IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 1 January 2014 (all times are UTC)
| 00:00 | <vagrantc> so, i should clean up the XS80-x-modes mess i just committed
| |
| 00:01 | and then upload?
| |
| 00:01 | oh, i wanted to fix resolvconf handling
| |
| 00:01 | <alkisg> Sounds good to me!
| |
| 00:01 | <vagrantc> alkisg: at least in debian, the resolvconf init script wipes out all our configs generated from init-ltsp.d
| |
| 00:02 | <alkisg> Ouch
| |
| 00:02 | So we should put it in RM_SYSTEM_SERVICES?
| |
| 00:02 | <vagrantc> er, rather, it wipes out resolvconf stuff generated in 10-resolv-conf
| |
| 00:03 | and it's been doing that since wheezy at least
| |
| 00:03 | <alkisg> vagrantc: maybe we need to put our configuration in /etc then instead of /run ?
| |
| 00:04 | <vagrantc> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=701647
| |
| 00:04 | * alkisg doesn't know exactly how resolvconf does its things... | |
| 00:04 | <vagrantc> yeah, i don't either
| |
| 00:04 | <alkisg> Maybe when it starts, it deletes everything under /run, but it should always respect /etc/resolvconf...
| |
| 00:06 | <vagrantc> maybe we drop an /etc/resolvconf/update.d script
| |
| 00:06 | or in /etc/resolvconf/resolv.conf.d/
| |
| 00:08 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 260 seconds) | |
| 00:14 | stgraber_ is now known as stgraber | |
| 00:15 | * vagrantc waves to stgraber | |
| 00:15 | * stgraber waves back to vagrantc | |
| 00:15 | <stgraber> and now time to head to bed (temporarily in Europe, so 1am here already)
| |
| 00:16 | <vagrantc> welll, apparently writing to /etc/resolvconf/resolv.conf.d/base works
| |
| 00:17 | dunno if that's bad for other reasons
| |
| 00:17 | <alkisg> And resolvconf then puts it in /run... sounds OK
| |
| 00:17 | <vagrantc> i think i'm going to commit that.
| |
| 00:17 | unless ubuntu behaves differently?
| |
| 00:19 | <alkisg> I don't think so
| |
| 00:19 | The manpage mentions "base", and it exists, it's an empty file
| |
| 00:26 | Time for bed here too... happy last 2013 coding and a happy new year :)
| |
| 00:26 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 00:30 | gbaman_ has joined IRC (gbaman_!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 00:33 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 245 seconds) | |
| 00:52 | gdi2k has joined IRC (gdi2k!~gdi2k@222.127.254.113) | |
| 01:35 | lee_ is now known as lee | |
| 02:19 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.154.124) | |
| 02:21 | PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.154.124, Client Quit) | |
| 02:22 | gbaman_ has left IRC (gbaman_!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Remote host closed the connection) | |
| 02:35 | gdi2k has left IRC (gdi2k!~gdi2k@222.127.254.113, Ping timeout: 272 seconds) | |
| 03:12 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 03:14 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Read error: Connection reset by peer) | |
| 03:14 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 03:23 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 260 seconds) | |
| 04:19 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 04:23 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 240 seconds) | |
| 05:10 | lnsand has joined IRC (lnsand!~AndChat70@173-16-199-235.client.mchsi.com) | |
| 05:18 | lnsand has left IRC (lnsand!~AndChat70@173-16-199-235.client.mchsi.com, Remote host closed the connection) | |
| 05:20 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 05:24 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 246 seconds) | |
| 05:44 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 05:49 | <alkisg> Good morning
| |
| 05:57 | <vagrantc> heh
| |
| 05:57 | happy new year ...
| |
| 05:58 | alkisg: well, i've rewritten that modeline support patch a bit
| |
| 06:04 | <alkisg> vagrantc: pastebin?
| |
| 06:05 | <vagrantc> alkisg: whole script, or diff?
| |
| 06:05 | <alkisg> Better whole script
| |
| 06:06 | <vagrantc> alkisg: http://paste.debian.net/73599/
| |
| 06:06 | i think that's the last thing i want to bother to test...
| |
| 06:07 | <alkisg> Just for the fun of it, env | sed | sort is faster of env | sort | sed
| |
| 06:08 | printf "\"$mode\" " => better always use "%s" in printf
| |
| 06:09 | <vagrantc> alkisg: urg. removed some of the %s stuff based a previous comment of yours ... but maybe i misunderstood
| |
| 06:09 | alkisg: i tried doing it with only one env loop, but i couldn't get the variables back out again...
