[12:16] <cbx33> really is nn time now
[12:16] <cbx33> nn guys
[12:17] <sbalneav> hey again
[02:59] <bddebian> Hello
[03:42] <sbalneav_> me on another machine
[03:44] <LaserJock> nooooo
[03:44] <LaserJock> we ... can't ... take ... 2
[03:44] <sbalneav> Muaahahahahahaha
[03:44] <sbalneav> Ignore that other fellow, he's an imposter!
[03:45] <sbalneav_> No!! Don't listen to him.  I'M the real sbalneav!
[03:45] <sbalneav_> oh, drat
[03:45] <sbalneav_> I would have made it if it hadn't have been for you pesky kids!
[03:46] <sbalneav_> Well, obviously my Scooby-doo reference fell flat. :P
[03:49] <LaserJock> sorry, we were having a discussion of nasty French hotel food in -motu
[03:49] <LaserJock> it has since devolved
[03:49] <LaserJock> ;-)
[03:49] <sbalneav_> Serves you right for eating there.  :)
[03:50] <sbalneav_> you won't make that mistake again next time, riiiiight? :)
[03:50] <sbalneav_> "Do as we do, say as we say"
[03:50] <LaserJock> "Stick with the LTSP guys"
[03:50] <LaserJock> "The only thing thin, is the clients"
[03:50] <LaserJock> ;-)
[03:50] <sbalneav_> We'll be in U.S. of A this time, so there will be thick steaks a-plenty on dinner forays.
[03:51] <bddebian> w00t
[03:51] <sbalneav_> Lobster, perhaps.
[03:52] <sbalneav_> And beer. Well.  That goes without saying.
[03:52] <sbalneav_> And for desert: Cheesecake.
[03:52] <sbalneav_> Yum
[03:52] <sbalneav_> Well, lets see if my little mod to Ogra's cdpinger works.
[03:54] <cafuego> Wheee, 210 Edubuntu CDs just arrived :-)
[03:55] <sbalneav_> 210?
[03:55] <sbalneav_> yikes!
[03:55] <cafuego> 3x70
[03:55] <cafuego> Spacial order
[03:55] <cafuego> s/a/e/
[03:55] <sbalneav_> SOMEONE's doing a big install, I'd say.
[03:56] <bimberi> cafuego: for SFD?
[03:56] <cafuego> They're going here: http://www.acec2006.info/
[03:56] <cafuego> bimberi: Nope
[03:57] <cafuego> Kattekrab is speaking at that ^^ conference.
[03:57] <bimberi> ah yes, of course :)
[03:57] <cafuego> So she can hand the CDs out to teachers directly.
[04:47] <sbalneav_> 4
[05:57] <sbalneav_> Man, when I get something working late at night, and there's no-one here to share it with, I shed a little tear.
[05:57] <sbalneav_> ;_;
[05:58] <sbalneav_> :)
[05:58] <bimberi_> :)
[05:58] <sbalneav_> cdrom support on thin client is now officially AWESOME
[05:58] <sbalneav_> memory sticks --> DONE
[05:58] <sbalneav_> cdroms --> DONE
[05:59] <sbalneav_> floppies --> COMIN' RIIIIGHT UP
[06:00] <bimberi_> fantastic!
[06:00] <bimberi_> i see lots of questions about local devices
[06:01] <sbalneav_> ogra's cdpinger was great, but it had one or two TINY bugs in it, which I've now squashed.
[06:02] <sbalneav_> Well, I'm an "upstream" developer for LTSP, and I developed the localdevice support there.
[06:02] <sbalneav_> It's going to be even better un Ubuntu.
[06:02] <sbalneav_> So, I guess I'm kicking my own butt :)
[06:02] <bimberi_> lol
[06:02] <LaserJock> hehe
[06:23] <LaserJock> sbalneav_: yo, imbrandon's got a question for you
[06:23] <LaserJock> ;-)
[06:23] <sbalneav_> certainly
[06:23] <imbrandon> heh
[06:23] <imbrandon> [23:21]  <imbrandon> do you know if a edubuntu server ( ltsp ) can server a ppc and amd64 client from a i386 ?
[06:23] <sbalneav_> the answer to that would be "yes"
[06:24] <sbalneav_> What you'd have to do is boot a ppc client on ubuntu, 
[06:24] <imbrandon> cool ok, was just about to try it but all the "thin clients" i would use would be ppc, and all servers i would use woudl be i386 ;)
[06:24] <sbalneav_> do the "ltsp-build-standalone" on it,
[06:25] <sbalneav_> then transfer the /opt/ltsp/ppc dir to /opt/ltsp on the 386 server.
[06:25] <imbrandon> ok so on the ppc load a full ubuntu ?
[06:25] <sbalneav_> just one.  To allow you to build the ppc tree.
[06:25] <imbrandon> ahh ok /me listens
[06:25] <sbalneav_> After that, and when you transfer over the ppc dir to the server, they can all net boot.
[06:26] <imbrandon> cool sounds simple enough
[06:26] <sbalneav_> See, you can't actually BUILD the ltsp chroot on a 386 server because it can't run the ppc bins.
[06:26] <imbrandon> right
[06:26] <sbalneav_> But, if you build it on a ppc, then transfer it over, from there on in, you're just serving it out with nfs
[06:26] <sbalneav_> so, it doesn't matter.
[06:26] <imbrandon> but if i put the ltsp server on a amd64 i could build both x64 and i386 
[06:27] <sbalneav_> Yes.  you should be able to do that.
[06:27] <imbrandon> cuz really its just chroots 
[06:27] <imbrandon> cool ok
[06:27] <sbalneav_> yep.
[06:27] <imbrandon> ok one last "dumb question" 
[06:27] <sbalneav_> sorry, did you /msg me?  I'm on my alter ego here, and not authenticaed on this id, so I dont get /msg
[06:28] <imbrandon> i have tes to get ltsp to work yes so i'm trying the edubuntu way since it all works out of the box pretty much
[06:28] <imbrandon> nope i dident
[06:28] <imbrandon> someoen might have but not me
[06:28] <sbalneav_> ah, ok, just wondering why i didn't see your question.
