[01:25] Hi, I was wondering if there is a URL on how to setup a client with a HDD so it can be used as a server backup ? [09:58] ogra_cmpc, ogra : bug 201536 [09:58] Launchpad bug 201536 in italc "FF exception, new release of iTalc (1.0.7)" [Undecided,New] https://launchpad.net/bugs/201536 [10:16] stgraber, seen the bug update ? [10:16] :) [10:32] ogra_cmpc: yeah !!! [10:32] :) [10:32] ogra_cmpc: Upstream will be happy :) [10:32] and we too i hope :) [10:33] I'm just waiting for a patch fixing screen resizing on amd64 (MMX thing) and will upload the new package somewhere so you can have a look at it [10:33] good [10:42] do you think this: /etc/udev/rules.d/88-ltsp.rules - interferes with local automount on a low fat client? [10:44] Nubae: i just removed the ltsp-client-core init script instead. [10:44] entirely? [10:44] actually, my solution wa snot to install ltsp-client, i just install ltsp-client-core instead AFAIK [10:44] i think [10:44] let me check ;) [10:45] ok, I'm not sure if automounting isn't working because of thin client init scripts or real local automounting not working [10:45] but I managed to get local printers to work yesterday, which I'm happy about [10:47] yeah, ltsp-client-core doesn't install that file... ltspfsd comes with the /lib/udev/add_fstab_entry scripts [10:48] i don't need ltspfsd on my clients so i don't install it [10:56] I can try and take it out as well then, not sure if I need it or not, I'm guessing not [10:57] so when ltsp-client-core starts, normally it install ltspfsd and you took that out? [10:58] it wont work without ldm tunnel in place anyway [10:59] and the udev rule would prevent you from using any kind of removable device [10:59] so remove it [10:59] ltsp-client-core is an init script.. if i ever find a init script on my box which apt-get installs packages, it'll be removed in no time :) [10:59] Nubae: i suggest you review the dependencies for ltsp-client and check if it's really needed. maybe ltsp-client-core will be good enough for you. [11:00] last time i looked ltsp-client-core was a package [11:00] you'll also find that /etc/fstab gets overwritten on boot [11:00] but it contains an initscript :) [11:00] ogra_cmpc: /etc/init.d/ltsp-client-core ;) [11:00] yeah taking out ltspfsd takes out ltsp-client also [11:00] ogra_cmpc: and packages don't "start" ;) [11:00] it never installed any packages though [11:00] so does /etc/hosts [11:01] but indeed its for configuring a parebone thin client system ... [11:01] wrt overwriting of fstab: this will hopefully be configurable once ogra upload a new ltsp package to hardy ;) [11:01] which means there is no /etc/hosts, no /etc/fstab and the like [11:01] fstab is overwritten in /etc/init.d/ltsp-client-setup [11:01] what about /etc/hosts? [11:01] read what i wrote :) [11:01] which i'd like to keep actually because it contains some useful things [11:02] the ltsp-client packages purpose is to configure a barebone thin client system which is usually not configured at all and lacking these files [11:03] yes. i'm not complaining about that. [11:03] if you use ltsp in any other context the first thong your script should do is: update-rc.d ltsp-client-setup remove [11:04] laga, for now patching is fine, but i would like to move the common functions out of the initscript [11:04] so you can source a -functions file and just call thin_client_fstab() which then would create a default fstab [11:05] sounds more versatile so it's good [11:05] so your scripts can just source the functions and you should switch off the actual ltsp related stuff [11:05] that avoids confusion and results in cleaner code [11:06] i'd actually remove ltsp-client-core instead of ltsp-client-setup from init.. ltsp-client-core seems more harmful for my purposes [11:06] yup [11:06] * laga shouldn't be discussing technical stuff when X is taking 100% CPU, making his box lag like hell [11:08] get a better graphics card or stop playing WoW while discussing :P [11:08] grr [11:08] i blame hardy [11:35] ogra_cmpc: looks like I have a fix for the MMX thing [11:35] yay [11:36] he basically was trying to use MMX on amd64 ... [11:36] ouch [11:39] let's upload that to the local package builder and see if that also work on i386 [11:46] no !!!, my pbuilder chroots have the new libc ... [11:53] wait a moment, the fix is already built for x86 ... waiting for the other arches [12:56] is there a page I can look at to know exactly how update-rc.d works, I've tried adding cupsys to it, doing update-rc.d cupsys default, but that doesn't seem to work [12:59] man [12:59] or man update-rc.d [12:59] :) === Paladine is now known as AHanff [13:00] does anyone know where I can get some more info on update-rc.d, looking through google is not bringing up anything [13:04] man update-rc.d [13:04] Nuba1, ^^^ [13:04] :-P [13:04] Nuba1, btw is your wiki name Dvanassche ? [13:05] yeah [13:05] there was someone looking for you last night [13:05] * ogra_cmpc looks through the log [13:05] .. subsume|work [13:05] heh, well I have kde4 installed at home and weirdly, kopete has no irc plugins!! [13:06] yeah, he seems to be working on the low fat client wiki too [13:06] right, i thinkl he wanted to discuss something with you [13:07] k, well, I'll try to stay around a bit more, see if I can run into him [13:15] is there something wrong with calling the ltsp chroot something other than i386 (ie, in my case I call it fati386)? [13:19] no [13:19] if you find that something chokes on it, please file a bug [13:19] all scripts should work with a changed chroot name [13:20] (with teh respective options indeed) [13:30] its been fine for me, there's just a mention on the wiki that people have had problems calling it anything but i386 [13:34] well, we wouldnt have added the --chroot option if we wanted to force that path upon users :) === AHanff is now known as Paladine [15:19] ogra_cmpc: iTalc 1.0.7 test packages : http://www.stgraber.org/download/ubuntu/italc/ [15:21] ogra, ogra_cmpc : I have tested it on amd64, I'll try it on i386 later today [15:21] thanks a lot :) [21:21] ogra_cmpc: ping RE: "profiling". I was sitting at work thinking about a workaround, and I was wondering if you (or other devs) have investigated it. [21:22] crimsun, intrepid is supposed to have a focus on performance improvement and size reduction so that will be an essential part then [21:23] ogra_cmpc: ah ok [21:23] ogra_cmpc: I started implementing something at work that takes the post-install initramfs and compares it to what's actually used from lsmod. [21:24] ogra_cmpc: it's pretty hacky, but I can already strip considerable modules from the initramfs [21:24] heh [21:24] ogra_cmpc: e.g., pci_hotplug and friends cause a 60-second probe/stall here [21:24] thats pretty much what i do on the classmate to speed up booting [21:25] even though i additionally cut down that list manually [21:25] ogra_cmpc: how are you doing it programatically? [21:25] ah [21:26] but alone cating the output from lsmod into /etc/modules and redoing the initramfs brought about 30% speedup [21:26] right. [21:26] I suspect you're going further and actually comparing the usage field? [21:28] * Paladine is off to Skegness Grammar School tomorrow :) [21:29] hmm, no, thats actually something i havent thought of ... the thing with the classmate is that it doesnt need to be dynamic the hw doesnt change [21:29] right. [21:29] at least not during boot [21:29] do you plan to be at UDS this time ? [21:31] ogra_cmpc: nope. I was at the last one. [21:32] i know ... [21:32] there i was very occupied by edu stuff ... this time will be more tech for me ... [21:35] ogra_cmpc: I will see about documenting something using the wiki in the meantime. [21:36] i'll ping you if i put up specs for UDS so that gets discussed and included at the right place :) [21:50] ogra_cmpc: excellent, thanks!