IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 4 November 2015 (all times are UTC)
| 00:28 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
| 01:52 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 255 seconds) | |
| 01:55 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 02:00 | FrankBlues has joined IRC (FrankBlues!~alex@c-67-171-127-41.hsd1.ut.comcast.net) | |
| 02:07 | FrankBlues has left IRC (FrankBlues!~alex@c-67-171-127-41.hsd1.ut.comcast.net, Ping timeout: 240 seconds) | |
| 02:17 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 264 seconds) | |
| 02:20 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 02:28 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 255 seconds) | |
| 02:31 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 02:45 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 255 seconds) | |
| 02:48 | Freejack has joined IRC (Freejack!~Freejack@cpe-67-248-0-202.nycap.res.rr.com) | |
| 02:48 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 02:53 | <fnurl> test
| |
| 04:11 | Parker955_Away has left IRC (Parker955_Away!~parker@2607:5300:60:8425::2d62:a8e6, Ping timeout: 240 seconds) | |
| 04:11 | andygraybeal has left IRC (andygraybeal!~andy@h52.205.189.173.dynamic.ip.windstream.net, Ping timeout: 240 seconds) | |
| 04:11 | sbalneav has left IRC (sbalneav!~sbalneav@wnpgmb0412w-ds01-45-12-121.dynamic.mtsallstream.net, Ping timeout: 240 seconds) | |
| 04:12 | andygraybeal has joined IRC (andygraybeal!~andy@h52.205.189.173.dynamic.ip.windstream.net) | |
| 04:14 | Parker955_Away has joined IRC (Parker955_Away!~parker@2607:5300:60:8425::2d62:a8e6) | |
| 05:16 | karchungsek has joined IRC (karchungsek!766593f7@gateway/web/freenode/ip.118.101.147.247) | |
| 05:16 | work_alkisg is now known as alkisg | |
| 05:22 | <karchungsek> i am new user . i want to install adobe flash player but i cant . Can some one teach me step by step. Thanks a lot
| |
| 05:26 | karchungsek has left IRC (karchungsek!766593f7@gateway/web/freenode/ip.118.101.147.247, Ping timeout: 246 seconds) | |
| 06:16 | beakburke has joined IRC (beakburke!~burke@www.thealmquists.net) | |
| 06:17 | ricotz has joined IRC (ricotz!~ricotz@p5B2A8217.dip0.t-ipconnect.de) | |
| 06:17 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
| 06:21 | karchungsek has joined IRC (karchungsek!71d2004a@gateway/web/freenode/ip.113.210.0.74) | |
| 06:22 | sbalneav has joined IRC (sbalneav!~sbalneav@wnpgmb0412w-ds01-45-12-121.dynamic.mtsallstream.net) | |
| 06:22 | <karchungsek> aorry disconnected.
| |
| 06:30 | <alkisg> karchungsek: why do you think that installing flash is specific to ltsp?
| |
| 06:46 | teacher has joined IRC (teacher!c23fefeb@gateway/web/freenode/ip.194.63.239.235) | |
| 07:18 | book`_ has joined IRC (book`_!~book`@105.ip-167-114-152.net) | |
| 07:27 | book` has left IRC (book`!~book`@105.ip-167-114-152.net, *.net *.split) | |
| 07:31 | karchungsek has left IRC (karchungsek!71d2004a@gateway/web/freenode/ip.113.210.0.74, Ping timeout: 246 seconds) | |
| 07:34 | schlady has joined IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de) | |
| 07:36 | schlady has left IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de, Client Quit) | |
| 07:38 | schlady has joined IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de) | |
| 07:40 | * fnurl is pleased to get a good result from pandoc today :D | |
| 07:41 | <alkisg> Hi fnurl :)
| |
| 07:41 | <fnurl> heya - how you been?
| |
| 07:42 | <alkisg> All fine! I've sent the last LTSP build to the PPA and we're ready to go whenever you want
| |
| 07:42 | The pandoc chart is impressive, it has a lot of formats there
| |
| 07:42 | http://pandoc.org/diagram.jpg
| |
| 07:42 | <fnurl> ya - very very cool tool
| |
| 07:43 | it's not 100% perfect, but it does a helluva good job regardless
| |
| 07:43 | lemme pm you 1s
| |
| 07:51 | <maldridge> I've got pandoc plugged into a mail-to-print system and it provides very good results most of the time; its one of the best programs I'd never heard of
| |
| 08:00 | <fnurl> i read about it recently, and I'm well impressed, i'm trying to overhaul all of our documentation, and using pandoc is likely to save me hundreds of hours
| |
| 08:01 | moving from email as version control (OMG kill me) with MS docs -> markdown & git versioning with pandoc for IDML/HTML
| |
| 08:16 | adrianorg has left IRC (adrianorg!~adrianorg@187.113.244.122, Ping timeout: 250 seconds) | |
| 08:18 | adrianorg has joined IRC (adrianorg!~adrianorg@187.113.219.158) | |
| 08:21 | <alkisg> !vnc-dide | echo fnurl:
| |
| 08:21 | <ltsp`> fnurl: vnc-dide: To share your screen with me, run this: sudo apt-get --yes install x11vnc; x11vnc -connect srv1-dide.ioa.sch.gr - this is a reverse connection, it doesn't need port forwarding etc.