[06:28] <sbalneav_> anway, go on
[06:29] <imbrandon> ahh i havent asked yet , and the first one i was talking to Laser_away in -motu ;)
[06:29] <imbrandon> ok anyhow ummm *thinks*
[06:29] <imbrandon> ohh the netboot thing
[06:29] <imbrandon> ok i load a enu server up and hop on an i386 with netbot nic
[06:30] <imbrandon> it should find it ok without conflicting with my router ?
[06:30] <imbrandon> or do i need to do the dns trickery
[06:30] <sbalneav_> Is your router handing out dhcp addresses?
[06:30] <imbrandon> like with plain ltsp ?
[06:30] <imbrandon> yea its a netgear cable modem router wireless all in one job that i kinda have to keep dhcp on becouse 
[06:31] <imbrandon> i have some non pc devices that need dhcp from it
[06:31] <imbrandon> i think that was my main hickup in the past to useing ltsp
[06:31] <imbrandon> i was hoping edubuntu took care of that somehow
[06:32] <sbalneav_> If you've got something else handing out dhcp addresses, then dhcp trickery will be a must.  However, dhcp trickery can be avoided by either a) using edubuntu to hand out the dhcp addresses, or b) installing a second net card, and hooking the thin clients up to that.
[06:32] <sbalneav_> Nope, that's pretty much impossible to just "magically" avoid :)
[06:32] <imbrandon> hrm well hrm .....
[06:32] <imbrandon> heh yea i figured that
[06:33] <imbrandon> heh ok .... *thinks* sorry still a bit rusty at this
[06:33] <sbalneav_> dhcp's easy.  There's both here and #ltsp to help you with it.  Not like we haven't helped, oh, say, a few THOUSAND people with it before :)
[06:33] <imbrandon> give me a some code and package i'm in heaven , give me a dhcp server and i need a 12pack of beer to even look at the conf LOL
[06:33] <imbrandon> hehe true
[06:33] <sbalneav_> meh, I code dhcpd.conf stanzas in my sleep.
[06:34] <imbrandon> i bet ;)
[06:34] <imbrandon> ok i'm gonna go try to do this with letting edubuntu hand out the ip addresses, mind if i pick you brain for one more minute
[06:35] <imbrandon> with a theoetical setup to make sure i dont have any major flaws in my thinkgin
[06:35] <imbrandon> so first i turn dhcp off on the router and let it jsut be a router .......
[06:35] <imbrandon> i can set all my pc static for now no probs
[06:36] <imbrandon> the non pc devices can get an ip from edubuntu still correct
[06:36] <imbrandon> and use the router as the gateway
[06:36] <sbalneav_> yep
[06:36] <sbalneav_> no problem
[06:36] <imbrandon> k cool, i think that should work then
[06:36] <sbalneav_> Eminently workable :)
[06:36] <imbrandon> ;)
[06:37] <imbrandon> do you konw if the apple nic's in ibooks ( orinoco drivers ) will netboot, or will i have to have some kinda skel system on them
[06:37] <imbrandon> to boot
[06:37] <imbrandon> or is google my friend on that one ;)
[06:38] <sbalneav_> Um, that I don't know.
[06:38] <imbrandon> np i can grab that via google, but say i do have to have a skel system setup for them
[06:38] <imbrandon> can that be generated from the ubuntu install ?
[06:38] <sbalneav_> I *just* bought a mac ppc a week or two ago to start testing things, and to be honest, I haven't even had a chance to boot it yet.
[06:38] <imbrandon> heh np
[06:39] <sbalneav_> imbrandon: yep.  In fact, if you simply install edubuntu-ppc, it'll do it for you.
[06:39] <imbrandon> i'm guessing i will becouse i would like to use the wireless in them
[06:39] <imbrandon> nice
[06:40] <imbrandon> so there is a like a "client" option in the install ? forgive my ignorance i'm very new to ltsp but not linux in general ( i'm a MOTU among other things )
[06:40] <imbrandon> heh
[06:40] <sbalneav_> Ah, you want wireless netboot?  heh
[06:41] <sbalneav_> there's a LOT of people who want that.  
[06:41] <imbrandon> if possible heh
[06:41] <sbalneav_> Not possible :)
[06:41] <imbrandon> maybe i can help try and figure out a way ;)
[06:41] <imbrandon> heh
[06:41] <sbalneav_> No wireless netboot card currently netboots :(
[06:41] <imbrandon> who know but no ammount of questions are gonna prepare me so i guess i'm gonna dive in and chekc back if i have probs ;)
[06:42] <sbalneav_> yep.
[06:42] <imbrandon> thanks for the quick start though ;)
[06:42] <sbalneav_> Not at all.
[06:43] <imbrandon> yea a wireless card that netbooted would sell like hotcakes i bet ;)
[06:43] <bimberi_> imbrandon: i think you can hold down the N key at boot to make (some) Macs netboot (non wireless)
[06:43] <imbrandon> bimberi_: cool thanks i'll try that
[06:44] <imbrandon> yea it has both, but i would RATRHER use the wireless but that might be a no-go atleaste for now ;)
[06:45] <imbrandon> the 3 i have here are semi older g3 800mhz 640mb ram iBooks so they may or may not but i'll try it ;)
[06:45] <imbrandon> and the "server" is gonna be a amd64x2 4400+ ( running in 32bit mode )
[06:46] <imbrandon> so it "should" work out ok ;)
[06:46] <imbrandon> hrm how is that gonna work though, i thought the apps ran on the server, i guess they do and are piped to the local X running
[06:46] <imbrandon> so X and the kernel + drivers are all that needs to be ppc ?
[06:47] <sbalneav_> Well, X, kernel, drivers, plus a whole buncg of glue
[06:47] <imbrandon> hehe yea i'm generizing a bit but 
[06:48] <imbrandon> yea the WM like gnome/kde + apps run from the server though
[06:48] <imbrandon> and are piped
[06:48] <sbalneav_> yep, that's how X works.
