the free software world has gotten heavy...
18:28
LTSP is only as lightweight as the components it's built from
18:29
<sbalneav>
+1
18:31
<vagrantc>
so, we either need to pick lighter components, or get patches into the components we're using to lighten them up a bit...
18:31
<awilliams>
vagrantc: indeed
18:31ltspuser_52 has left IRC (ltspuser_52!c31772ca@gateway/web/freenode/ip.195.23.114.202, Quit: Page closed)
18:31
<awilliams>
sbalneav: Doing pretty well, can't complain at all
18:32
<jammcq>
in ltsp-4.2, we streamlined that boot process down to the smallest possible set of scripts. Now... streamlined is not what i'd use to describe it
18:32
but... i don't wanna sound like i'm complaining, cuz it really works pretty well
18:32
but it really takes a powerful workstation to run with any kind of performance now
18:33
<vagrantc>
powerful workstation?
18:33
<jammcq>
works great with a 1.6ghz cpu with 4gb of ram
18:33
but I found a 1.2ghz box with 512mb really slow
18:34
<vagrantc>
most of the thin clients we're using at freegeek are under 1GHz with 128-512MB of ram...
18:34
i've even seen people using multimedia on these, much to my surprise
18:35
for fatclients, sure, it needs a bit more resources than a disked machine ... but there was no such thing as fatclients for ltsp 4.x, no?
18:36
<jammcq>
true
18:36
so like you said, it's not necessarily LTSP that's getting bloated, it's Linux in general
18:39
<vagrantc>
firefox in particular... eesh.
18:42
<awilliams>
FF18 is a total dog
18:43* jammcq prefers chrome
18:43* awilliams too
18:45* vagrantc isn't too fond of self-surveilance
18:46
<awilliams>
heh, which is why I wish FF did not do the chrome chase
18:47* vagrantc is still running iceweasel 10
18:47
<vagrantc>
but yeah, the absurd release every 6 weeks thing is insane.
18:48
6 weeks of feature development and QA... no problem.
18:50
<jammcq>
vagrantc: earlier today, someone had a problem with a local printer. alkisg pointed him to the bug that was fixed back in august. I understand wheezy was in a freeze. when do you suppose that fix would make it into wheezy?
18:51
<vagrantc>
jammcq: i'd have to see an actual bug report to have a clue
18:51
jammcq: in wheezy's LTSP, or in some other printing system?
18:52
<jammcq>
09:52 < alkisg> https://bugs.launchpad.net/ltsp/+bug/996533
18:52
it's an ltsp bug as far as I can tell
18:52
thin client can't boot if it has set PRINTER_0_DEVICE in lts.conf
18:54
<vagrantc>
jammcq: i had planned to get a fix in for that... but it may have fallen off the radar
18:55
jammcq: ahh, that's a bit too invasive to fix, i think...
18:55
<jammcq>
so what's the solution?
18:55
people with wheezy just live with not being able to print to local printers?
18:55
<vagrantc>
i daresay, that may be the case.
18:55
<jammcq>
unless they hack the script themselves?
18:56
<vagrantc>
but first step would be to actually report the bug to debian
18:56
otherwise i can't realistsically get a freeze exceptionn
18:58
<jammcq>
is the freeze just leading up to final release? or is it frozen for ever?
18:59
<vagrantc>
major bugs can get fixed, as long as they're not too invasive
18:59
i'd classify this as "important" although maybe it could be "serious" since it actually results in a broken boot.
19:00
jammcq: just for clarity, you've actually confirmed that this bug exists in debian wheezy?
19:00
<jammcq>
the guy in here earlier today said he was using it
19:00
i haven't plugged a printer in. I don't have any.
19:00
<vagrantc>
it's not broken in the default configuration, or in most cases, so probably only important
19:01
jammcq: from the bug report, it sounds like setting the value at all can trigger the bug
19:01
printer or not
19:01
<jammcq>
but again, is it frozen because it's about to be released ? or does this restriction last forever?
19:01
<vagrantc>
jammcq: yes, it's about to be released, so changes get more review.