| |
| 08:24 | teacher has left IRC (teacher!c23fefeb@gateway/web/freenode/ip.194.63.239.235, Quit: Page closed) | |
| 08:29 | fnurl has left IRC (fnurl!3d3edc4a@gateway/web/freenode/ip.61.62.220.74, Ping timeout: 246 seconds) | |
| 08:34 | fnurl has joined IRC (fnurl!3cf8605f@gateway/web/freenode/ip.60.248.96.95) | |
| 08:34 | <fnurl> bacl, not sure what happened there :S
| |
| 08:35 | schlady has left IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de, Remote host closed the connection) | |
| 08:39 | schlady has joined IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de) | |
| 09:50 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 252 seconds) | |
| 09:50 | schlady has left IRC (schlady!~schlady@ip1f111304.dynamic.kabel-deutschland.de, Remote host closed the connection) | |
| 10:02 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 240 seconds) | |
| 10:02 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 10:06 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 10:28 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 260 seconds) | |
| 10:38 | solarnetone has left IRC (solarnetone!~scott@rrcs-67-79-182-229.se.biz.rr.com, Read error: No route to host) | |
| 10:40 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 10:57 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 260 seconds) | |
| 11:07 | lbssousa has joined IRC (lbssousa!~laercio@177.143.58.26) | |
| 11:10 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 11:34 | schlady has joined IRC (schlady!~schlady@141-53-223-35.ip.uni-greifswald.de) | |
| 11:35 | schlady has joined IRC (schlady!~schlady@141-53-223-35.ip.uni-greifswald.de) | |
| 11:37 | schlady has joined IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de) | |
| 11:37 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 246 seconds) | |
| 11:39 | schlady has left IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de, Read error: Connection reset by peer) | |
| 11:42 | schlady has joined IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de) | |
| 11:44 | schlady has left IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de, Read error: Connection reset by peer) | |
| 11:45 | schlady has joined IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de) | |
| 11:50 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 12:25 | Gremble has joined IRC (Gremble!~Ben@bmex-gw.bristolwireless.net) | |
| 12:25 | Gremble is now known as Guest96266 | |
| 12:34 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 250 seconds) | |
| 12:42 | <muppis> !raspberry
| |
| 12:42 | <ltsp`> I do not know about 'raspberry', but I do know about these similar topics: 'raspberrypi'
| |
| 12:42 | <muppis> !raspberrypi
| |
| 12:42 | <ltsp`> raspberrypi: (#1) Ubuntu/LTSP on Pi 2: https://help.ubuntu.com/community/UbuntuLTSP/RaspberryPi, or (#2) Debian/LTSP (with raspbian chroot) on Pi: http://cascadia.debian.net/trenza/Documentation/raspberrypi-ltsp-howto/, or (#3) unofficial Ubuntu/LTSP (with raspbian chroot) on Pi: http://pinet.org.uk/
| |
| 12:45 | schlady has left IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de, Read error: Connection reset by peer) | |
| 12:46 | schlady has joined IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de) | |
| 12:47 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 13:01 | danau11 has joined IRC (danau11!~durban@static-72-89-243-139.nycmny.fios.verizon.net) | |
| 13:06 | danau11 has left IRC (danau11!~durban@static-72-89-243-139.nycmny.fios.verizon.net) | |
| 13:51 | schlady has left IRC (schlady!~schlady@141-53-32-98.ip.uni-greifswald.de, Remote host closed the connection) | |
| 14:12 | FrankBlues has joined IRC (FrankBlues!~alex@c-67-171-127-41.hsd1.ut.comcast.net) | |
| 14:17 | Phantomas has joined IRC (Phantomas!~ftsamis@ubuntu/member/phantomas) | |
| 14:30 | larryni has left IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com, Ping timeout: 250 seconds) | |
| 14:43 | larryni has joined IRC (larryni!~larryni@host86-151-7-247.range86-151.btcentralplus.com) | |
| 14:55 | ben_roose has joined IRC (ben_roose!~roose@roose.cs.wichita.edu) | |
| 15:44 | hydruid has joined IRC (hydruid!9c6e22ba@gateway/web/freenode/ip.156.110.34.186) | |
| 15:45 | <hydruid> Good morning. What's the best approach so that a thin client can access their local hard drive?
| |
| 15:45 | I enabled LocalDev and Local_Storage in lts.conf but that didn't seem to work
| |
| 15:45 | I found this troubleshooting guide but it's dated, https://wiki.ubuntu.com/DebugLocalDev
| |
| 15:50 | myrdd has joined IRC (myrdd!~quassel@55d42f35.access.ecotel.net) | |
| 15:59 | alexxtasi has joined IRC (alexxtasi!5e40a53e@gateway/web/freenode/ip.94.64.165.62) | |
| 16:05 | Guest96266 has left IRC (Guest96266!~Ben@bmex-gw.bristolwireless.net, Ping timeout: 260 seconds) | |
| 16:06 | Freejack has left IRC (Freejack!~Freejack@unaffiliated/freejack, Ping timeout: 240 seconds) | |
| 16:08 | Freejack has joined IRC (Freejack!~Freejack@unaffiliated/freejack) | |
| 16:10 | alexxtasi has left IRC (alexxtasi!5e40a53e@gateway/web/freenode/ip.94.64.165.62, Changing host) | |
| 16:10 | alexxtasi has joined IRC (alexxtasi!5e40a53e@unaffiliated/alexxtasi) | |
| 16:10 | alexxtasi has joined IRC (alexxtasi!5e40a53e@gateway/web/freenode/ip.94.64.165.62) | |
| 16:17 | <||cw> hydruid: I haven't used that, but maybe it only works on usb disks? or only fat32? what is on the local disk?
| |
| 16:18 | <hydruid> ||cw: it's a local sata drive with ntfs
| |
| 16:18 | <||cw> if it's ntfs you probably need to add ntfs-3g package to the client image
| |
| 16:18 | <hydruid> ||cw: ahhh good point, thank you I didn't think of that
| |
| 16:18 | <||cw> unless that pipes the raw block device, which IIRC it doens't
| |
| 16:19 | <hydruid> ||cw: I know before it worked without a hitch but I think that was with a fat client
| |
| 16:20 | <||cw> fat client runs the session locally, so yeah
| |
| 16:20 | much easier to make that work
| |
| 16:21 | <hydruid> ||cw: yes it is, I'm in the process of building a fat client now, I'm limited on time
| |
| 16:21 | ||cw: thank you for your help, it's good to see the LTSP community still has helpful members
| |
| 16:25 | alexxtasi has left IRC (alexxtasi!5e40a53e@gateway/web/freenode/ip.94.64.165.62, Ping timeout: 246 seconds) | |
| 16:25 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 16:29 | <sbalneav> vagrantc: ping
| |
| 16:29 | <vagrantc> sbalneav: heya!
| |
| 16:30 | <sbalneav> http://bazaar.launchpad.net/~lightdm-team/lightdm-webkit-greeter/trunk/revision/24
| |
| 16:30 | The patches I came up with at BTS for making lightdm-webkit-greeter work properly with pam were accepted upstream.
| |
| 16:31 | If you're building fat clients, and want proper pam messages, you can now use lightdm-webkit-greeter.
| |
| 16:31 | Only problem is, there isn't a package in debian for it.