| |
| 06:10 | <alkisg> Yeah I wasn't talking about %s then, I was talking about printf not being needed to read the variables
| |
| 06:10 | <vagrantc> while seems to make them local in scope
| |
| 06:10 | <alkisg> That's when the EOF is needed
| |
| 06:10 | If you want the variables value to be remembered, then don't use a subshell (| while), but <<EOF
| |
| 06:11 | read -r mode modeline <<EOF
| |
| 06:11 | $s
| |
| 06:11 | EOF
| |
| 06:13 | <vagrantc> don't understand how to put the output from "env | sort | sed" to that
| |
| 06:14 | alkisg: i used "env | sed | sort" first, but it stripped out the variable names and only sorted based on the values
| |
| 06:14 | <alkisg> Ah you're right my bad sorry
| |
| 06:14 | read <<EOF
| |
| 06:14 | $(command)
| |
| 06:14 | EOF
| |
| 06:14 | <vagrantc> while read -r mode modeline <<EOF
| |
| 06:15 | $(command)
| |
| 06:15 | EOF
| |
| 06:15 | ; do ...
| |
| 06:15 | ?
| |
| 06:16 | admittedly, i've been fighting minutiae all day, maybe i should just push what i've tested
| |
| 06:16 | gdi2k has joined IRC (gdi2k!~gdi2k@222.127.254.113) | |
| 06:17 | <alkisg> Sure, I don't think such small implementation details matter
| |
| 06:18 | while read -r mode modeline ; do
| |
| 06:18 | mode=xx etc
| |
| 06:18 | done <<EOF
| |
| 06:18 | $(env | sort | sed -n '/^X_MODE_[0-9]=/s///p')
| |
| 06:18 | EOF
| |
| 06:18 | Also, the "eval" part isn't needed, you can just define the function
| |
| 06:20 | <vagrantc> then i could just do a single while loop ... that seems better.
| |
| 06:20 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 06:23 | <alkisg> vagrantc: here's a new trick I just found out:
| |
| 06:23 | exec 5<&0
| |
| 06:23 | exec <<EOF
| |
| 06:23 | whoami
| |
| 06:23 | sleep 1
| |
| 06:23 | exec 0<&5
| |
| 06:23 | EOF
| |
| 06:23 | * vagrantc is terrified | |
| 06:23 | <alkisg> That allows us to execute strings instead of using eval, thus avoiding all the weird escaping that eval needs
| |
| 06:24 | So we can have e.g.:
| |
| 06:24 | display_hacks="$display_hacks
| |
| 06:24 | some command | instead of a function"
| |
| 06:25 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 240 seconds) | |
| 06:25 | * alkisg was looking for something similar since a long time... that would be useful in epoptes too | |
| 06:39 | <vagrantc> hrm.
| |
| 06:40 | <alkisg> Meh it only works with interactive shells, so it's only good for epoptes
| |
| 07:04 | gdi2k has left IRC (gdi2k!~gdi2k@222.127.254.113, Ping timeout: 272 seconds) | |
| 07:06 | <vagrantc> almost got this modeline support cleaned up
| |
| 07:08 | or maybe sent it into an infinite loop
| |
| 07:08 | * alkisg feels responsible for vagrantc taking so much time with it and thinks he shouldn't have provided so much feedback... it's just wasn't that significant, just silly coding ideas... | |
| 07:08 | <alkisg> pastebin?
| |
| 07:09 | <vagrantc> the code is much better...
| |
| 07:09 | shorter, easier to read...
| |
| 07:14 | alkisg: http://paste.debian.net/73600
| |
| 07:16 | <alkisg> Yup, you're right, it's easier to read
| |
| 07:16 | if [ -n "$modes" ]
| |
| 07:17 | And 4 spaces indentation :P
| |
| 07:17 | <vagrantc> alkisg: with $modes it doesn't matter so much if there's a space
| |
| 07:17 | <alkisg> I mean [] instead of test
| |
| 07:17 | <vagrantc> the only not-so-easy-to-read is the EOF bits
| |
| 07:18 | <alkisg> We don't use "if test" so much in the code... either "if []" or plain "test"
| |
| 07:18 | <vagrantc> alkisg: test -n is ok for short one-liners
| |
| 07:18 | <alkisg> Sure, but we don't use "if test", just plain "test" there
| |
| 07:18 | <vagrantc> ah, didn't see that.
| |
| 07:18 | nice catch
| |
| 07:21 | * alkisg will stop proposing silly tiny improvements when vagrantc tells him to... | |
| 07:21 | <alkisg> Smaller sed: sed -n 's/^X_MODE_[0-9]=//p'
| |
| 07:21 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 07:22 | <vagrantc> stole that from the HOSTS_* code
| |
| 07:23 | if i'm not careful, i'll upload at precisely midnight (in my timezone)
| |
| 07:24 | <alkisg> Finally, /me prefers "unset var" instead of "var=", because at the later case "var" shows up in `set`
| |
| 07:25 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 252 seconds) | |
| 07:30 | * vagrantc conceedss | |
| 07:30 | <vagrantc> alkisg: did you re-enable translations?
| |
| 07:31 | <alkisg> vagrantc: no, I thought you did!
| |
| 07:31 | <vagrantc> nope
| |
| 07:31 | <alkisg> Ah
| |
| 07:31 | wait, that's ltsp-trunk, we never disabled them there
| |
| 07:31 | <vagrantc> heh
| |
| 07:31 | <alkisg> Only ldm has issues because of intltool
| |
| 07:31 | (so for ltsp6 we'll be fine :P)
| |
| 07:32 | <vagrantc> alkisg: it was a pretty pointless translation commit
| |
| 07:32 | just updated line numbers in files...