19:01
and invasive changes aren't allowed
19:02
once it's released, similar issues apply
19:02
<jammcq>
what happens after release? does it get easier to fix then?
19:03
<vagrantc>
and again, someone (ideally not the maintainer) should report a bug, if there's any chance to get it fixed at all.
19:03
<jammcq>
ok, where do I do that?
19:03
<vagrantc>
jammcq: after release, it's probably a little harder to get fixes in, yes.
19:04
<jammcq>
oh, so it's hard now, but really hard later :)
19:04
<vagrantc>
jammcq: http://www.debian.org/Bugs/Reporting
19:04
it'll also be really slow later
19:04
i.e. it could be months and months before it gets in
19:06
looks like alkisg committed a fix that, if it actually worked, would probably be acceptible: https://bazaar.launchpad.net/~ltsp-upstream/ltsp/ltsp-trunk/revision/2382
19:07
but the stgraber rewrote jetpipe to not use python-daemon
19:07
<jammcq>
is there not a web-based interface to the bug reporting system?
19:08
<vagrantc>
https://bazaar.launchpad.net/~ltsp-upstream/ltsp/ltsp-trunk/revision/2395
19:08
jammcq: not for reporting bugs... for viewing them.
19:08
<jammcq>
oye
19:08
<vagrantc>
jammcq: for making changes, it's all email based.
19:09
<jammcq>
I don't have an smtp server so I can't submit it
19:10
at least not with reportbug
19:10
<vagrantc>
you can save the file and cut-and-paste into an email
19:11
jammcq: basically reportbug tries to gather useful information ... although you could craft it by hand, too.
19:14
<jammcq>
vagrantc: what would the severity be for this?
19:14
'important' ?
19:15
<vagrantc>
jammcq: yeah
19:15
jammcq: it's pretty severe in it's consequences, but doesn't affect the default or most installs ... only ones that use that feature.
19:15
<jammcq>
does it apply to an upstream part of the package?
19:16
<vagrantc>
how do you mean?
19:16
<jammcq>
it's asking me if the following apply to this report:
19:16
and it lists 7 different choices
19:16
<vagrantc>
ah, yeah, it's an upstream issue
19:16
<jammcq>
6 upstream
19:16
5 patch
19:16
am I supplying the patch or do I just list the upstream bug info?
19:17
<vagrantc>
you could claim there's a patch ... pointing to the two bzr commits that could potentially be used to fix the issue.
19:18
<jammcq>
k
19:18* vagrantc will try to confirm the bug...
19:23
<vagrantc>
yup, the bug is definitely there...
19:23* vagrantc thinks init-ltsp.d is a bad place to start a daemon from...
19:25* vagrantc wonders if you ran it after the full boot process was done if it would work better.
19:26
<jammcq>
for USB printers, what if we just run it from udev ?
19:26* vagrantc nods
19:27
<vagrantc>
well, the simple one-line "jetpipe ... &" at least allows it to boot. no idea if it actually works ...
19:27
i also don't have a printer handy to test...
19:29
maybe for wheezy it'd be ok to move it back into /etc/init.d/ltsp-client*
19:32komunista has left IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk, Quit: Leaving.)
19:34
<vagrantc>
ok, my hunch was right ...
19:34
running it from /etc/init.d/ltsp-client-core works fine.
19:35
<jammcq>
so it's just running at the wrong time?
19:35
<vagrantc>
yeah.
19:35
<jammcq>
well... seems like a simple fix
19:35
<vagrantc>
python-daemon probably requires something to be initialized.
19:35
<jammcq>
I filed the bug
19:35
<vagrantc>
jammcq: thanks!
19:35
<jammcq>
no prob, thanks for helping me through it
19:35
next up...
19:36
I still don't understand that setuid problem I found.
19:36
<vagrantc>
me neither.
19:36
<jammcq>
on your workstation, are you running X or Xorg ?
19:37
<vagrantc>
looks like X ...
19:37
<jammcq>
and on mine, it's Xorg
19:37
i'm guessing yours is pretty current
19:37
<vagrantc>
i don't think that behavior has changed for years.