| |
| 16:31 | <vagrantc> with libpam-sshauth, or arbitrary pam modules?
| |
| 16:32 | <sbalneav> Both
| |
| 16:32 | Hence my patches :D
| |
| 16:32 | <vagrantc> nice!
| |
| 16:33 | <sbalneav> Do you know anyone (hint-hint) who'd sponsor it in debian?
| |
| 16:33 | It WORKS in debian, I'm using it now on our thin clients
| |
| 16:34 | But I had to do the old manual compile shuffle
| |
| 16:34 | <vagrantc> well, i can give it a look-over!
| |
| 16:34 | ah, you want packaging, too? :)
| |
| 16:35 | sbalneav: do you know if they're likely to make a "release" anytime soon?
| |
| 16:36 | <sbalneav> the 0.2 version was recently released
| |
| 16:36 | <vagrantc> but that seems to be missing some useful patches!
| |
| 16:36 | <sbalneav> http://bazaar.launchpad.net/~lightdm-team/lightdm-webkit-greeter/trunk/revision/21
| |
| 16:36 | Yeah
| |
| 16:36 | That's the problem.
| |
| 16:37 | * vagrantc git clones it | |
| 16:37 | <sbalneav> The next version of debian's a while off, so I'm not too worried about it getting in, like TOMORROW or anything.
| |
| 16:37 | I'll ask him if he can release a 0.2.1 or something.
| |
| 16:40 | <vagrantc> not urgent
| |
| 16:40 | but it would help to get it through the NEW queue
| |
| 16:42 | sbalneav: would probably be good to contact the team(s) already maintaining lightdm related packages.
| |
| 16:43 | or not, they're apparently still using svn
| |
| 16:44 | sbalneav: how confident are you it'll be maintainable in the long-term?
| |
| 16:45 | <sbalneav> Which, lightdm-webkit-greeter?
| |
| 16:46 | It's VERY simple; when you look at the C code, it's essentially a giant piece of plumbing, just mapping lightdm API calls onto calls to javascript.
| |
| 16:46 | The greeter's VERY simple.
| |
| 16:47 | Plus, the bonus for LTSP is: writing an "ltsp greeter" to ship with LTSP means we can dump LDM and just write a greeter that will work with all the pam stuff we want, AND it's just HTML and Javascript
| |
| 16:47 | No compiled code.
| |
| 16:47 | <vagrantc> ah, so we can ship a custom greeter using lightdm-webkit-greeter as an engine/backend?
| |
| 16:48 | <sbalneav> bingo.
| |
| 16:48 | It's a huge win/win
| |
| 16:48 | * vagrantc isn't entirely sure javascript is a selling point | |
| 16:48 | <vagrantc> but i guess it's what all the cool kids are using
| |
| 16:48 | <sbalneav> Well, it's better than compiled C.
| |
| 16:48 | <vagrantc> or all the cool teens, surely the cool kids have moved onto something else
| |
| 16:49 | <sbalneav> If we can get just about all of our "compiled code" down to libpam_sshauth, we'll be laughin.
| |
| 16:52 | <vagrantc> i've been wondering if libpam_sshauth is the way to go for a while ... if it wouldn't work better to figure out how to use LDAP, winbind, or other existing PAM infrastructure
| |
| 16:53 | or a sql-pam backend, or ...
| |
| 16:53 | hydruid has left IRC (hydruid!9c6e22ba@gateway/web/freenode/ip.156.110.34.186) | |
| 16:55 | <vagrantc> sbalneav: i saw that you "approved" some libpam-sshauth merge request ... did you end up merging it?
| |
| 16:57 | alkisg has thrown around a lot of ideas regarding some future design... it might not necessarily hinge on ssh so much
| |
| 16:57 | * vagrantc isn't quite sure what could spark some fresh LTSP6 development | |
| 16:58 | <vagrantc> but it seems like we're a little fragmented, mostly discussing development on irc whenever we happen to have ideas
| |
| 17:02 | <sbalneav> vagrantc: Nort use
| |
| 17:02 | Not yet
| |
| 17:02 | Yeah, well, I've been busy as hell for the last 3 -4 years. I keep meaning to jump back in.
| |
| 17:03 | * vagrantc nods | |
| 17:03 | <sbalneav> I'll get back to it.
| |
| 17:03 | I retire in 7 years 11 months.
| |
| 17:03 | <vagrantc> heh!
| |
| 17:03 | <sbalneav> I'm serious.
| |
| 17:04 | One of my plans for retirement is to spend a significant amount of time doing free software devel
| |
| 17:04 | <vagrantc> won't that be a time to run off into the woods and forget about computers? :)
| |
| 17:05 | <sbalneav> Winter: Free Software dev.
| |
| 17:05 | Summer: boating :D
| |
| 17:05 | <vagrantc> two seasons
| |
| 17:05 | <sbalneav> I live in Canada :D
| |
| 17:06 | There's "winter" and everything else :D
| |
| 17:09 | <vagrantc> here the seasons are basically wet and dry, though it's been a little confused the past couple years
| |
| 17:09 | alkisg: any thoughts on the lightdm-webkit-greeter mentioned above?
| |
| 17:11 | <maldridge> vagrantc: I haven't read the scrollback, but lightdm-webkit-greeter is old, mostly unmaintained, and very, *very* buggy
| |
| 17:12 | <vagrantc> maldridge: looks like it's had a burst of development since october ... prior to that commits back in 2012
| |
| 17:12 | <maldridge> ah, just read scrollback; you want no part of it
| |
| 17:12 | I'm getting ready to refactor it out of a large lab config set because it has a bad habit of not exiting
| |
| 17:13 | and has very little capability to expose fields to the greeter
| |
| 17:13 | bbiaf
| |
| 17:16 | <sbalneav> maldridge: Funny, becase as of right now, it's the ONLY greeter in which you have a chance of implementing the full pam conversation.
| |
| 17:18 | Neither lightdm-gtk-greeter nor lightdm-kde-greeter pass through the pam convo properly; they're expecting a userid/password -> pass/fail, and can't handle the pam stack asking subsequent questions like what will happen if you do password expiry and password quality checks via pam_cracklib or pam_passwdqc
| |
| 17:25 | vervelak has left IRC (vervelak!~vervelak@139.91.248.3, Ping timeout: 260 seconds) | |
| 17:26 | vervelak has joined IRC (vervelak!~vervelak@139.91.248.3) | |
| 17:28 | <vagrantc> sbalneav: giving a quick go at packaging...