| |
| 07:33 | or, even worse, used the same line numbers, but rearranged the order of the line numbers!
| |
| 07:33 | er, not line numbers, but files
| |
| 07:33 | in comments.
| |
| 07:34 | there was some actual content in at least one of the files, at least.
| |
| 07:37 | <alkisg> Hrm. And it didn't update ltsp.pot, so it might just do that again...
| |
| 07:38 | <vagrantc> alkisg: grep -q "root=/dev/nbd." /proc/cmdline || exit 0
| |
| 07:38 | root=/dev/nbd. ???
| |
| 07:38 | <alkisg> . == any number
| |
| 07:38 | Or better yet, [0-9]...
| |
| 07:38 | <vagrantc> well, anything
| |
| 07:38 | * vagrantc nods | |
| 07:38 | <alkisg> It's a bit confusing though, so yeah [0-9] would be better
| |
| 07:40 | * alkisg will commit that in a couple of minutes | |
| 07:41 | <alkisg> Meh I have /dev/nbd0 hardcoded for root, I should have used 0 there instead
| |
| 07:41 | <vagrantc> we shouldn't have 0 hard-coded for root either...
| |
| 07:42 | <alkisg> Sure, so OK i'll leave that for when it's properly rewritten
| |
| 07:42 | Although that script shouldn't even exist, but anyways...
| |
| 07:42 | * vagrantc nods | |
| 07:43 | <alkisg> !learn alkisg-todo as ltsp-config aoe (for swap too)
| |
| 07:43 | <ltsp> The operation succeeded.
| |
| 07:43 | <vagrantc> swap is going to be harder, isn't it? it's not like multiple clients can share that...
| |
| 07:44 | <alkisg> Same as in nbd
| |
| 07:44 | The real trouble is managing the numbers
| |
| 07:44 | * vagrantc nods | |
| 07:45 | <vagrantc> alkisg: so, are you going to commit that nbd-disconnect fix, or should i, or... ?
| |
| 07:45 | <alkisg> The good thing is that aoe knows the client's mac address
| |
| 07:45 | vagrantc: nah just leave it as it is now
| |
| 07:45 | <vagrantc> ok.
| |
| 07:45 | <alkisg> Unless you want me to fix it for any root, not just 0
| |
| 07:46 | <vagrantc> no, that's too much work for short order, although i think it's worth catching the nbd-disconnect now
| |
| 07:46 | would be easy to forget later
| |
| 07:46 | <alkisg> It does work now
| |
| 07:46 | It just matches any character instead of any digit, nothing significant
| |
| 07:47 | <vagrantc> sure
| |
| 07:47 | <alkisg> Fortunately I've discovered the kvm's "-no-shutdown" parameter, that allowed me to see the output of those last commands
| |
| 07:47 | <vagrantc> nice
| |
| 07:47 | <alkisg> ...and, "console output" in the ltsp-client-core upstart job
| |
| 07:48 | (otherwise echo >&2 shows nothing)
| |
| 07:48 | * vagrantc commits it anyways :P | |
| 07:48 | <alkisg> :)
| |
| 07:48 | vagrantc: ok at least put /dev/nbd0 there, not [0-9]
| |
| 07:48 | I.e. "we only care about roots that we know about"
| |
| 07:49 | Wait wait
| |
| 07:49 | vagrantc: /dev/nbd[1-9]) => put that 0-9 too
| |
| 07:49 | vagrantc: leave it to me
| |
| 07:49 | We can easily do the "any root" change
| |
| 07:50 | <vagrantc> forget it, time to release :)
| |
| 07:50 | <alkisg> OK
| |
| 07:50 | <vagrantc> eesh, haven't even started on the changelog.
| |
| 07:50 | <alkisg> Ah you want to do it before 2014 :D
| |
| 07:51 | <vagrantc> yup
| |
| 07:53 | <alkisg> vagrantc: http://pastie.org/8591167
| |
| 07:54 | Or http://pastie.org/8591168 in .diff form
| |
| 07:57 | Err for param in $(cat /proc/cmdline) sorry
| |
| 07:57 | <vagrantc> too late, already tagged and pushed 5.5.0
| |
| 07:58 | alkisg: is nbd swap handled by something else?
| |
| 07:58 | <alkisg> Nice!!!
| |
| 07:59 | <vagrantc> upload will likely be late
| |
| 07:59 | <alkisg> (updated) http://pastie.org/8591174
| |
| 07:59 | I think there's a service that just runs swapoff -a on shutdown
| |
| 08:03 | initscripts: /etc/init.d/umountfs
| |
| 08:04 | Happy new year vagrant :)
| |
| 08:05 | * vagrantc thanks alkisg and continues to frantically write a changelog | |
| 08:15 | * vagrantc wonders if the NBD swap thresholds are a bit too high | |
| 08:20 | <alkisg> vagrantc: I've had crashes with 1 gb fat clients
| |
| 08:20 | <vagrantc> 512MB ram + 512MB swap for a thin client?