[06:49] <sbalneav_> The X calls are networkable, so an x89 server can display on a ppc client no problem
[06:49] <imbrandon> cool ok, yea like i said still kinda new to ltsp , tried it on other distro's before i started deving for kubuntu but never got it to work, maybe today is my lucky day
[07:01] <jsgotangco> we dont doubt the demigod of LTSP
[07:02] <imbrandon> ;)
[07:16] <sbalneav> jsgotangco: :)
[07:19] <highvoltage> mhuhahahaha
[07:21] <jsgotangco> there goes one of the acolytes
[07:28] <highvoltage> acolytes?
[07:28] <highvoltage> what's that?
[07:29] <sbalneav> highvoltage: One who attends; an assistant. "With such chiefs, and with
[07:29] <sbalneav>       James and John as acolytes." --Motley.
[07:30] <sbalneav> It has a religous connotation. 
[07:32] <highvoltage> aah
[07:38] <sbalneav> Long fruitful night, but now I need sleep.
[07:38] <sbalneav> night all
[08:29] <RichEd> pygi :)
[08:44] <pygi> RichEd, what I did this time? :)
[08:45] <RichEd> ? nothing bad springs to mind ;)
[08:46] <jsgotangco> RichEd: just for info, I will be doing a presentation to GILAS (www.gilas.org) regarding Edubuntu on the 30th. Gina hooked me up for this
[08:46] <pygi> RichEd, o good, for the first time in history ^_^
[08:46] <pygi> RichEd, o, btw. you gonna visit Ubuntu conf. in Hungary next month?
[08:46] <pygi> one of my presentations is about Edubuntu, ofcourse
[08:52] <RichEd> morning guys ...
[08:53] <jsgotangco> :)
[08:53] <jsgotangco> RichEd: really they're *so* excited about htis
[08:53] <RichEd> pygi: would love to ... but that will need to be next year, same time, same place, busy with foundation building blocks in this phase
[08:54] <pygi> RichEd, ah, oki :'(
[08:54] <highvoltage> morning RichEd 
[08:54] <RichEd> jsgotangco: as per all my conversation threads, I am excited & supportive ... you are my community resource main main #1
[08:54] <RichEd> main man #1 :P
[08:55] <RichEd> hi highvoltage 
[08:55] <jsgotangco> if we make a dent on GILAS this would be big
[08:55] <jsgotangco> afaik, they are in bind at the moment with MS Phil
[08:55] <jsgotangco> there seems to be a stalemate
[08:55] <RichEd> any news on the "personal full time association" opportunity ? 
[08:56] <jsgotangco> she called me up today she said she'll get back to me this week
[08:56] <RichEd> jsgotangco: I will make time to help you with the presentation, or find people or resources to assist
[08:56] <RichEd> we need to show support for you ... and I *WILL* make that happen
[08:57] <jsgotangco> i have to fix this up by month's end or else i'll get screwed by my current work
[08:57] <RichEd> MS stalemate ? how ? what ? more ?
[08:57] <jsgotangco> well they seem to have some issues on distributing the software then supporting it
[08:58] <RichEd> jsgotangco: tomorrow, 9:00-10:00 am my time belongs to you on this topic
[08:58] <jsgotangco> i dunno the exact details, GILAS isn't actually a well advertised initiative
[08:58] <RichEd> it is about to go 9:00 am for me now, to give you a time reference
[08:58] <jsgotangco> ahh okay
[08:58] <jsgotangco> 3pm here
[08:58] <RichEd> so chalk it up in your diary ...
[08:58] <jsgotangco> okay
[08:59] <jsgotangco> we can do skype if you want
[08:59] <RichEd> we'll get a private window open, and research chat side by side ...
[08:59] <jsgotangco> ok
[08:59] <RichEd> skype would be good ... i need to get set up ... can you send me a quick start pointer ?
[09:00] <RichEd> i'm on ADSL so b/w is fine ... just need to know where to register / download etc.
[09:00] <bimberi_> !skype
[09:00] <ubotu> To install Skype on Ubuntu, see https://wiki.ubuntu.com/Skype To record on Skype, check: https://wiki.ubuntu.com/SkypeRecordingHowto
[09:00] <RichEd> pygi ... how far is hungary from georgia
[09:00] <bimberi_> 'morning RichEd :)
[09:00] <jsgotangco> http://www.skype.com/download/
[09:01] <RichEd> that uboto is a most helpful bot ! it seems to do word reconition & rapid response !
[09:01] <RichEd> recognition <- reconition
[09:01] <jsgotangco> you can grab the Debian package
[09:01] <RichEd> thanks ubuto ;)
[09:01] <RichEd> hi bimberi_ : do we get beaten by you or the men in black this week (sobbe)
[09:02] <bimberi_> RichEd: it's our turn up against the all blacks this week
[09:03] <bimberi_> the wallabies trained here in Canberra this week, probably because it's colder, to acclimatise
[09:03] <bimberi_> hehe
[09:07] <jsgotangco> cricket?
[09:07] <RichEd> no rugby ... or in recent weeks, more like buggery
[09:08] <RichEd> my team is playing 2 levels below badly
[09:09] <bimberi_> RichEd: ubotu is a bot - i triggered it with !skype
[09:09] <bimberi_> !botsnack
[09:09] <ubotu> Yum!
[09:10] <RichEd> ah ... at first i thought it was a person ... then for a moment I though it actually did grammar recognition :)
[09:10] <jsgotangco> !die
[09:10] <ubotu> Sorry, I don't know anything about die - try searching on http://bots.ubuntulinux.nl/factoids.cgi
[09:10] <RichEd> !springbokrugby
[09:10] <ubotu> Sorry, I don't know anything about springbokrugby - try searching on http://bots.ubuntulinux.nl/factoids.cgi
[09:11] <RichEd> :)
[09:11] <RichEd> !Sorry
[09:11] <ubotu> Sorry, I don't know anything about Sorry - try searching on http://bots.ubuntulinux.nl/factoids.cgi
[09:12] <jsgotangco> hmmm epiphany took over my desktop lol
[09:12] <bimberi_> RichEd: you can actually /msg it - btw it ignores after a certain number of bad queries :)
[09:13] <jsgotangco> RichEd: this looks like partly useful
[09:13] <jsgotangco> http://www.tsf.org.za/index.php?option=content&task=view&id=246&Itemid=33
[09:13] <RichEd> you mean so you can ask in a private window without appearing too stoopid ?