19:38
<jammcq>
I dont' understand why mine is runnign Xorg
19:38
I just built it from the latest packages on saturday
19:38
<Hyperbyte>
Hi everyone. :) Good evening!
19:38
<jammcq>
hey Hyperbyte
19:39
<Hyperbyte>
How are things? :)
19:39
<vagrantc>
jammcq: and nothing in lts.conf other than SCREEN_07=kiosk ?
19:39
<jammcq>
right
19:39
well... KIOSK_WM = /usr/bin/icewm-session
19:40
and KIOSK_EXE = /usr/bin/xterm
19:40
but no X config stuff
19:40* vagrantc tries with the same...
19:40komunista has joined IRC (komunista!~slavko@adsl-195-168-244-224.dynamic.nextra.sk)
19:44
<vagrantc>
jammcq: still works for me...
19:45
jammcq: so ... what exactly is calling Xorg directly, i wonder ... can you set SCREEN_02=shell and run "pstree -a" just to see what called it?
19:45
<jammcq>
and you are running X and it has setuid?
19:45
sure, gimme a couple minutes
19:46
<vagrantc>
jammcq: ah, with "pstree -a" i see Xorg
19:46
jammcq: but with "pgrep -l -f X" i see X
19:47
that's ... odd.
19:47
jammcq: X has setuid and Xorg doesn't.
19:47
jammcq: or, X has setuid and setgid, Xorg doesn't have either.
19:48
-rwsr-sr-x 1 root root 14264 Jun 25 2012 /opt/ltsp/amd64/usr/bin/X
19:48
-rwxr-xr-x 1 root root 2044664 Nov 29 11:34 /opt/ltsp/amd64/usr/bin/Xorg
19:48
<jammcq>
right
19:48
that's what I saw
19:48
and for some reason, mine is running Xorg and yours is running X
19:48
<vagrantc>
from what i remember, "X" is essentially a thin wrapper around Xorg z...
19:49
jammcq: well, it's not clear to me now ... pstree shows "Xorg", pgrep shows "X"
19:49
to make matters worse ... /etc/X11/X is a symlink to Xorg.
19:50
<jammcq>
pstree shows that Xorg is running from xinit
19:50
oh?
19:50
<vagrantc>
same here.
19:51
<jammcq>
does your xinit line say: .... -config /var/run/ltsp-xorg.conf
19:51
or is it -config var/run/ltsp-xorg.conf
19:51
without the leading slash
19:51
<vagrantc>
nope.
19:51
neither.
19:51
<jammcq>
?
19:51anunnaki has joined IRC (anunnaki!~anunnaki@c-174-54-115-236.hsd1.pa.comcast.net)
19:51
<jammcq>
no -config ?
19:51
<vagrantc>
no -config
19:51
probably a difference in video card...
19:52
<jammcq>
hmmm, why would mine have a -config and yours not?
19:52
<vagrantc>
intel video?
19:52
<jammcq>
ummm
19:52
<vagrantc>
jammcq: what's in your ltsp-xorg.conf ?
19:52
<jammcq>
yeah
19:52
<vagrantc>
there are some settings that autodetect settings, some of which require an xorg.conf
19:53
<jammcq>
Section "Screen"
19:53
IDENTIFIER "Screen0"
19:53
SubSection "Display"
19:53
EndSubSection
19:53
EndSection
19:53
that's it
19:54
<vagrantc>
jammcq: probably something in /opt/ltsp/amd64/usr/share/ltsp/screen-sessions.d/ is triggering a "needed" xorg.conf setting...
19:55
jammcq: i'd suspect a combination of XS20-intel-virtual and XS85-virtual
19:56
jammcq: try setting X_VIRTUAL_AUTO=false
19:57
<jammcq>
booting
19:57
<vagrantc>
jammcq: but it sounds like there's a bug with kiosk mode and "X -config ..."
19:58
probably also affects rdesktop, maybe others
19:59
<jammcq>
yeah, when I booted, there wasn't a -config anymore
19:59
so now I removed setuid on Xorg
19:59
and i'm booting again
20:03
yes, without the -config, Xorg doesn't need setuid
20:03
but... I could see there would be times when you need the -config, and then Xorg would be broken without setuid
20:04
<vagrantc>
so, this might be worth commenting on http://bugs.debian.org/587900
20:04
jammcq: right.