| |
| 17:31 | <sbalneav> Cool
| |
| 17:36 | <maldridge> sbalneav: interesting; whie I care about those features, I care more that my users don't get launched into a session with a dead fullscreen white window with title "lightdm-webkit-greeter"
| |
| 17:37 | <sbalneav> I'm using it production right now.
| |
| 17:37 | Not seeing that at all.
| |
| 17:37 | <maldridge> interesting, any suggestions you can offer?
| |
| 17:38 | <sbalneav> I recently submitted patches.
| |
| 17:38 | <maldridge> we're running straight ubuntu 14.04, tons of problems with it not exiting
| |
| 17:39 | <sbalneav> One of the key features that wasn't implemented was the "in_authentication" callback, so you could tell in your scripts if it wasn't authenticated yet.
| |
| 17:39 | Also, they hadn't implemented a key lightdm feature which was authenticating with a NULL userid, thus causing PAM to do the prompting for the username.
| |
| 17:40 | http://bazaar.launchpad.net/~lightdm-team/lightdm-webkit-greeter/trunk/revision/24
| |
| 17:40 | <maldridge> hm, the thing I'm more interested in is pulling arbitrary information from 3rd sources; most notably we store loginShell in ldap at my site, and i don't currently have a way to have that follow users
| |
| 17:40 | <vagrantc> sbalneav: https://cascadia.debian.net/~vagrant/debian/pool/main/l/lightdm-webkit-greeter/
| |
| 17:40 | sbalneav: how do those look?
| |
| 17:40 | <sbalneav> I'm using libpam_ldapd and loginShell works for me.
| |
| 17:40 | vagrantc: looking
| |
| 17:41 | vagrantc: after lunch I'll install from your pkgs, and check.
| |
| 17:41 | In 10 minutes I gotta pop out for some errands.
| |
| 17:43 | maldridge: You're more than likely using the older version. There's been some new dev in the last couple of months. If you could either try vagrant's packages, or just do a bzr branch and compile of the upstream, I'd be interested to know if your problems go away.
| |
| 17:44 | <vagrantc> sbalneav: those packages were built on sid ... not sure if they'll install on jessie
| |
| 17:47 | <sbalneav> vagrantc: I can put together a sid vm
| |
| 17:49 | vervelak has left IRC (vervelak!~vervelak@139.91.248.3, Ping timeout: 240 seconds) | |
| 17:52 | <maldridge> an interesting suggestion, unfortunately local policy prevents us from installing locally built/maintained versions. Is there a PPA or similar available?
| |
| 17:54 | <vagrantc> not yet
| |
| 17:54 | vervelak has joined IRC (vervelak!~vervelak@139.91.248.3) | |
| 17:57 | vervelak has left IRC (vervelak!~vervelak@139.91.248.3, Read error: Connection reset by peer) | |
| 18:19 | <bennabiy> vagrantc, sbalneav: That is very encouraging. I would like to see how this progresses. I might be getting more time soon to be able to work on LTSP in the somewhat near future (within the next couple months).
| |
| 18:21 | <alkisg> vagrantc: about the greeter, I don't know, I haven't understood the pam stuff well enough yet, I thought we wanted to use pam because it would work with any DM, if it'll only work on a single DM we might as well keep ours...
| |
| 18:21 | About ltsp6 ideas etc, we're starting to work on configd with Phantomas
| |
| 18:21 | Initially we'll use it alongside ltsp,
| |
| 18:22 | and after a couple of years it might be good enough to completely replace ltsp-client in a smooth migration, whenever each site is ready for it
| |
| 18:22 | configd => complements/replaces lts.conf, ldminfod, config-client => initially complements ltsp-client, in the future it may replace it
| |
| 18:23 | People may also use that outside of ltsp, without netbooted pcs
| |
| 18:23 | There's too much legacy in ltsp, it's difficult to care about backwards compatibility etc when you want a major upgrade
| |
| 18:24 | So a parallel configuration system will work best imho
| |
| 18:27 | Server-side=python/https, client-side=wget+shell script sourcing. We'll clearly define some phases where the client contacts the server, similar to ltsp-cluster, and as part of the configuration it'll be possible for the sysadmin to also setup forward ssh connections to the clients for tools like "exec-on-clients"
| |
| 18:27 | <bennabiy> alkisg: where is more info on configd / config-client?
| |
| 18:27 | <alkisg> bennabiy: there's none; the initial thought was for ltspd, but I think there's too much legacy there, http://wiki.ltsp.org/wiki/Dev:Ltspd
| |
| 18:28 | So that part might be aborted
| |
| 18:28 | We're now trying to document configd in a private google doc (will be a wiki in the future, of course), and have a few code snippets around
| |
| 18:29 | It's the part where we mostly chat for hours to define the infrastructure, then code a bit for proof of concept
| |
| 18:30 | <bennabiy> That is great
| |
| 18:31 | <vagrantc> alkisg: yeah, ideally what we'd have would work with any display manager ... but like you've suggested before, having a sane default display manager would be nice.
| |
| 18:31 | <bennabiy> Dev for it going to be in launchpad, or through something like github?
| |
| 18:32 | <alkisg> Something between those 2, I don't think it's very significant, as long as translations are easy for end users
| |
| 18:32 | <bennabiy> base language going to be greek? ;)
| |
| 18:33 | <alkisg> Haha, no, not at all :)
| |
| 18:34 | <bennabiy> I keep trying to find a reason to learn greek, but have only learned a little (I get too easily distracted by other languages)
| |
| 18:35 | <alkisg> Greek is difficult though, we have too much grammar... I think you'd need to find someone to talk with in greek in order to get motivated
| |
| 18:36 | <bennabiy> Yes, which is part of my problem with languages... without someone to communicate with, no reason to learn to communicate.
| |
| 18:36 | <alkisg> For us it's a bit difficult when we're developing in english, and there's no native english speaker in the team, and we're not sure if the template english strings that we're sending to the translators are good enough
| |
| 18:37 | <bennabiy> english is about 1/2 greek at least anyways
| |
| 18:38 | <alkisg> Well, "difficult" englsh words like empathy or anthropology are easy for us, but easy, latin-based ones initially are not
| |
| 18:38 | <bennabiy> so, would you say it is all latin to you, if it is all greek to us?