[09:17] <RichEd> jsgotangco: i'm going to shower and will be back to give you my concentration for 20 mins
[09:18] <RichEd> i'm also seeing highvoltage & willvdl for lunch today ... so will ask them for some face-to-face help with references ... there must be something in the TSF / tuXlabs arsenal on the topic
[09:18] <jsgotangco> thanks
[09:19] <bimberi_> RichEd: lol, yes that's right
[09:27] <highvoltage> on which topic? WSD?
[09:29] <RichEd> highvoltage: jsgotangco needs case study reference on how the introduction of IT has improved academic results
[09:29] <RichEd> any reference material that will support a presentation to a foundation in Philippines on why computes should be funded into education
[09:31] <RichEd> jsgotangco: i've got a conversation happening with the Canonical / Intel partner manager ... and his recent mail says that PRC is on their short term radar
[09:31] <jsgotangco> PRC
[09:31] <jsgotangco> china?
[09:31] <RichEd> PRC is Pacific Rim Countries, no ? and that's you no ?
[09:32] <highvoltage> ah
[09:32] <jsgotangco> ohh
[09:32] <jsgotangco> well its all known as People's Republic of China ;)
[09:32] <jsgotangco> there's East Asia and ASEAN
[09:32] <RichEd> oh ... I better check to see what he means by his abbreviation ...
[09:33] <RichEd> But we are talking about Vietnam now, and that area, so you are within that sort of geographicasl region
[09:35] <jsgotangco> that's part of SEA (South East Asia)
[09:36] <RichEd> yep ... this is his territory focus: Vietnam, Latin America, PRC, and Southeast Asia.
[09:39] <RichEd> jsgotangco: google search on [vietnam phillipines]  tells me: "Throughout the past week more than one million people were evacuated due to Typhoon Chanchu in China, the Philippines and Vietnam"
[09:39] <RichEd> so you are in the same bowl so to speak ?
[09:40] <RichEd> anyway ... what I can say now is that there is some growing co-operation between around Intel Education programme & Ubuntu
[09:40] <RichEd> so we may be able to add value above the o/s into your pitch to Gina ... 
[09:40] <jsgotangco> RichEd: the typhoon path usually starts in Philippines then goes to Taiwan or China
[09:41] <jsgotangco> RichEd: i just saw this http://business.newsforge.com/article.pl?sid=06/07/10/2115242&tid=37
[09:42] <jsgotangco> http://anycec.org/main/node/668?PHPSESSID=087555993070778a7c912ae21859116a
[09:43] <jsgotangco> this is exactly what were looking for
[09:46] <RichEd> jsgotangco: quick, what it what link on Gina ? i have a very short email time window ... got the intel guys attention now
[09:47] <jsgotangco> ah wait
[09:47] <cbx33> Mornin all
[09:47] <RichEd> and confirm ... she is a director of http://www.abs-cbnfoundation.com/ ?
[09:47] <cbx33> what a great meeting yesterday :D
[09:48] <jsgotangco> RichEd: 
[09:48] <jsgotangco> RichEd: yes
[09:48] <jsgotangco> http://www.planetphilippines.com/archives/2005/jul1-15-05/current/features_current/feature3.html
[09:48] <RichEd> hello cbx33 : that meeting with highvoltage & willvdl happens today. case study & advocacy will be on the topic list
[09:48] <cbx33> exshellent
[09:48] <cbx33> what time?
[09:48] <RichEd> lunch time
[09:48] <cbx33> cool
[09:49] <cbx33> after that will we know a little more about how to proceed with the packs?
[09:49] <jsgotangco> RichEd: http://www.newsflash.org/2004/02/sb/sb003803.htm
[09:50] <jsgotangco> RichEd: she's from a media tycoon clan
[09:57] <RichEd> cbx33: we'll start a discussion which will lead to a proj plan ... and I now commit to agree to a timeline & deadline by end of tomorrow.
[09:57] <RichEd> so that will be a stake in the ground, where we can haggle over details, but non-negotiable that is it going forward
[09:58] <RichEd> towards a conclusion I mean (not just wanderin' along somewhere)
[10:00] <cbx33> good
[10:16] <RichEd> jsgotangco: see email ... 1st baby step ... but support for you :)
[10:22] <jsgotangco> wooooo
[10:22] <jsgotangco> thanks!!!!
[10:25] <jsgotangco> "kitty"
[10:25] <jsgotangco> hahaha
[10:26] <jsgotangco> man if i get to move in this organization i have to fix up that website
[10:35] <RichEd> see the next email ... lots of resonance ... if you play this info right ... you'll be the 3way man on the ground for CBS & Ununtu & Canonical ralationship manager
[10:35] <RichEd> talk to me before you spread anything. we'll tie it up peoperly to server all of our aims
[10:35] <RichEd> properly to serve
[10:35] <jsgotangco> sever lol
[10:36] <jsgotangco> thanks im sending email now regarding some news i found
[11:37] <jsgotangco> i gotta go
[11:37] <jsgotangco> ciao
[12:24] <valkyrie> hi
[01:04] <rodarvus> good morning
[01:11] <bimberi> hi rodarvus
[01:11] <rodarvus> hi bimberi 
[01:12] <bimberi> wow the local device support (in ogra's email of a few minutes ago) sounds fantastic
[01:14] <ogra> its really sexy :)
[01:14] <bimberi> sbalneav was very happy earlier :)
[01:14] <ogra> well, he broke a lot tonight :P
[01:15] <bimberi> lol
[01:15] <ogra> but i'm already fixing ... he also improved a lot ;)
[01:17] <ogra> ouch ... that mail went to the wrong list ...