20:06
<jammcq>
I don't see how to comment on that bug
20:06
<vagrantc>
email 587900@bugs.debian.org
20:06
<jammcq>
seriously?
20:07
<vagrantc>
you could even download the mbox and reply to it... :)
20:07
<jammcq>
why not just use gopher to download the isos?
20:07
<vagrantc>
jammcq: because email is useable, and gopher is not?
20:08
<jammcq>
ummm, yeah
20:09
<vagrantc>
it may not be your preference, but it's the way it is :P
20:10
i definitely struggle with the dozens of poorly implemented web interfaces that can't even do simple things without javascript.
20:11
or require you to actually sign up with an account
20:12
<Enslaver>
my raspi ltsp image is almost done compiling ;)
20:13
<vagrantc>
takes me about 5 minutes to build mine :)
20:13
<Enslaver>
how so?
20:13
<vagrantc>
ltsp-build-client --arch armhf ...
20:14
using qemu-user emulation on a reasonably fast x86 machine
20:14
although i don't have all the firmware and such... that takes a little longer manually.
20:14
<jammcq>
vagrantc: I don't think I understand enough about this -config problem to comment on it. it's still confusing me
20:14
<Enslaver>
that just grabbing pre-compiled binaries?
20:15
<jammcq>
I couldn't boot, because Xorg didn't have setuid
20:15
<vagrantc>
Enslaver: yes, from raspbian
20:15
<Enslaver>
Ah
20:15
<jammcq>
then, when I gave it setuid, I still couldn't because the -config pathname started with a '/'
20:15
<vagrantc>
Enslaver: just like any other LTSP build
20:15
<Enslaver>
I'm compiling the entire kernel from source along with all binaries, takes me a little while, lol
20:15
I shoulda mock'd it
20:15
<vagrantc>
Enslaver: yes, i'm glad to have others handle that part :)
20:16
<Enslaver>
vagrantc: btw whats the debian equivalent of a spec file?
20:16
<ogra_>
arent you on fedora ? there should be binaries for it
20:16
<vagrantc>
jammcq: it sounds like a separate bug to me.
20:16
Enslaver: the debian directory
20:16
Enslaver: it's not a single file, but a whole mess of files.
20:16
<Enslaver>
there probably is binaries, but noting as custom as i want
20:16
<ogra_>
and a big book of policies
20:16
<Enslaver>
and I'm on el6, not fedora
20:17
<vagrantc>
ogra_: it's not so much as a book as a short story
20:17
<ogra_>
an adventure game :)
20:17
<vagrantc>
exactly
20:17
<Enslaver>
sounds like where's waldo
20:17* vagrantc open door south
20:17* vagrantc is eaten by a groo
20:17
<ogra_>
a grue is in your way
20:18
<vagrantc>
groo, policy violation. same diff.
20:18
<ogra_>
haha
20:19
<alkisg>
vagrantc, jammcq: I think the jetpipe/init-ltsp.d/& issue also involves that, when started from init-ltsp.d with &, it only works once!!!
20:19* Enslaver is sorry to say, you have been eaten by a Grue.
20:19
<alkisg>
So maybe it should internally be translated into RCFILE_0 and run from rc.local
20:19* alkisg doesn't use jetpipe at all though so he never tested that
20:20
<vagrantc>
jammcq: yeah, so that sounds like another upstream bug ... i.e. kiosk (and probably rdesktop) don't work when a -config file is specified.
20:20
alkisg: i.e. it prints only one thing?
20:20
<ogra_>
alkisg, you could bind it to a udev rule instead of starting it from init
20:20
<alkisg>
That's what the last wheezy user here said
20:20
<vagrantc>
alkisg: i moved it back into /etc/init.d/ltsp-client-core and it seemed to actually run...
20:21
<alkisg>
ogra_: I think some dev that has actually used it at least once should take over that bug...