| |
| 08:21 | <alkisg> Fat clients with 1 GB RAM didn't get NBD swap previously due to low thresholds,
| |
| 08:21 | and when I changed the thresholds and they got NBD_SWAP, then the crashes stopped
| |
| 08:22 | Assuming firefox/libreoffice might need 500 mb ram on a thin client,
| |
| 08:22 | and the rest of the OS might need ~200 MB RAM...
| |
| 08:22 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 08:22 | <vagrantc> might be right.
| |
| 08:22 | <alkisg> That's 700 mb without using localapps
| |
| 08:22 | <vagrantc> it's just hard to even think of that as a thin client
| |
| 08:23 | <alkisg> It's silly that xorg doesn't take `free` into account when apps tell it to use local pixmaps
| |
| 08:23 | Firefox wants 300 mb in local pixmaps? and I only have 2 mb free? tell firefox "no I can't do that"
| |
| 08:27 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 272 seconds) | |
| 08:33 | <vagrantc> no doubt
| |
| 08:34 | X.org needs to learn how to say no.
| |
| 08:34 | <alkisg> Actually I think that fixing things like those are even more important than ditching ldm... at some point we need to put some resources into making some upstream projects respect remote desktop solutions more
| |
| 08:34 | <vagrantc> of course, they'll just move on to wayland or mir or who knows what
| |
| 08:34 | * vagrantc nods | |
| 08:35 | <vagrantc> sorry, we don't support systems without 2TB of video ram anymore...
| |
| 08:42 | * alkisg would like it if we branched ltsp6 now, right after tagging ltsp 5.5 | |
| 08:43 | <alkisg> ...actually just being able to say "ltsp 5.5 is the last version that runs with ldm and isn't developed anymore, it's only getting some backports from ltsp6"
| |
| 08:47 | * alkisg publishes the latest versions of ldm, ltspfs and ltsp in the greek schools *proposed* ppa, for *precise* :) | |
| 08:52 | <alkisg> vagrantc: schools here have the versions shipped around wheezy, do you think that if one upgrades the server only (and not the chroot) to the latest ldm/ltspfs/ltsp versions, will have issues?
| |
| 08:52 | vmlintu has joined IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi) | |
| 08:56 | <alkisg> http://packages.qa.debian.org/l/ltsp.html
| |
| 08:56 | <vagrantc> don't think there are any incompatibilities
| |
| 08:57 | <alkisg> I think http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=701916 can be closed now, it's committed/released, even if it never were actually reproduced in debian
| |
| 08:58 | <vagrantc> ok.
| |
| 09:00 | <alkisg> ...so, tomorrow we start ltsp6? :D
| |
| 09:00 | <vagrantc> seems like it
| |
| 09:02 | bennabiy has left IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com, Ping timeout: 240 seconds) | |
| 09:03 | bennabiy has joined IRC (bennabiy!~Thunderbi@24.181.55.79) | |
| 09:23 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 09:27 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 245 seconds) | |
| 09:33 | * alkisg wonders how easy would it be to backport ltsp5.5 for squeeze/10.04 for use with an aoe root, for schools that want gnome 2, have 64mb ram clients etc | |
| 09:35 | * vagrantc cringes | |
| 09:36 | <vagrantc> wheezy with one of the external mate repositories is probably a more sane bet...
| |
| 09:37 | <alkisg> Why oh why didn't those mate folks invest on gnome-flashback instead... :(
| |
| 09:40 | * vagrantc can't keep track of all the variants | |
| 09:42 | <alkisg> I think it's possible to have multiple vlade instances that serve different swap files for the same slot, if one defines the mac address that each vblade instance should reply to
| |
| 09:43 | ...but the best approach there would be an upstream patch to vblade
| |
| 09:45 | <vagrantc> right
| |
| 09:46 | well, uploaded ltsp 5.5.0-1 to debian unstable
| |
| 09:46 | <alkisg> :)
| |
| 09:47 | <vagrantc> my first upload of the year.
| |
| 09:48 | <alkisg> vblade is using stdlib instead of glib, I love it
| |
| 09:49 | * vagrantc doesn't know which alkisg prefers | |
| 09:49 | <alkisg> So... plans for 2014: compressed btrfs with embedded swap partition over aoe with server-side cow support
| |
| 09:50 | <vagrantc> it's about time!
| |
| 09:50 | <alkisg> glib is gnome specific and I don't want to invest there as I think qt should be preferred for UI apps and plain libc for non ui apps
| |
| 09:51 | Or at least that's my initial understanding, when I'm ready to write .c code I'll check again :)
| |
| 09:54 | <vmlintu> sounds interesting.. so nbd is going out?
| |
| 09:55 | <vagrantc> alkisg: are there other AoE implementations besides vblade?
| |
| 09:55 | alkisg: we could also boot with iPXE and just boot directly from the image
| |
| 09:55 | could also do that with NFS
| |
| 09:56 | <alkisg> vmlintu: of course not, but if aoe doesn't suffer from ip issues, persistence issues etc etc, why not default to it at some point in the future?