[01:19] <bimberi> ogra: -users? that's ok isn't it? or are you worried about expectations
[01:19] <rodarvus> ogra, putting a default lts.conf in edubuntu edgy ?
[01:20] <ogra> rodarvus, yes, we discussed it yesterday in the meeting
[01:20] <rodarvus> ogra, yes, I remember :)
[01:20] <ogra> its one step in the K12 direction
[01:21] <rodarvus> I was asking if this is the email you were referring to (which you said you sent to the wrong mailing list)
[01:21] <rodarvus> but anyhow
[01:21] <ogra> together with dchpd.conf autogeneration and localdev that makes it possible for us to compete feature wise
[01:21] <rodarvus> the defaults seem just fine
[01:21] <ogra> great :)
[01:21] <rodarvus> SWAP is the one likely getting changed the most
[01:22] <ogra> well, we need a default size you can set anywhere 
[01:22] <rodarvus> we could think about tweaking it automatically (without user interaction), but this is surely stuff for edgy+1
[01:22] <ogra> err
[01:22] <rodarvus> ogra, yeah. 64mb is surely a good choice
[01:22] <ogra> but thast the target of the fully automatic swapserver spec
[01:22] <ogra> no user interaction at all
[01:22] <rodarvus> ogra, yes
[01:23] <rodarvus> but what I'm mentioning is changing the default setup automatically
[01:23] <ogra> hmm
[01:23] <rodarvus> (if this is part of the spec  I missed it and need to stufy it with more care :) )
[01:24] <rodarvus> ogra, considering disk space on the server, memory setup for clients, type of applications you are going to run, etc
[01:24] <ogra> right
[01:24] <ogra> but it should be an adjustable value ...
[01:24] <rodarvus> indeed
[01:25] <ogra> either thrugh a config on the server (/etc/nbd-something.conf) or (better) triggered from the client, so you can set individual swap sizes for different clients
[01:26] <ogra> its one option more to use in the clients swapon code ... nbd-server needs just to pick it up for the automatic swapfile creation
[01:26] <rodarvus> nice, that was my next question :)
[01:26] <rodarvus> (if you could setup swap sizes different for each client)
[01:26] <ogra> you should be able to ...
[01:27] <ogra> indeed that doesnt need to be in the initial implementation ... a fixed size is fine in the beginning
[01:27] <ogra> but we offer per client settings in the lts.conf ... so we should keep it in mind :)
[01:27] <rodarvus> *nods*
[01:34] <bimberi> ah, _that_ email (only just got it :/ )
[01:34] <jsgotangco> wow my daughter counted to 200 without failure
[01:34] <jsgotangco> for a 4 1/2 year old that seems impressive
[01:35] <ogra> totally
[01:35] <bimberi> wow, it is!
[01:35] <jsgotangco> i started her teaching to use an abacus last week
[01:38] <jsgotangco> heh im in awe as well
[02:17] <lucasvo> ogra: you are the maintainer of the -user list, right? could you please find out why I ge tmy mails twice? it's lucasvo@vincisolutions.ch
[03:10] <RichEd> very quiet here again today ... hmmmmm
[03:14] <jsgotangco> hey
[03:14] <jsgotangco> ive been looking at the websites
[03:16] <RichEd> hey ... what do you think ?
[03:18] <RichEd> if you were pitching to a lightweight organisation on your own .. then I would say that Intel would say ... ho-hum ... another plea for help
[03:20] <RichEd> but seeing as it is (Ed)Ubuntu + a media related organisation with exposure possibilities & a good philanthropic track record ... in a country where they have alrady donme some good work ...
[03:21] <RichEd> I am sure we can put something together that serves the aims of all 3 parties ?
[03:21] <RichEd> --- Edubuntu boot CD question --- from Marilize ------------------------
[03:22] <RichEd> Does the current LTS 6.06 allow a live CD boot option ? Or only an install ?
[03:22] <jsgotangco> RichEd: interestingly the people i will meet on the 30th are the heads of the two biggest foundations
[03:22] <jsgotangco> RichEd: the current LTS is only install (even the one on shipit)
[03:22] <jsgotangco> RichEd: we have a liveCD option but only for workstation
[03:22] <jsgotangco> its because we ship a server solution primarily
[03:22] <jsgotangco> rather than a desktop one
[03:22] <RichEd> The web site says "includes a Live CD" which I think is a remnant from the 2 CD pack.
[03:23] <jsgotangco> is that from our website?
[03:23] <RichEd> And people are moaning .....
[03:23] <RichEd> Ahhhh .... Ohhhhhhhh ..... Whine .....
[03:23] <RichEd> So I guess we must correct the site then.
[03:24] <jsgotangco> if its on our website i can change it
[03:24] <jsgotangco> (edubuntu.org)
[03:24] <jsgotangco> i was planning on doing a review a few days ago after i made some minor typo corrections but never got to
[03:25] <RichEd> will get the link from marilize now ... tx
[03:25] <jsgotangco> yes please im in the cms now
[03:27] <RichEd> asking ... waiting ...
[03:27] <jsgotangco> hhehe
[03:28] <RichEd> send me these names : <jsgotangco> RichEd: interestingly the people i will meet on the 30th are the heads of the two biggest foundations
[03:28] <jsgotangco> gina lopez of abs-cbn and vicky garchitorena of ayala foundation (for gilas)
[03:29] <jsgotangco> gina hooked us up with vicky
[03:29] <jsgotangco> because of gilas
[03:29] <RichEd> i've just got the names & email addressess of the business development managers for this region at intel ...
[03:30] <jsgotangco> well i think intel is part of gilas
[03:30] <RichEd> i'll drop your meeting names in the intro mail ... who knows ... maybe on the off chance they will want to join you ?
[03:30] <jsgotangco> so our meeting with gilas is preliminary and introduce edubuntu/ubuntu
[03:31] <RichEd> please can you mail me with as much deatil as you can ?