| |
| 18:39 | <alkisg> Here we say "it's chinese to us" :)
| |
| 18:39 | <bennabiy> HAH
| |
| 18:39 | <alkisg> Those even have completely different writing systems.... chaos!
| |
| 18:40 | <bennabiy> Heh. I like Japanese, not so much chinese
| |
| 18:40 | Japanese has 4 writing systems
| |
| 18:40 | all interchangeable (to some degree)
| |
| 18:41 | <alkisg> I think thats a real challenge, to learn japanese, chinese, or something of those arabic-based writing systems where you don't even understand where one letter ends and the other begins
| |
| 18:42 | ...but again, no real motive for that other than curiosity :D
| |
| 18:42 | <bennabiy> Arabic confuses me (writing wise) but I understand some arabic because of its similarity to hebrew
| |
| 18:43 | <alkisg> The hebrew character (letter) differences are to subtle for me, I think many of them are the same!
| |
| 18:43 | *too
| |
| 18:47 | <bennabiy> They can get confusing
| |
| 18:47 | especially when the print starts to rub off, or poor copy
| |
| 18:47 | <alkisg> vagrantc (and the others): feature-wise, (not caring about how it will be implemented), what are you missing from ltsp?
| |
| 18:49 | <maldridge> alkisg: reliable documentation for how to customize; reliable method for implementing kiosk computing
| |
| 18:49 | (that doesn't involve generating a ton of accounts)
| |
| 18:49 | * bennabiy seconds documentation... | |
| 18:50 | * vagrantc concurs that documentation is lacking | |
| 18:50 | <alkisg> Right. So, while ltsp developers think of ltsp6 as a chance to clean up the code and have it better designed and more maintainable etc, from the user side, that won't offer anything...
| |
| 18:50 | I think that's one of the reasons that ltsp6 is advancing so slow...
| |
| 18:51 | <vagrantc> main feature is implementing better guest-login functionality ...
| |
| 18:51 | <bennabiy> vagrantc: I was thinking something similar
| |
| 18:52 | <maldridge> that! I want ot have a lab of ~35 machines but allowing a single volatile account to log in everywhere
| |
| 18:52 | <vagrantc> i've cobbled together a few parts to have anonymous logins with lightdm, with a local tmpfs homedir, no authentication or homedir mounting on the server
| |
| 18:52 | <bennabiy> I would like to have a guest login that has a core set of files available, and restrictions put on it, but that nothing from the home directory lasts past the session
| |
| 18:52 | <alkisg> E.g. why would switching to the lightdm-webkit-greeter help us with guest accounts? With the current ltsp code base, I don't think guest accounts are more than 50 lines of code away...
| |
| 18:53 | <vagrantc> alkisg: you mean with LDM, or with lightdm, or what?
| |
| 18:53 | <maldridge> well, I for one would enjoy lightdm-webkit because then I could plug in my existing greeter skin from our fat installs
| |
| 18:53 | <alkisg> With the current LDM code base
| |
| 18:53 | <maldridge> which I am unwilling to port back to a gtkrc based theme
| |
| 18:53 | <vagrantc> i found it was easier to implement with libpam-mount & lightdm than implementing in LDM
| |
| 18:53 | * bennabiy would like to do away with gtkrc... | |
| 18:54 | <vagrantc> alkisg: i would like to move to a more standard display manager so that themeing can be shared with other projects
| |
| 18:54 | maybe lightdm-webkit will just make that it's own special themeing again...
| |
| 18:55 | <bennabiy> vagrantc - alkisg: what currently are the "inflexible" parts of ltsp? Where does the current LTSP not allow customization or flexibility towards configuration?
| |
| 18:55 | LDM and...
| |
| 18:56 | <maldridge> one thing that I am dissapointed with currently is how hard it is to script ltsp deployment
| |
| 18:56 | <alkisg> Oh, it's very flexible. It's just not always easy for non-developers to find where to put that code you want :)
| |
| 18:56 | <maldridge> there seem to be too many systems available and no clear winner
| |
| 18:56 | * bennabiy nods | |
| 18:56 | <alkisg> What do you mean "ltsp deployment"?
| |
| 18:56 | ...script what, ltsp-build-client?
| |
| 18:56 | <maldridge> alkisg: I use ansible, I'd like to know exactly what the best method is, and how to update the things within it
| |
| 18:57 | i.e. chroot vs. pnp / nfs vs. nbd
| |
| 18:57 | <alkisg> maldridge: ansible is about maintaining a lot of pcs. ltsp is about maintaining one pc...
| |
| 18:57 | <bennabiy> I more mean what would stand in the way of implementing customizations? I said LDM, because it is hard-coded.
| |
| 18:57 | <vagrantc> maldridge: my most recent LTSP deployment is all done with ansible
| |
| 18:57 | <maldridge> alkisg: yeah, but I manage all of my ltsp servers along with my other servers
| |
| 18:57 | <bennabiy> is it the only compiled code in LTSP?
| |
| 18:57 | <maldridge> vagrantc: are your playbooks public anywhere?
| |
| 18:58 | <vagrantc> maldridge: no, they're too site-specific at the moment ... first project with ansible, need to clean it up a bit
| |
| 18:58 | or rewrite it from scratch :)
| |
| 18:58 | <alkisg> maldridge: I don't understand how ltsp would help ansible replicate ltsp servers
| |
| 18:58 | <maldridge> ok, I've got mine on github, and while they work, I haven't figured a good way yet to determine if there needs to be an image rebuild
| |
| 18:58 | <alkisg> You can netboot ltsp servers from a golden ltsp server template, but that's not related to ansible...
| |
| 18:59 | <maldridge> alkisg: basically a clear pipeline that can be followed to stand up an ltsp server (see above comment about multiple paths) and then a clear pipeline for keeping that server updated once its up (see above comment about chroots)
| |
| 18:59 | <vagrantc> maldridge: ah, yes, the image rebuild i just manually remove the image file if i want to trigger a rebuild
| |
| 19:00 | <maldridge> ah, so you are switching on stat.exists==true then?
| |
| 19:00 | <vagrantc> maldridge: also, just doing ltsp-pnp-style
| |
| 19:00 | maldridge: checks if the file exists
| |
| 19:01 | <maldridge> ok, makes sense
| |
| 19:01 | <alkisg> The "multiple paths to install an ltsp server" is getting in the way of selecting one of those paths?