| |
| 09:57 | vagrantc: no idea, but it seems very easy to fork it
| |
| 09:58 | (11:55:47 πμ) vagrantc: alkisg: we could also boot with iPXE and just boot directly from the image => you mean if the cow is server-side?
| |
| 09:58 | How would we do that with NFS?
| |
| 09:58 | <vagrantc> alkisg: looking at how much extra space nbi.img files take, i'm really thinking we should drop support by default
| |
| 09:58 | <alkisg> +1000
| |
| 09:58 | They just don't work
| |
| 09:59 | Lets start ltsp6 tomorrow and drop a whole lot of things :D
| |
| 09:59 | <vmlintu> alkisg: nbd certainly has had its share of problems.. and it would be nice to have fail-over for the image server..
| |
| 09:59 | <alkisg> nbi, nbd-proxy etc, and anyone that wants them can send working patches
| |
| 09:59 | <vagrantc> i guess just removing the package dependency will at least cause them not to generate
| |
| 09:59 | <alkisg> Ah true... wish you could do that for ltsp 5
| |
| 09:59 | 5.5
| |
| 09:59 | <vagrantc> alkisg: but stgraber says nbd-proxy works! :)
| |
| 10:00 | <alkisg> Oh well
| |
| 10:00 | <vagrantc> alkisg: i think the code supports the wrappers not being present
| |
| 10:00 | alkisg: for nbi.img
| |
| 10:00 | <alkisg> I'd say that anyone who wants a proxy, should send a patch for xnbd-proxy instead
| |
| 10:01 | <vagrantc> or split nbd-proxy into it's own project
| |
| 10:01 | at the very least
| |
| 10:01 | <alkisg> Sure, but it'll be harder, just using xnbd sounds much easier (and more stable, and extra benefits for local caching, etc etc)
| |
| 10:01 | I'd love to have support for xnbd in ltsp-trunk, but I don't like nbd-proxy
| |
| 10:02 | Also, the reason it was invented was persistence, I think there were several upstream nbd-client commits for persistence... maybe it doesn't even have a reason to exist anymore
| |
| 10:03 | I would like to have at least someone that uses it and sends feedback on why it's still needed... anyways
| |
| 10:03 | <vmlintu> Have you thought about using the same image to netboot ltsp servers?
| |
| 10:04 | <alkisg> I think ltsp-pnp made that possible, no?
| |
| 10:04 | Same image not only for ltsp-servers but for ltsp-clients as well :D
| |
| 10:07 | <vmlintu> I haven't really been using ltsp-pnp, but we are now using the same image for thin and fat clients, ltsp servers and laptops
| |
| 10:09 | <alkisg> ltsp-pnp was about doing stuff from init-ltsp.d
| |
| 10:09 | I.e. removing services, configuration files etc on boot, not statically on ltsp-build-client
| |
| 10:09 | <vagrantc> one of the better innovations in recent years
| |
| 10:12 | <vmlintu> yes, the init-ltsp.d approach is the key to that
| |
| 10:12 | I needed to do some big changes to that to init-ltsp.d parts get laptop boots without network. Would that be something that you'd want in ltsp6?
| |
| 10:13 | <alkisg> vmlintu: say more about that... what changes?
| |
| 10:13 | freedomrun has joined IRC (freedomrun!~freedomru@unaffiliated/freedomrun) | |
| 10:13 | <alkisg> !local-boot
| |
| 10:13 | <ltsp> local-boot: If you want LTSP fat clients on a low-speed network, you can put i386.img on e.g. C:\Boot\LTSP\i386.img and use this command line in pxelinux.cfg: APPEND ro initrd=ltsp/i386/initrd.img init=/sbin/init-ltsp root=/dev/sda1 rootflags=ro loop=/Boot/LTSP/i386.img; IPAPPEND 3
| |
| 10:13 | <alkisg> I thought that was all that's needed...
| |
| 10:14 | (oh well, similar, without the pxe kernels part)
| |
| 10:14 | * vagrantc wonders where loop= is implemented | |
| 10:14 | <alkisg> in initramfs-tools
| |
| 10:14 | <vagrantc> not in debian's initramfs-tools
| |
| 10:14 | <alkisg> Really?!! That's ubuntu-ism?!
| |
| 10:14 | <vmlintu> normally the init-ltsp.d scripts require that network is available during boot time, but when you run local grub + image on harddrive, dhclient and others fail
| |
| 10:15 | when booting without network, that is
| |
| 10:15 | <alkisg> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=468114
| |
| 10:16 | vmlintu: sure, could you file a bug report and send patches there?
| |
| 10:17 | It's actually a good thing that debian doesn't support loop=, it'll be easier to convince maks to support root=<multiple,roots,to,be,executed,sequentially> :D
| |
| 10:18 | Or at least mount=x,y,z to be executed sequentially before root=
| |
| 10:18 | Ah. mount=x mount=y mount=z
| |
| 10:19 | <vmlintu> alkisg: the "bad" thing about that is that it kind of requires local storage to store some information like hostname and lts.conf etc..