[03:32] <RichEd> what's gilas, what's ayala, why are they joining you, and how do you mean intel is part of gilas ?
[03:32] <RichEd> REQUEST:
 http://www.edubuntu.com/Download
 it should not say Live CD.....for shipping
 thanks
[03:32] <RichEd> END REQUEST
[03:32] <RichEd> thanks ...
[03:32] <RichEd> out for 30 mins ...
[03:33] <jsgotangco> okay thanks
[03:34] <highvoltage> jsgotangco: kewl :)
[03:34] <jsgotangco> highvoltage: slacker
[03:35] <highvoltage> jsgotangco: :(
[03:35] <highvoltage> i prefer it if you call me a slacker
[03:35] <highvoltage> *joke*
[03:35] <highvoltage> :)
[03:35] <jsgotangco> RichEd: GILAS is a privately funded project to connect to the internet the 5,000+ public schools in the country
[03:36] <jsgotangco> RichEd: its a 5 year project i think its on its 2nd year and only 1.200+ has been connected
[03:37] <jsgotangco> let's also add a point release annoucement even if its a bit late
[03:38] <jsgotangco> we shouldn't even have to link to cdimage since this is more of the developmental download links
[03:47] <highvoltage> bye!
[03:47] <jsgotangco> ok the page is done
[03:50] <jsgotangco> RichEd: please check if its ok now
[04:10] <RichEd> marilize will test ... will let you know ...
[04:10] <RichEd> soon
[04:18] <jsgotangco> yeah
[04:18] <jsgotangco> i also added news of the point relase
[04:21] <jsgotangco> ahh fudge
[04:37] <sbalneav> Morning all!
[04:37] <sbalneav> Hey ogra!
[04:37] <bddebian> Heya
[04:37] <jsgotangco> hi
[04:37] <bddebian> Hi sbalneav
[04:37] <RichEd> gun tag herr grawert
[04:37] <sbalneav> hey bddebian jsgotangco RichEd 
[04:38] <RichEd> hi sbalneav ... & cbx33 
[04:38] <cbx33> hi RichEd
[04:39] <bddebian> Hi jsgotangco, RichEd, cbx33 :-)
[04:39] <cbx33> hi bddebian 
[04:40] <jsgotangco> hey
[04:42] <ogra> hey sbalneav 
[04:42] <sbalneav> hey!
[04:43] <sbalneav> Sorry, I started the cdpinger from the startup, but what I was REALLY thinking was starting it from udev, when a cdrom is created
[04:43] <ogra> you forgot to close the (f) filehandle in cdpinger btw :)
[04:43] <sbalneav> but I couldn't think of the udev rule :(
[04:43] <ogra> has funny effects since you open a new f ever 3 seconds
[04:43] <ogra> *every
[04:44] <ogra> my client dies after some time with a "too many open files" error on the console
[04:45] <ogra> hmm, an udev rule ...
[04:45] <ogra> do we know that the blockdevice is a cdrom ? 
[04:45] <ogra> ah, yes, indeed we do
[04:45] <ogra> thats trivial ... we can at least start it from the add_fstab script
[04:46] <ogra> i think i hand over the device type anyway already, let me look
[04:47] <ogra> LABEL=$(echo ${ID_BUS}${ID_TYPE}-$1|tr " " "_") 
[04:47] <ogra> right
[04:48] <ogra> so if ${ID_TYPE} = cdrom we just start a cdpinger 
[04:49] <sbalneav> that's what I was thinking.
[04:49] <sbalneav> then, if there's NO cdrom in the system, you're not running a pinger.
[04:49] <ogra> its just not easy to stop it ;)
[04:49] <ogra> at least in case of multiple cdroms
[04:49] <ogra> you cant just kill all of them ...
[04:50] <ogra> i dont want to do uglyy grepping stuff through the processlist ...
[04:50] <cbx33> ogra, Iwas looking at the ACL
[04:50] <sbalneav> Hmmm.
[04:50] <cbx33> for the scp
[04:51] <ogra> but even if you write a pid file, you wont know which one is right without looking in the processlist
[04:51] <cbx33> from what I can see there are no restrictions on who can run scp
[04:51] <sbalneav> I think I bypassed the check symlink thing that you were doing every pass.  We could just re-position that, and if the symlink goes away, we just exit!
[04:51] <sbalneav> howzzt?
[04:51] <ogra> cbx33, they are in scp itself, it doesnt start up with uid!=0
[04:51] <cbx33> hmmm
[04:52] <cbx33> I'll have to check that
[04:52] <cbx33> the server side works great
[04:52] <sbalneav> pitti's got a few more requests :)
[04:52] <cbx33> I have an ACL working
[04:52] <sbalneav> so I need to do that today :)
[04:52] <ogra> i lie the idea of dynamically strating it better though :)
[04:52] <ogra> *like
[04:52] <sbalneav> yeah, if they don't have a cdrom, no sense using up the ram.
[04:53] <sbalneav> simple and clean.  I'm really linking this implementation.
[04:53] <ogra> yeah
[04:53] <ogra> it gets better every day ;)
[04:53] <ogra> additionally to being sexy in the beginning ;)
[04:53] <ogra> *already
[04:54] <sbalneav> Well, we have the two sexiest developers in Linux working on it, so how could it fail? 
[04:55] <bddebian> heh
[04:56] <ogra> :)
[04:57] <sbalneav> 3)'s the biggie.  Any idears on that one?
[04:57] <ogra> damned ... the udev script wont work this way ...
[04:58] <sbalneav> verdampt?
[04:58] <ogra> yeah
[04:58] <ogra> the device wond be called cdrom ... so we cant use the symlink
[04:59] <ogra> at least fro usb devices you get scdX
[04:59] <sbalneav> ah, I'm already talking with keybuk about that, we've got the same problems with floppies.
[05:00] <sbalneav> I think there's a udev magical bit where you can pass the symlink name instead of the kernel name.