20:21
<vagrantc>
meanwhile, it'd be good to have it not totally broken on wheezy ... :|
20:21
<alkisg>
I just committed a small fix that people here reported that it worked, without testing, just to let them at least boot
20:21
<vagrantc>
i could probably dig up a USB printer to test with
20:22
but not sure about other kinds of printers
20:22
<jammcq>
vagrantc: is it because normally, X runs as root, but in kiosk mode it runs as a normal user?
20:22
<ogra_>
yes, the backgrounding should work, what dio you mean by only runs once ? it dies after printing ?
20:22
<vagrantc>
jammcq: yeah, i think that's basically what's triggering your problem.
20:22
<alkisg>
jammcq: why not run X as root then and the browser as local user afterwards?
20:22
<vagrantc>
running daemons from init-ltsp.d seems like a bad idea.
20:22
<jammcq>
alkisg: i'm just using the kioskSession script and that's what it's doing
20:23
<vagrantc>
alkisg: having to mangle the XAUTHORITY stuff then...
20:23
<jammcq>
when we switch to lightdm, doesn't that run X as the user?
20:23
<vagrantc>
jammcq: probably a hack to work around it would be to use the relative path stuff that you found as a workaround.
20:23
<jammcq>
so we'd still have a problem
20:23
<alkisg>
jammcq: no, as root
20:23
root 1359 2.5 0.6 56856 26564 tty7 Rs+ 12:22 15:15 /usr/bin/X :0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch -background none
20:23
<vagrantc>
jammcq: and it handles all the XAUTHORITY stuff, so we don't have to do it manually :)
20:24
<alkisg>
I don't think we should be the ones reporting bugs about Xorg not running OK as user
20:24
Let other DMs do that first, and we follow afterwards :D
20:25
<vagrantc>
meanwhile, solving bugs today...
20:25
<jammcq>
so really, like alkisg said, xinit/X should run as root, and the window manager should run as the user
20:25
<vagrantc>
xinit typically handles that ...
20:25
<jammcq>
so it means moving where the 'su - $kioskuser' happens
20:25
<vagrantc>
the other option would be to symlink /etc/X11/xorg.conf to /var/cache/ltsp/ltsp-xorg.conf
20:26
<alkisg>
ogra_: when specifying init=/bin/bash, I think I get a prompt about backgrounding not supported etc, so maybe that's why & doesn't work well there. I think it's the only case where we launch a daemon though, so moving it somewhere else sounds like a good idea...
20:26
<vagrantc>
and then not specify -config at all
20:26
alkisg: yes! we shouldn't launch daemons from init-ltsp.d!
20:26
<ogra_>
alkisg, oh, you boot with init=/bin/bash ?
20:26
<alkisg>
ogra_: similar, init=/bin/init-ltsp
20:26
*sbin
20:26
<ogra_>
yeah, that indeed has no job control
20:27
<knipwim>
Enslaver: looking at 00-overlay now, is there a special reason you check for /etc/ltsp_chroot ?
20:27
<vagrantc>
it should only configure daemons and stuff and use the standard mechanism for starting daemons.
20:27
<ogra_>
right, that bit should be donr by init
20:27
*done
20:27
<vagrantc>
checking for /etc/ltsp_chroot is the old way to make sure it's an LTSP environment ...
20:27
but we stopped using that recently.
20:27
<Enslaver>
whats the new way?
20:27
<vagrantc>
in order to support ltsp-pnp style environments.
20:28
checking /proc/cmdline for init=/sbin/init-ltsp
20:28
or variations on that
20:28
<alkisg>
The idea is that an ltsp chroot can also be booted "locally"
20:28
Either on a template client, a VM, or an "ltsp-pnp" ltsp server
20:29
<Enslaver>
yeah i like that concept, that's where the ltsp-cleanup comes in I'm guessing
20:29
to remove user accounts and such?
20:29
<alkisg>
Yeah, although that can be useful on chroots too
20:29
<vagrantc>
it is one of the best innovations LTSP has seen in a while :)
20:29
<Enslaver>
later down the road we can prob set that up as 2 different 'sandbox' filesystems
20:30
checks based on boot hardware or /proc/cmdline
20:30
<vagrantc>
i wouldn't make assumptions about hardware, at least not as upstream.