| |
| 10:19 | <alkisg> vmlintu: I don't think that's bad...
| |
| 10:20 | They could go in kernel parameters, or inside the squashfs image, or even in C:\ltsp\lts.conf :D
| |
| 10:20 | <vmlintu> Right now I'm using a specially named lvm partition, but the kernel parameter sounds better
| |
| 10:21 | <vagrantc> alkisg: have fun with ltsp6! :)
| |
| 10:21 | <alkisg> Ooooh /me checks trunk...
| |
| 10:22 | <vagrantc> alkisg: there are probably a handful of smallish fixes that might make sense to push first
| |
| 10:22 | <alkisg> Ah, I thought you created the ltsp5 branch
| |
| 10:22 | Sure, let's do that
| |
| 10:22 | We can start ltsp6 when we're out of ideas that apply to ltsp5
| |
| 10:22 | <vagrantc> heh
| |
| 10:23 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 10:23 | <vagrantc> i'd kind of like to split up update-kernels and ltsp-client-core initscript
| |
| 10:24 | or at least add some crude plugin support to both
| |
| 10:25 | <alkisg> OK, I'll start testing ltsp5.5 on 12.04 to see if any fixes are needed
| |
| 10:25 | <vagrantc> sounds good!
| |
| 10:26 | <vmlintu> do you feel like using fat clients as wifi access points?
| |
| 10:27 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 10:28 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 260 seconds) | |
| 10:38 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Remote host closed the connection) | |
| 10:38 | mattcen has left IRC (mattcen!~mattcen@c110-22-201-130.sunsh4.vic.optusnet.com.au, Ping timeout: 272 seconds) | |
| 11:19 | gdi2k has joined IRC (gdi2k!~gdi2k@222.127.254.113) | |
| 11:24 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 11:27 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Read error: Operation timed out) | |
| 12:09 | mattcen has joined IRC (mattcen!~mattcen@c110-22-201-130.sunsh4.vic.optusnet.com.au) | |
| 12:25 | alexqwesa has left IRC (alexqwesa!~alex@109.172.12.47, Ping timeout: 248 seconds) | |
| 12:25 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 12:28 | alexqwesa has joined IRC (alexqwesa!~alex@109.172.12.47) | |
| 12:30 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Ping timeout: 265 seconds) | |
| 12:45 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 12:57 | cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg) | |
| 13:05 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Remote host closed the connection) | |
| 13:05 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 13:24 | gdi2k has left IRC (gdi2k!~gdi2k@222.127.254.113, Ping timeout: 260 seconds) | |
| 13:53 | gdi2k has joined IRC (gdi2k!~gdi2k@222.127.254.113) | |
| 14:20 | PhoenixSTF has joined IRC (PhoenixSTF!~rudi@78.29.154.124) | |
| 14:34 | willianmazzardo has joined IRC (willianmazzardo!~textual@187.4.15.116) | |
| 14:56 | khildin has joined IRC (khildin!~khildin@ip-80-236-219-149.dsl.scarlet.be) | |
| 16:14 | Phantomas has joined IRC (Phantomas!~Phantomas@ubuntu/member/phantomas) | |
| 16:16 | shogunx has left IRC (shogunx!~shogunx@rrcs-67-79-182-229.se.biz.rr.com, Ping timeout: 252 seconds) | |
| 16:17 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 16:21 | Fenuks has joined IRC (Fenuks!~Fenuks@176.51.127.69) | |
| 16:38 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Remote host closed the connection) | |
| 17:02 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 17:27 | freedomrun has left IRC (freedomrun!~freedomru@unaffiliated/freedomrun, Ping timeout: 240 seconds) | |
| 17:28 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Remote host closed the connection) | |
| 17:29 | vagrantc has joined IRC (vagrantc!~vagrant@freegeek/vagrantc) | |
| 17:41 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 17:41 | freedomrun has joined IRC (freedomrun!~freedomru@unaffiliated/freedomrun) | |
| 17:42 | <alkisg> vagrantc: I can't install ltsp-client 5.5 on 12.04: http://packages.ubuntu.com/search?keywords=fonts-dejavu-core
| |
| 17:42 | ...recommends? :)
| |
| 17:43 | Also, please remove zram-config, it's very bad for ltsp clients
| |
| 17:43 | alexqwesa has left IRC (alexqwesa!~alex@109.172.12.47, Remote host closed the connection) | |
| 17:43 | <alkisg> (assuming that we do have nbd swap in ltsp... which is way better)
| |
| 17:45 | <vagrantc> alkisg: got patches in one of your packaging branches?
| |
| 17:45 | <alkisg> Meh the package name has changed, it was: ttf-dejavu-core
| |
| 17:45 | So, it can just be one | the other
| |
| 17:46 | <vagrantc> put the newest named one first
| |
| 17:46 | <alkisg> No I don't, do you want me to?
| |
| 17:46 | OK
| |
| 17:46 | <vagrantc> alkisg: it's easier to merge exactly what you're asking for if you do :)
| |
| 17:46 | <alkisg> Should I put zram-config in suggests?