[05:01] <sbalneav> if nothing else we could do a RUN+="cdpinger cdrom%n"
[05:03] <ogra> have a look at 60-symlink.rules
[05:03] <ogra> will a 65 script override that ? 
[05:03] <sbalneav> yes. should, that's what keybuk recommended.
[05:03] <ogra> oh, ah
[05:03] <ogra> heh
[05:04] <ogra> ENV{ID_CDROM}=="?*",            SYMLINK+="cdrom"
[05:04] <ogra> that means that the symlink will always point to the most recent created device if i'm not wrong
[05:04] <sbalneav> right
[05:05] <ogra> which would mean its completely unreliable anyway
[05:05] <sbalneav> If we wanted separate ones, we'd wand SYMLINK+="cdrom%n"
[05:05] <sbalneav> Which would give us cdrom0 cdrom1, etc
[05:05] <ogra> yep
[05:05] <sbalneav> which would still work with g-v-m, and solve all the problems.
[05:05] <ogra> i wonder why the script doesnt do that yet already
[05:06] <ogra> looks buggy
[05:07] <jsgotangco> the 13th planet?
[05:07] <cbx33> anyone who has bought a tosh laptop will have seen that plastic on the top
[05:07] <cbx33> that peals off
[05:07] <jsgotangco> yeah okay
[05:07] <cbx33> it;s quite sticky and static
[05:07] <jsgotangco> uh huh
[05:07] <cbx33> if you put 15-20 sheets of that together
[05:07] <cbx33> it's like flamin kevlar
[05:07] <cbx33> and extremely light
[05:07] <cbx33> I tried to stab a screwdriver through it
[05:08] <cbx33> seriously full force
[05:08] <cbx33> didn't get through
[05:08] <RichEd> jsgotangco: link looks fine ... thanks
[05:08] <cbx33> that's pretty tough stuff when in multiple sheets.....quite flexible too :p
[05:08] <jsgotangco> okay so you're saying let's buy a lot of toshiba laptops and use them plastics to create a DIY protection suit
[05:08] <jsgotangco> RichEd: cheers
[05:09] <cbx33> jsgotangco, exactly
[05:09] <jsgotangco> nice we'll be able to replace ogra's roof with toshiba laptops as well
[05:09] <cbx33> heheh
[05:11] <sbalneav> ogra: why don't we just handle the direct name, and just have cdpinger create directories like cdrom%n, and just scan the directory for the first available free n?
[05:12] <ogra> sounds like the best option :/
[05:12] <sbalneav> so, we se scd[whatever] . and we look through /tmp/drives.  If there's no cdrom0, then create it, if there is, then create cdrom1, etc etc/
[05:12] <cbx33> ogra, the implementation of the scp dbus I found a problem with
[05:12] <cbx33> got a sec?
[05:12] <sbalneav> shouldn't be that much extra work.
[05:13] <ogra> no but  its an ugly limitation i wouldt have expected 
[05:13] <ogra> cbx33, shoot
[05:13] <cbx33> a) does the server know when a message has completed?
[05:13] <ogra> why should it need to ?
[05:13] <cbx33> b) what happens if two people send a request from scp at the same ich time
[05:14] <sbalneav> yeah, I wasn't aware that %n was so flakey a number.
[05:14] <cbx33> how shorta poll time should we have?
[05:14] <cbx33> and should messages have timeout periods?
[05:14] <ogra> the second instalce should fail to connect, that needs internal handling in scp
[05:14] <cbx33> ahh
[05:14] <ogra> it shouldnt even start up 
[05:14] <cbx33> ok
[05:15] <cbx33> so you're not even contemplating that two versions could be running at once
[05:15] <cbx33> I was thinking of having a messaging bus with incrementing message numbers so the client can see which messages it has read
[05:15] <ogra> exactly
[05:15] <cbx33> and have a TTL on the message
[05:16] <ogra> its limited to a single ltsp server anyway
[05:16] <ogra> i dont think you need message numbers
[05:17] <ogra> the client should just listen (poll every n secs) for a message from the namespace ...
[05:17] <ogra> if a list message comes and it finds $USER in there, it will accept the next message as well (which should be a command message)
[05:18] <cbx33> how long till the mesage is deleted?
[05:18] <ogra> if theer doesnt come a second message after a certain time (5 secs or so) it starts to listen to "list" messages again
[05:18] <cbx33> hmmm
[05:19] <ogra> some seconds whould suffice for all clients to pick it up
[05:19] <cbx33> how does the client know it's just done a message?
[05:19] <ogra> *should
[05:19] <cbx33> and not to do the same message again
[05:19] <ogra> why shoudlnt it ? 
[05:19] <cbx33> ??
[05:19] <cbx33> if some one puts exec firefox in the queue
[05:19] <ogra> if i click a second time on execute it should execute a second time
[05:19] <cbx33> of course it should
[05:20] <cbx33> but how does the mesasge get cleared from the queue?
[05:20] <cbx33> or fomr the bus
[05:20] <cbx33> the scp server deletes it after a certain amount of time
[05:20] <cbx33> ?
[05:20] <cbx33> it can't just sit there, otherwise the client will keep acting on it?
[05:20] <ogra> for example ... doesnt dbus have a ttl mechanism ? 
[05:20] <cbx33> not at the level we are using
[05:21] <cbx33> some of the things you wrote were extremely low level
[05:21] <ogra> then delete it after n seconds
[05:21] <cbx33> and we don;t need to worry about
[05:21] <cbx33> but then the poll time needs to match that exactly !!
[05:21] <cbx33> if it's under the command can be run twice
[05:21] <ogra> well, mdz and iwj wanted it like that, i found it overkill as well ;)
[05:21] <cbx33> if it's over it could get skipped
[05:22] <cbx33> ogra, can I propose another system?