20:30
assumptions based on hardware, rather
20:31
<alkisg>
An LTSP server can also be booted as an LTSP client with no hardware differences...
20:31* vagrantc nods
20:32
<Enslaver>
yah i've been playing a bit with revisor, kinda similar concept, builds a chroot for different environments that can be both live-filesystem and live-cd
20:33
<vagrantc>
the greatest thing about the ltsp-pnp mode is setting up fatclients by installing what you want on the server, and then building your fatclient image out of it.
20:34* vagrantc wonders of NBD will chain off of another NBD server ...
20:34
<Enslaver>
hmm, maybe i need to look into that before i go too deep in the fat client route
20:34
Where can i grab the rpms?
20:35
<vagrantc>
it's all in upstream LTSP ...
20:35
it's essentially a different way to build a fatclient, all the stuff you need for fatclient support is the same.
20:36
!ltsp-pnp
20:36
<ltsp>
ltsp-pnp: ltsp-pnp is an alternative (upstream) method to maintain LTSP installations for thin and fat clients that doesn't involve chroots: https://help.ubuntu.com/community/UbuntuLTSP/ltsp-pnp
20:36
<Enslaver>
what does it use to build it?
20:38
<vagrantc>
ltsp-update-image --cleanup /
20:38
it's basically a chroot-less way to build an NBD image
20:39
<Enslaver>
so it builds a squashfs of /
20:41
<vagrantc>
essentially
20:45
<Enslaver>
one question about the cmdline=/sbin/init-ltsp check, what if you're chroot'd to the chroot? How are the scripts able to determine?
20:45
<vagrantc>
Enslaver: that would only be an issue if you're booted as a thin client, chrooting to a chroot, and not wanting it to behave as a thin client ...
20:46
<Enslaver>
or if i type chroot /opt/ltsp/i386 from the server
20:47
<vagrantc>
then it shouldn't behave as a thin client...
20:47
because it's a chroot
20:47
<Enslaver>
same file structure
20:47
<vagrantc>
but it's not a thin client
20:48
that's where you'd do maintenance of the thin client, not running the thin client itself.
20:49
<Enslaver>
yeah I'm not saying use that check for running thin client apps, i use it more as a safety precaution
20:49
<alkisg>
The cmdline needs to be checked from the initramfs
20:49
In the bootem system, /sbin/init-ltsp can save some var that it has been called
20:49
No need to check cmdline there
20:51
So if e.g. we save that to /run, and a chroot doesn't bind-mount /run, it won't be considered a "booted client"
20:51
<vagrantc>
Enslaver: also, we've worked to make installing ltsp-client* packages outside of the thin client environment non-destructive ... so they only enable thin-client behaviors when booted as a thin client
20:51
<knipwim>
Enslaver: the 00-overlays says "not done" at elif [ "$INSIDE_LTSP_INIT" ]
20:51
is it really not done, or did you forget to remove the return and comment?
20:52
<alkisg>
(and ltsp-build-client plugins too, we tried to move some stuff from there to init-ltsp.d)
20:52
<Enslaver>
I commented above that, it needs to be filled in by someone who understands initramfs a little better before those 2 will be active
20:53
<knipwim>
but that one is inside the ltsp init
20:53
<Enslaver>
i prefer to escape out of an untested function, rather than have it bomb
20:53
the ltsp-init one I'm testing now
20:53
<knipwim>
kk
20:53
<vagrantc>
jammcq: do you have access to local printers to test that bug?
20:54
jammcq: or know someone who does?