| |
| 19:01 | E.g. ltsp-build-client is needed for raspberry pis, while ltsp-pnp is suited for plain i386 installations...
| |
| 19:01 | <maldridge> since there is very arguably no clear winner, I find writing lots of special scripts and special casing for various labs
| |
| 19:01 | <alkisg> I don't think it will be an improvement to limit the methods there
| |
| 19:01 | <vagrantc> alkisg: my biggest frustration with NBD is when doing development, i can quickly run out of disk space because of deleted image files holding open inodes...
| |
| 19:01 | <maldridge> for example, nfs vs nbd
| |
| 19:02 | vagrantc: I haven't had that issue on btrfs, but ymmv
| |
| 19:02 | <alkisg> vagrantc: then use nbd + btrfs or ext4
| |
| 19:02 | Not squashfs
| |
| 19:02 | <vagrantc> fair enough
| |
| 19:02 | NFS vs. NBD may be a moot point if overlay FS in mainline linux continues to be broken with NFS as a backend
| |
| 19:03 | <maldridge> alkisg: I'm not suggesting limited choices, I'm suggesting a clear comparison of why each exists and why I would like to use each of them
| |
| 19:03 | <alkisg> Documentation again then
| |
| 19:03 | <vagrantc> the out-of-the box experience does lead one right to squashfs ... which for the most part works fine ...
| |
| 19:04 | <alkisg> I think that's mostly up to "good users" than to the ltsp developers, though
| |
| 19:04 | <vagrantc> alkisg: well, good developers could document the code when they implement it...
| |
| 19:04 | at least a stub that can be expanded on
| |
| 19:05 | * vagrantc is to blame as much as anyone | |
| 19:05 | <vagrantc> easiest time to document a feature is when it goes into the VCS
| |
| 19:05 | <alkisg> If they get paid for it, sure. If they have limited time for it, it's better to spend it on coding and let someone else document, with some help for it
| |
| 19:05 | <vagrantc> true enough
| |
| 19:05 | <bennabiy> at least even something which could prompt someone to pop in here and ask "What does this option do?" rather than having to know where in the code to look to try to implement feature x
| |
| 19:06 | <alkisg> At one time, I suggested in the edubuntu list that I would help people with their issues as long as they documented the solutions in the wiki
| |
| 19:07 | The result was that a single person was interested :)
| |
| 19:07 | <vagrantc> heh
| |
| 19:07 | tech support with a documentation agreement
| |
| 19:07 | <alkisg> Many, many floss users just don't want to really contribute back
| |
| 19:07 | <vagrantc> alkisg: you're too helpful
| |
| 19:08 | i suspect many don't have the confidence that they could actually contribute back
| |
| 19:08 | <maldridge> ^^
| |
| 19:08 | * vagrantc has seen it enough | |
| 19:08 | <maldridge> the fear of failure is a very strong thing among people just coming into dev from an operations side
| |
| 19:08 | <alkisg> That's true in several cases, but there are also many cases where people could contribute and just don't want to
| |
| 19:09 | <maldridge> can you produce an example?
| |
| 19:09 | <alkisg> Sure, localization
| |
| 19:09 | <bennabiy> I would like to contribute, but often run into time conflicts
| |
| 19:09 | <alkisg> Most greeks know english well enough to help there
| |
| 19:09 | Every time a new gnome comes out though, it's very difficult to find contributors for localization
| |
| 19:09 | <bennabiy> perhaps there is something I could provide towards documentation
| |
| 19:10 | I guess I just do not know the best way to go about it
| |
| 19:10 | <maldridge> personally, for ltsp6 I would like to see a shift away from dnsmasq
| |
| 19:10 | <alkisg> For example, maldridge asked for guest accounts
| |
| 19:10 | I could point him to 2-3 solutions we've already given in the ML, or to 2-3 other ideas
| |
| 19:10 | And even code some new ones
| |
| 19:11 | ...then he could put the solutions in the ltsp.org wiki
| |
| 19:11 | myrdd has left IRC (myrdd!~quassel@55d42f35.access.ecotel.net, Remote host closed the connection) | |
| 19:11 | <alkisg> finally, when someone comes here, we could just do:
| |
| 19:11 | <maldridge> alkisg: sure, and I think I've currently got one of the ML ones in use, but the fact remains that guest accounts have been asked for enough at this point to probably be an actual feature
| |
| 19:11 | <bennabiy> My times are scattered. I would not mind updating the wiki with a few things which I know about, but if there was a way to even know what needs to be updated, or how to go about it...
| |
| 19:11 | <alkisg> !guest-accounts
| |
| 19:11 | <ltsp`> Error: "guest-accounts" is not a valid command.
| |
| 19:11 | <alkisg> ...and point him there
| |
| 19:12 | Someone would have to search the wiki, search what people are asking help for, and document what needs to be done etc
| |
| 19:12 | <maldridge> actually no, I'm logging everyone in under a single account to get around a flaw in some closed source code
| |
| 19:12 | * maldridge is at lunch | |
| 19:12 | <bennabiy> alkisg, vagrantc: how does one get write access to the wiki? What is the way to go about it?
| |
| 19:12 | <alkisg> Guest accounts shouldn't relate to the server except for temporary disk space, not for authentication
| |
| 19:12 | It's already open (the wiki)
| |
| 19:13 | You just sign up or something
| |
| 19:13 | <bennabiy> ok
| |
| 19:13 | I probably did that at one point
| |
| 19:13 | * bennabiy cant remember yesterday... | |
| 19:13 | <alkisg> :)
| |
| 19:13 | (09:10:18 μμ) maldridge: personally, for ltsp6 I would like to see a shift away from dnsmasq ==> why is that?
| |
| 19:13 | <bennabiy> Alkisg: when are you coming to the states again?
| |
| 19:13 | <alkisg> For of all, dnsmasq is the only linux software that supports proxydhcp
| |
| 19:14 | bennabiy: no idea, but I'd love to come when I get the chance
| |
| 19:14 | <bennabiy> alkisg: first of all?
| |
| 19:14 | <alkisg> *first, yes sorry
| |
| 19:14 | There's something wrong when I'm trying to type in english while thinking something else in english... my hands type whatever they want :)
| |
| 19:15 | <bennabiy> hah!