| |
| 17:46 | <vagrantc> dunno.
| |
| 17:47 | <alkisg> It would help installations that for some reason want to avoid nbd swap
| |
| 17:47 | <vagrantc> alkisg: do these packages work fine in 14.04 ?
| |
| 17:47 | <alkisg> The zram-config package?
| |
| 17:47 | <vagrantc> no, the others
| |
| 17:47 | <alkisg> The dejavu ones? They're just fonts, they renamed a package... yeah I do see dejavu fonts...
| |
| 17:48 | alexqwesa has joined IRC (alexqwesa!~alex@109.172.12.47) | |
| 17:51 | <alkisg> vagrantc: can we lower the fonts to recommends?
| |
| 17:51 | Ubuntu now defaults to installing recommends in ltsp chroots, so it should be fine
| |
| 17:52 | And, I don't see why I must absolutely have thai fonts in my chroot...
| |
| 17:52 | <vagrantc> lower all the fonts to recommends ... that makes some sense
| |
| 17:53 | <alkisg> dejavu wasn't the only package that got renamed, so I'll check all packages now
| |
| 18:11 | bennabiy has left IRC (bennabiy!~Thunderbi@24.181.55.79, Ping timeout: 240 seconds) | |
| 18:11 | bennabiy has joined IRC (bennabiy!~Thunderbi@24-181-55-79.dhcp.gnvl.sc.charter.com) | |
| 18:31 | <alkisg> vagrantc: http://bazaar.launchpad.net/~ts.sch.gr/sch-scripts/ltsp-debian-packaging/revision/1189
| |
| 18:37 | <stgraber> alkisg: we can't do a direct sync from unstable but I'll try to do a merge over the next couple of days
| |
| 18:37 | alkisg: I'll be travelling tomorrow so I'll grab the source packages on my laptop and see if I have some time during my flight for that
| |
| 18:38 | <alkisg> stgraber: thanks, in the meantime I've built it with a recipe on launchpad and I'll be testing that on 14.04 and also 12.04
| |
| 18:38 | Btw it would be nice to have it autosync from debian, i.e. to completely eliminate the delta...
| |
| 18:39 | (not from unstable of course...)
| |
| 18:39 | stgraber: please also see the link I just posted above ^ where I lower the fonts to recommends, and zram to suggests
| |
| 18:42 | <stgraber> alkisg: I'll take a look. We're not very far from being syncable with Debian though some bits are going to be hard to handle without the occasional manual merge.
| |
| 18:42 | * alkisg has been using the debian packaging for the last 2 years in ubuntu | |
| 18:42 | <stgraber> assuming vagrantc applied some of the patches I sent him last time I did the merge, the delta should be pretty small nowadays (some d-i changes and dhcpd)
| |
| 18:43 | <alkisg> stgraber: if you want, cc me when you have the delta ready, I think it's worth the trouble to get it completely synced
| |
| 18:43 | I.e. to find ways to ship different dhcpd packages or whatever else is needed
| |
| 18:46 | <vagrantc> depends how ugly it is...
| |
| 18:46 | i mean, we can do all manner of hacks in debian/rules conditionally based on Vendor, but...
| |
| 18:46 | i did merge most of the ubuntu changes, but i forget what all was leftover
| |
| 18:46 | <alkisg> vagrantc: just simple file overlaying
| |
| 18:47 | Same as we currently do for init-ltsp.d
| |
| 18:47 | <vagrantc> we'd probably need to use dh_exec to do that sanely.
| |
| 18:48 | <alkisg> I think stgraber wants to keep the old 192.168.0.1 and .254 server IPs instead of the new 192.168.67.1 one
| |
| 18:50 | <vagrantc> could probably handle that with sed...
| |
| 18:50 | at build-time
| |
| 18:50 | <alkisg> btw having `ltsp-config dnsmasq` always create the 192.168.67.1 range makes it pretty simple to tell the users "if you have a 2 nic setup, just declare that static IP, done"
| |
| 18:50 | sch-scripts even do NAT-ting there automatically for fat clients
| |
| 18:51 | GEEGEEGEE has left IRC (GEEGEEGEE!~Charles@cpc8-sprt2-2-0-cust26.17-2.cable.virginm.net, Ping timeout: 272 seconds) | |
| 18:56 | <vagrantc> alkisg: sorry for misspelling your name!
| |
| 18:57 | <alkisg> No worries I don't like my surname anyways :)
| |
| 19:00 | <stgraber> alkisg: yeah, I need to figure out an upgrade path...
| |
| 19:01 | <alkisg> stgraber: I think ltsp6 will be so different that we should declare it needs manual upgrading from ltsp5
| |
| 19:01 | So then it will be a good time to break the upgrade path :)
| |
| 19:01 | <stgraber> alkisg: sounds reasonable :)
| |
| 19:10 | danishman has joined IRC (danishman!~kvirc@62-243-156-218-static.dk.customer.tdc.net) | |
| 19:15 | <vagrantc> alkisg: needed to +Suggests: ${vendor:Suggests}
| |
| 19:15 | <alkisg> Sorry, didn't see that
| |
| 19:15 | Fenuks has left IRC (Fenuks!~Fenuks@176.51.127.69, Quit: KVIrc 4.3.1 Aria http://www.kvirc.net/) | |
| 19:16 | <vagrantc> no worries
| |
| 19:21 | alkisg: i reworked your patch slightly and committed it.