20:54
<Enslaver>
if you can paste bin me the lsinitrd from an initramfs made from initramfs-tools that would be helpful
20:54
<jammcq>
ummm
20:54
<vagrantc>
jammcq: i think i could put together something that might work, if you've got a way to test
20:54
<Enslaver>
(i.e. i'm checking for dracut-lib.sh for INSIDE_DRACUT)
20:54
<knipwim>
Enslaver: also, you can use "die" instead of defining a panic()
20:54
<jammcq>
the only printer I have here is a network printer
20:54
<alkisg>
Enslaver: $ lsinitramfs /boot/initrd.img-3.5.0-23-generic | pastebinit
20:54
http://paste.ubuntu.com/1678562/
20:55
<jammcq>
I could do it in vmware
20:55
<Enslaver>
knip: which rev you looking at?
20:55
<jammcq>
hmmm
20:55
<vagrantc>
jammcq: i can't figure out how to emulate a printer in libvirt ...
20:55
<knipwim>
Enslaver: 2463
20:55
<jammcq>
vmware lets me assign a device to a file
20:56
<vagrantc>
ah, i could do that too, i guess. no idea what valid printing output would be, though.
20:56* alkisg is happy to see more devs in ltsp again :)
20:56
<knipwim>
Enslaver: i was also wondering about the modprobe
20:56
and if a /proc/config.gz could also be done for an in-kernel module
20:56
<vagrantc>
alkisg: it's almost like having meetings has spurned activity or something :)
20:57
<Enslaver>
knip: what about the modprobe?
20:57
<vagrantc>
knipwim: at least on debian systems, modprobe will return true if it's a built-in.
20:57
which is a little confusing
20:58vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 255 seconds)
20:59
<knipwim>
vagrantc: nice
21:00
<Enslaver>
knip: just pushed 2465, the panic was giving me problems, i'd rather just try and mount tmp before i panic
21:03highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage)
21:10alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
21:11anunnaki has left IRC (anunnaki!~anunnaki@c-174-54-115-236.hsd1.pa.comcast.net, Ping timeout: 248 seconds)
21:19
<knipwim>
Enslaver: it seems to be working
21:20
<Enslaver>
in dracut?
21:21
<knipwim>
no network though, but that's probably another problem
21:21
no, in the init-ltsp.d
21:21PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.130.7)
21:21
<Enslaver>
ah nice, i can't seem to get mine working
21:22
although its a completely new way I'm building a chroot, could be unrelated to the script
21:23
<knipwim>
apparantly proc is already mounted on gentoo, so that could use a check
21:24
<Enslaver>
k
21:25
<knipwim>
also, i removed the if [ -n "$OVERLAYFS" ], since you already set a default earlier
21:27
<Enslaver>
you removed the -n or the whole line?
21:27
OVERLAY: bool, OVERLAYFS: string
21:29
i'll prob redesign that into a case statement eventually
21:30
wow, my pi is still comPiling
21:31
<vagrantc>
Enslaver: i would guess it would take days...
21:31Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 255 seconds)
21:31
<Enslaver>
I was hoping for a little sooner :( This server has 40g ram and quad xeon
21:31
<vagrantc>
Enslaver: considering the kernel alone took several hours on faster arm hardware
21:31
ah, cross-compiling
21:31
<Enslaver>
yah
21:32
omg compiling a kernel on the actual arm pi, yeah i could see that taking a minute or two
21:32
<knipwim>
Enslaver: the whole line, because fo OVERLAYFS=${OVERLAYFS:-unionfs}
21:32
so -n "$OVERLAYFS" is always true
21:34
<Enslaver>
wanna hear how smart I am? mine wasn't working because i left the return in there...
21:34
<vagrantc>
figuring out one's own mistakes is a sign of intelligence :)
21:35
<knipwim>
yeah, i saw that on my first try :)
21:35
<vagrantc>
fail better, fail faster :)
21:35
er, fail faster, fail better
21:35
<Enslaver>
fool myself once, shame on me
21:37highvoltage has left IRC (highvoltage!~highvolta@ubuntu/member/highvoltage, Quit: Lost terminal)
21:39
<Enslaver>
for initramfs-tools, where is the rootfs mounted inside it?
21:42
<vagrantc>
typically /root, if i remember correctly.
21:42
it's definited in a variable ${rootmnt}
21:43
<Enslaver>
[ -d /scripts/local-top ] && [ -n "${rootmnt}" ] && INSIDE_INITRAMFS=1
21:43
something like that safe to assume?