| |
| 19:15 | Happens to english speakers as well
| |
| 19:15 | <vagrantc> stick to typing in C.UTF-8, then
| |
| 19:15 | <alkisg> :P
| |
| 19:15 | * bennabiy laughs | |
| 19:17 | * alkisg waves for now, bye all! | |
| 19:18 | alkisg is now known as work_alkisg | |
| 19:22 | * bennabiy waves. | |
| 19:22 | <bennabiy> Time for me to depart as well
| |
| 19:22 | Back either tomorrow or Friday.
| |
| 19:29 | lbssousa has left IRC (lbssousa!~laercio@177.143.58.26, Quit: lbssousa) | |
| 19:35 | Phantomas has left IRC (Phantomas!~ftsamis@ubuntu/member/phantomas) | |
| 19:36 | <maldridge> work_alkisg: it does not, imho, subscribe to the UNIX philosophy of do one thing and do it well, instead allowing multiple services to die at once from an issue with a single daemon
| |
| 19:43 | FrankBlues has left IRC (FrankBlues!~alex@c-67-171-127-41.hsd1.ut.comcast.net, Ping timeout: 240 seconds) | |
| 19:43 | Frank_Blues has joined IRC (Frank_Blues!~alex@c-67-171-127-41.hsd1.ut.comcast.net) | |
| 19:43 | Frank_Blues is now known as FrankBlues | |
| 20:49 | epoptes_user5 has joined IRC (epoptes_user5!4d1c1c5d@gateway/web/freenode/ip.77.28.28.93) | |
| 20:50 | epoptes_user5 has left IRC (epoptes_user5!4d1c1c5d@gateway/web/freenode/ip.77.28.28.93) | |
| 20:58 | epoptes_user2 has joined IRC (epoptes_user2!6c3de413@gateway/web/freenode/ip.108.61.228.19) | |
| 21:03 | <epoptes_user2> Can Epoptes be used remotely? I have eight children (not mine) with Edubuntu on laptops that I need to be abale to keep tabs on and administrate their computers. Yet we are in four different locations with internet access.
| |
| 21:05 | also i see i mayne in the wrong channel
| |
| 21:08 | danau11 has joined IRC (danau11!~durban@static-72-89-243-139.nycmny.fios.verizon.net) | |
| 21:12 | danau11 has left IRC (danau11!~durban@static-72-89-243-139.nycmny.fios.verizon.net, Ping timeout: 244 seconds) | |
| 21:21 | epoptes_user2 has left IRC (epoptes_user2!6c3de413@gateway/web/freenode/ip.108.61.228.19) | |
| 21:43 | work_alkisg is now known as alkisg | |
| 21:45 | <alkisg> vagrantc: I'm trying to install lubuntu-desktop in an armhf chroot, and it *indirectly* pulls some other packages like unity-control-center, which do not normally appear in a lubuntu installation.
| |
| 21:45 | Do you happen to know any apt command that will tell me which package it is that Recommends: unity-control-center?
| |
| 21:45 | I think `aptitude why` can do that but only after they are actually installed...
| |
| 21:46 | I'm reluctant to use --no-install-recommends because that may prevent other important packages from being installed..
| |
| 21:54 | <vagrantc> alkisg: apt-cache rdepends sort of works, though poorly ...
| |
| 21:54 | it treats recommends, depends and maybe even suggests as the same thing
| |
| 21:54 | * alkisg is trying a `for each package in $Depends; do apt-get install --simulate $package | grep unity`... | |
| 21:55 | <alkisg> thanks, I'll also check rdepends
| |
| 21:59 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |
| 21:59 | cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 260 seconds) | |
| 22:10 | <alkisg> lubuntu-desktop depends on lightdm, which recommends unity-greeter | kde-greeter, but not lightdm-gtk-greeter which is the one used by lubuntu... I'll install lightdm first without recommends
| |
| 22:14 | Nah, it's pulling a lot more packages that aren't there in a normal lubuntu installation
| |
| 22:14 | So, `apt-get install --no-install-recommends lubuntu-desktop^ lubuntu-core^` is a much better way to do that :)
| |
| 22:14 | <vagrantc> what's the ^ ?
| |
| 22:15 | <alkisg> Install a task's packages
| |
| 22:15 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 22:15 | <alkisg> We should make a note of that ^ somewhere for LTSP lubuntu users that use chroots instead of ltsp-pnp...
| |
| 22:20 | vagrantc, ogra_, so... I've found a cool way to boot ltsp clients without a network connection by using a local copy of the squashfs image. That's fine for raspberry pis, so that the SD card is always in read-only mode, whether they're booting from LAN or from the local SD.
| |
| 22:20 | But I'm having an issue in displaying a boot menu for users to select what to boot, ltsp or local
| |
| 22:21 | ...does it make sense to display a menu inside the initramfs, from e.g. init-premount/ltsp?
| |
| 22:21 | (and then modify cmdline with bind-mounts... to strip nbdroot...)
| |
| 22:21 | Is there any other way to display a boot menu in raspbery pis or similar devices?
| |
| 22:21 | <vagrantc> using u-boot ...
| |
| 22:22 | <alkisg> The usb keyboard doesn't seem to work with u-boot
| |
| 22:22 | I think they only support the high speed mode, for the usb network card...
| |
| 22:22 | <vagrantc> could be
| |
| 22:22 | * vagrantc notes that there are no similar devices | |
| 22:22 | <alkisg> The other way I've thought of, is to detect if the eth0 link is up, then boot into ltsp, otherwise boot locally....
| |
| 22:22 | Haha
| |
| 22:23 | So that the users would need to unplug the network cable to boot locally
| |
| 22:23 | <vagrantc> ouch
| |
| 22:23 | <alkisg> Currently, I only have those 2 solutions, cable or initramfs menu...
| |
| 22:24 | * vagrantc wonders if there's a way to reconfigure initramfs-tools without tweaking /proc/cmdline ... | |
| 22:24 | <alkisg> I couldn't think of any. Well, other than sed'ing though a lot of scripts which would be impossible to maintain
| |
| 22:26 | <vagrantc> ah, because so many scripts check /proc/cmdline
| |
| 22:26 | <alkisg> RIght
| |
| 22:27 | * alkisg imagines the boot script message... "Please remove your ethernet cable within 5...4..3..2..1... seconds if you want to boot locally..." | |
| 22:27 | <alkisg> :P
| |
| 22:27 | <vagrantc> and then plug it back in when you want internet access...