| |
| 19:25 | vagrantc has left IRC (vagrantc!~vagrant@freegeek/vagrantc, Quit: leaving) | |
| 19:32 | GEEGEEGEE has joined IRC (GEEGEEGEE!~Charles@cpc8-sprt2-2-0-cust26.17-2.cable.virginm.net) | |
| 19:51 | SmallR2003 has joined IRC (SmallR2003!~Robert@c-98-253-173-240.hsd1.il.comcast.net) | |
| 19:51 | rhptyosy has joined IRC (rhptyosy!iamparadox@c-71-206-130-134.hsd1.va.comcast.net) | |
| 19:53 | shawnp0w1rs has joined IRC (shawnp0w1rs!~spowers@151.236.4.166) | |
| 19:58 | |Paradox| has left IRC (|Paradox|!iamparadox@c-71-206-130-134.hsd1.va.comcast.net, *.net *.split) | |
| 19:58 | SmallR2002 has left IRC (SmallR2002!~Robert@c-98-253-173-240.hsd1.il.comcast.net, *.net *.split) | |
| 19:58 | shawnp0wers has left IRC (shawnp0wers!~spowers@linuxjournal/staff/shawnp0wers, *.net *.split) | |
| 20:03 | book` has left IRC (book`!~book`@li125-242.members.linode.com, Ping timeout: 248 seconds) | |
| 20:03 | monteslu has left IRC (monteslu!~monteslu@ip68-109-175-67.ph.ph.cox.net, Ping timeout: 248 seconds) | |
| 20:04 | elias_a has left IRC (elias_a!elias@hilla.kapsi.fi, Ping timeout: 248 seconds) | |
| 20:07 | ltsp` has joined IRC (ltsp`!bot@75.98.171.129) | |
| 20:09 | book` has left IRC (book`!~book`@li125-242.members.linode.com, *.net *.split) | |
| 20:09 | monteslu has left IRC (monteslu!~monteslu@ip68-109-175-67.ph.ph.cox.net, *.net *.split) | |
| 20:09 | ltsp has left IRC (ltsp!bot@ltsp.org, *.net *.split) | |
| 20:09 | elias_a has left IRC (elias_a!elias@hilla.kapsi.fi, *.net *.split) | |
| 20:11 | elias_a_ has left IRC (elias_a_!elias@hilla.kapsi.fi, Ping timeout: 240 seconds) | |
| 20:12 | elias_a has joined IRC (elias_a!elias@hilla.kapsi.fi) | |
| 20:14 | monteslu has joined IRC (monteslu!~monteslu@ip68-109-175-67.ph.ph.cox.net) | |
| 20:29 | Phantomas has left IRC (Phantomas!~Phantomas@ubuntu/member/phantomas, Ping timeout: 240 seconds) | |
| 20:45 | book`_ has left IRC (book`_!~book`@li125-242.members.linode.com, Ping timeout: 252 seconds) | |
| 20:51 | book` has joined IRC (book`!~book`@li125-242.members.linode.com) | |
| 20:53 | danishman has left IRC (danishman!~kvirc@62-243-156-218-static.dk.customer.tdc.net, Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/) | |
| 21:20 | book` has left IRC (book`!~book`@li125-242.members.linode.com, Ping timeout: 240 seconds) | |
| 21:20 | book` has joined IRC (book`!~book`@li125-242.members.linode.com) | |
| 21:36 | khildin has left IRC (khildin!~khildin@ip-80-236-219-149.dsl.scarlet.be, Quit: I'm gone, bye bye) | |
| 21:58 | alexqwesa has left IRC (alexqwesa!~alex@109.172.12.47, Read error: Connection reset by peer) | |
| 21:58 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Remote host closed the connection) | |
| 21:59 | alexqwesa has joined IRC (alexqwesa!~alex@109.172.12.47) | |
| 22:05 | GEEGEEGEE has left IRC (GEEGEEGEE!~Charles@cpc8-sprt2-2-0-cust26.17-2.cable.virginm.net, Ping timeout: 272 seconds) | |
| 22:10 | GEEGEEGEE has joined IRC (GEEGEEGEE!~Charles@cpc8-sprt2-2-0-cust26.17-2.cable.virginm.net) | |
| 22:39 | gbaman has joined IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com) | |
| 22:43 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.) | |
| 22:43 | vmlintu has left IRC (vmlintu!~vmlintu@nblzone-240-143.nblnetworks.fi, Ping timeout: 245 seconds) | |
| 22:55 | PhoenixSTF has left IRC (PhoenixSTF!~rudi@78.29.154.124, Quit: Leaving) | |
| 23:39 | gbaman has left IRC (gbaman!~gbaman@host81-130-35-88.in-addr.btopenworld.com, Remote host closed the connection) | |