21:44
<vagrantc>
looks like i've got a virtualized printer to test the bug with ... if i can just remember how to use a printer...
21:44
<Enslaver>
cat /dev/urandom > /dev/lp0
21:44
<vagrantc>
Enslaver: reasonably...
21:44
<Enslaver>
banner boobs?
21:44
<vagrantc>
Enslaver: well, i'm testing our network printing daemon, jetpipe...
21:45
<Enslaver>
jetpipe = jetdirect redirects?
21:45
<vagrantc>
basically, yeah
21:45
Enslaver: and, the whole "boobs" thing is a bit wearisome :P
21:46
<Enslaver>
I didn't mean boobs, i meant 80085
21:46Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas)
21:48
<Enslaver>
i have some lpadmin scripts you might like
21:48
<vagrantc>
i think i've got it working with netcat :)
21:48
<Enslaver>
gj
21:48
If i remember correctly you can telnet to 9100 and issue PRINT or something
21:49
<vagrantc>
seems like jetpipe isn't much more than netcat, really.
21:49
well, running from the ltsp-client-core init script, rather than init-ltsp.d, seems to fix it in my crude test case...
21:49
<Enslaver>
03:00:01 PM all 56.77 0.00 34.10 0.03 0.00 9.09
21:49
03:10:01 PM all 18.39 0.01 20.23 0.13 0.00 61.25
21:49
03:20:01 PM all 84.07 0.00 10.47 0.00 0.00 5.45
21:49
lol my SAR from the pi build
22:08
<ogra_>
the original jetpipe was just four lines of shell
22:08
piping the port directly to the printer inetrface
22:08
*interface
22:10
(and it used netcat actually)
22:15
<vagrantc>
:)
22:15
who says LTSP is bloated? :)
22:26
jammcq: so i think i've got a patch that should be acceptible to the release team.
22:26
<jammcq>
cool
22:26
how do I test it?
22:26
<vagrantc>
jammcq: i guess i could build and upload some packages for you to test
22:26anunnaki has joined IRC (anunnaki!~anunnaki@174.54.115.236)
22:26
<vagrantc>
jammcq: on amd64? i386?
22:26
<jammcq>
that would be most awesome
22:27
umm, either, but prefer amd64
22:32anunnaki has left IRC (anunnaki!~anunnaki@174.54.115.236, Ping timeout: 245 seconds)
22:35dead_inside has left IRC (dead_inside!~dead_insi@76.75.3.174, Ping timeout: 260 seconds)
22:39bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Read error: No route to host)
22:41
<vagrantc>
jammcq: ok, add to your chroot's sources.list: deb http://cascadia.debian.net/~vagrant/debian UNRELEASED main
22:41
jammcq: ltsp-chroot apt-get update && ltsp-chroot apt-get dist-upgrade
22:41
jammcq: the packages are unsigned, so you'll get warnings about that
22:43
jammcq: (presuming ltsp-chroot is hitting the correct chroot)
22:44
jammcq: and of course, set PRINTER_0_DEVICE and all that ...
22:47
<jammcq>
k, I'll add that now
22:48
I prolly can't test printing until tomorrow tho
22:48
I'll set the option tho and see if it boots
22:51anunnaki has joined IRC (anunnaki!~anunnaki@174.54.115.236)
22:53* vagrantc wonders what other bugs are lurking for wheezy
23:00PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.130.7, Quit: Leaving)
23:03khildin_ has left IRC (khildin_!~khildin@ip-80-236-222-115.dsl.scarlet.be, Quit: I'm gone, bye bye)
23:09
<jammcq>
vagrantc: looks good so far. I configured PRINTER_0_DEVICE = /tmp/printer.out
23:09
booted the client
23:09
from the server, I do: telnet 192.16.200.20 9100
23:09
and whatever I type was written to /tmp/printer.out on the client
23:11
<vagrantc>
that's about what i did, except with netcat instead of telnet
23:11highvoltage has joined IRC (highvoltage!~highvolta@ubuntu/member/highvoltage)
23:11highvoltage is now known as hv-onlogged