| |
| 22:27 | <alkisg> Yes, it's just for a second
| |
| 22:28 | * vagrantc wonders if there's a way to properly modify /proc/cmdline ... | |
| 22:28 | <vagrantc> seems... risky.
| |
| 22:29 | if kexec worked, that'd be an interesting workaround
| |
| 22:30 | alkisg: i think NOOBs does some crazy initramfs trickery ... but it may have an entirely custom initramfs... and is really just selecting different boot partitions/images
| |
| 22:30 | <alkisg> vagrantc: I think noobs has a custom bootcode.bin, that's loaded before u-boot/the kernel etc, and it loads recovery.elf instead of kernel.img
| |
| 22:30 | <maldridge> vagrantc: last I checked its doing some initramfs trickery and curling system files on the fly
| |
| 22:31 | <alkisg> So it's a menu before linux, too early...
| |
| 22:32 | kexec will be wonderful when it's completely stable
| |
| 22:32 | <vagrantc> why is that too early? that sounds perfect?
| |
| 22:32 | <alkisg> it's a binary blob, completely unmodifiable by me
| |
| 22:32 | No scripting/menuing capabilities
| |
| 22:32 | <vagrantc> how did they modify it?
| |
| 22:32 | <alkisg> They have the source code...
| |
| 22:33 | * vagrantc shrugs | |
| 22:33 | <alkisg> Well, if I wanted to go down to modifying their own bootloader, sure, but they don't even have a menu system, it'd take a heck of a lot of work to do it
| |
| 22:33 | They only load recovery if they can't find kernel.img
| |
| 22:33 | They don't present a choice to the user
| |
| 22:34 | <vagrantc> NOOBs definitely presented a boot menu ... though i thought it was in the initramfs
| |
| 22:34 | <alkisg> Later on the check for kernel.img on a second partition or something, and the prefer that one over the recovery in the first partition
| |
| 22:34 | The graphical menu is after the initramfs, much later than the bootloader
| |
| 22:35 | It's in the image.fs or however the are calling it, I don't remember
| |
| 22:35 | Similar to our i386.img
| |
| 22:35 | the check... the prefer ==> I meant they, meh...
| |
| 22:37 | <vagrantc> alkisg: could you boot differently if a file were present or missing? would that be good enough?
| |
| 22:37 | <alkisg> With the same /proc/cmdline mangling from init-premount? Sure, any input will do, either the cable, a mouse click or a mouse completely removed, or a file missing etc...
| |
| 22:38 | But how would the user for a file to be missing?
| |
| 22:38 | *force
| |
| 22:56 | <vagrantc> alkisg: well, the boot menu would involve mounting the SD card ...
| |
| 22:56 | alkisg: what's the use-case where you would want to both boot locally and over network?
| |
| 22:56 | <alkisg> vagrantc: I have a pi that I can test ltsp with, or I can also use it as a standalone media player...
| |
| 22:58 | Or just imagine a classroom where the ltsp server has failed, a hardware problem etc, until it's solved (could take days) why shouldn't they be able to use the pis
| |
| 22:59 | Actually, I could just put a panic handler there, to catch the "nbd mount failed", and continue with local boot...
| |
| 22:59 | <vagrantc> it's a lot of workarounds to support one buggy piece of hardware...
| |
| 22:59 | just happens to be a very popular piece of buggy hardware
| |
| 23:00 | <alkisg> The live local ltsp part is not specific to ltsp
| |
| 23:00 | It has a lot of benefits. It's an easily customizable live system, with an lts.conf and everything
| |
| 23:01 | E.g. imagine again the ltsp server failing, and the students using their desktop pcs as local "ltsp clients" from their local hard disk
| |
| 23:01 | That's only one command away, a dd if=/dev/nbd0 of=/dev/sda, from epoptes... provided the clients do have disks
| |
| 23:02 | <vagrantc> sure, i see value in those sorts of systems
| |
| 23:02 | but that's a lot easier to implement on x86, or even other arm boards ...
| |
| 23:03 | <alkisg> The scripts are just bash
| |
| 23:03 | <vagrantc> well, on x86, you could implement it in grub, or some other real bootloader
| |
| 23:03 | <alkisg> Nothing pi-specific there... the problem with pi is that it doesn't have a boot manager... in PCs I would have a grub menu, boot ltsp from the network or locally
| |
| 23:04 | * vagrantc nods | |
| 23:06 | <alkisg> Btw, /me would love it if our squashfs image did have a boot manager and a boot sector embedded, i.e. if it was directly dd'able
| |
| 23:06 | I don't think squashfs supports that though
| |
| 23:07 | <vagrantc> we'd have to have a separate partition for that
| |
| 23:07 | unless there's a bootloader that supports reading from squashfs
| |
| 23:09 | <alkisg> Squashfs could include an MBR, sure
| |
| 23:10 | Hmm a /boot partition would be needed, you're right
| |
| 23:10 | grub can read from squashfs, but it needs to be loaded first
| |
| 23:11 | OK then, again an ltsp-dd script to the rescue :)
| |
| 23:13 | vagrantc: I have a few of such scripts in sch-scripts, should I put them upstream in ltsp instead?
| |
| 23:14 | E.g. a client-side tool to dump the nbd image to the local disk, would that belong in upstream ltsp/
| |
| 23:14 | ?
| |
| 23:16 | vervelak has joined IRC (vervelak!~vervelak@139.91.248.3) | |
| 23:17 | <alkisg> !client-list
| |
| 23:17 | <ltsp`> client-list: to get a list of all nbd-clients (which sometimes is the same as ltsp clients), run: netstat -tn | sed -n 's/.*:10809 *\([0-9.]*\):.*/\1/p' | sort -Vu
| |
| 23:19 | alkisg is now known as work_alkisg | |
| 23:20 | <vagrantc> work_alkisg: a client-side tool to generate a bootable image you can drop on disk sounds useful to me...
| |
| 23:35 | FrankBlues has left IRC (FrankBlues!~alex@c-67-171-127-41.hsd1.ut.comcast.net, Ping timeout: 240 seconds) | |