[05:22] <ogra> sure
[05:22] <cbx33> well,
[05:22] <cbx33> I think a message queue, with a ttl on each message is a lot better,
[05:22] <cbx33> each message has an incremented number
[05:23] <cbx33> and the client just keeps count of what the last number it acted on was
[05:23] <ogra> ok
[05:23] <cbx33> if a message appears that applies to it, that is greater than the last count, run it
[05:23] <cbx33> if not ... ignore it
[05:23] <ogra> but the messages must disappear from the bus anyway
[05:23] <cbx33> yes
[05:23] <cbx33> after the TTL
[05:24] <cbx33> I'm just thinking this part of the system has to be pretty stable
[05:24] <cbx33> if it crashes teachers lose control etc
[05:24] <ogra> the message it acts on must be listmessagenumber+1
[05:24] <cbx33> yes
[05:24] <cbx33> and apply to it
[05:24] <cbx33> user name etc
[05:25] <ogra> sure thats out of question
[05:25] <cbx33> I also think that the 4 message system shoudl be changed for expandability
[05:25] <cbx33> why not send one type of message
[05:25] <cbx33> with a command as a prefix
[05:25] <cbx33> instead of list, kill, exec
[05:26] <cbx33> what do you think?
[05:27] <sbalneav> ogra: I have an idear
[05:27] <ogra> you mean: "exec, firefox [user1, user2, usern] " ?
[05:27] <cbx33> ogra, kinda yeh
[05:27] <ogra> sure, why not
[05:27] <sbalneav> instead of callig gnome-session directly...
[05:27] <cbx33> I'll write up a new spec
[05:27] <cbx33> for that section
[05:27] <ogra> we dot call gnome-session at all
[05:27] <sbalneav> Why not call a script that looks like:
[05:27] <sbalneav> #!/bin/sh
[05:27] <ogra> *dont
[05:27] <sbalneav> gnome-session
[05:27] <sbalneav> # clean up mounts.
[05:27] <ogra> its a prerequisite that we use the defaul XSession of the system
[05:28] <ogra> we never called gnome-session
[05:28] <sbalneav> Oh?
[05:28] <sbalneav> heh, never looked at what we called. 
[05:28] <sbalneav> oh!
[05:28] <sbalneav> Do we call the xsession script?
[05:29] <ogra> /etc/X11/Xsession
[05:29] <sbalneav> ah, perfect.
[05:29] <sbalneav> We can put some cleanup in /etc/X11/Xsession.d
[05:29] <ogra> so you can add it to XSession.d instead ;)
[05:29] <ogra> desktop independent ;)
[05:29] <sbalneav> hah, beat you to it.
[05:29] <sbalneav> problem solved.
[05:30] <ogra> and you even wrote the full path !
[05:30] <sbalneav> :)
[05:30] <cbx33> ogra, is that ok for me to come up with a new spec for it, or shall I just go ahead and code it
[05:30] <cbx33> and come up with a spec later :p
[05:30] <ogra> go ahead and code it and add it in an extra section to the spec, so i can wave it in
[05:30] <cbx33> ogra, ok, only if you are happy with that implementation?
[05:31] <ogra> yep, thats fine
[05:31] <ogra> it adresses some weaknesses ;)
[05:31] <cbx33> cool
[05:32] <cbx33> it'll be early next week now :p
[05:32] <cbx33> I'm away tomorrow and sat
[05:32] <cbx33> and I have to have logout sounds done for 3 themes by tuesday
[05:42] <ogra> sbalneav, what do you think about gconf integration ?
[05:44] <ogra> ltspfsmounter could check if a key is set for certain device types i.e. user is allowed to use local cd or local disks ... 
[05:44] <ogra> that would enable the admin to lock down the key in a gui way via pessulus ...
[05:44] <ogra> gives fine grained access control
[05:49] <sbalneav> That sounds awesome, but I don't know diddly bupkis about gconf.
[05:49] <sbalneav> You'd use something like gconf-tool2 or something to look up the keys.
[05:49] <sbalneav> ?
[05:49] <sbalneav> Sorry, I'm at work, I'm rummaging around for a spare terminal to play now at  work :)
[05:50] <ogra> heh
[05:50] <ogra> there is python-gconf
[05:50] <sbalneav> Oh, right, you were thinking of re-writing the mounter in python.
[05:51] <ogra> yep
[05:51] <sbalneav> I think that's an awesome solution.  It will integrate well with pessalus/sabayon, and gives fine-grained control on a user-by-user basis
[05:54] <jsgotangco> good night
[05:56] <ogra> well, we have that anyway by the fuse group already ;) but it would be finer grained
[06:13] <cbx33> why is all my text anti aliased now :S
[06:14] <Petaris> cbx33: its supposed to make it pritier
[06:14] <cbx33> hi LaserJock 
[06:14] <cbx33> Petaris, it makes it more diffiult to read on this LCD
[06:14] <cbx33> :S
[06:14] <Petaris> yeah, I have a monitor or two it utterly sucks on to
[06:14] <Petaris> makes it look fuzzy
[06:14] <sbalneav> aaaaaargh!!!!
[06:15] <sbalneav> 99xorg-common-startup:
[06:15] <sbalneav> exec $STARTUP
[06:15] <sbalneav> crumb!
[06:15] <sbalneav> ogra: That something we can change?
[06:16] <ogra> why ? 
[06:16] <sbalneav> well, it's sourced.
[06:17] <sbalneav> so if you do an exec of the startup (i.e. gnome session)
[06:17] <sbalneav> there's no hope of slotting anything in afterwards :(
[06:17] <cbx33> ogra, there is something wrong with taht ACL
[06:17] <cbx33> :(
[06:17] <sbalneav> because xsession script's gone :(
[06:17] <cbx33> I'll try and sort that out first
[06:17] <cbx33> 2/3 logout sounds done
[06:17] <cbx33> :D
[06:18] <ogra> sbalneav, i thought you wanted to clean up before 
[06:18] <sbalneav> Before?
[06:18] <ogra> before the session starts
[06:18] <ogra> we cant change the ubuntu defaults there ...
[06:18] <sbalneav> So when the user logs out, you leave the dead bindmounts and all around?
[06:19] <sbalneav> We could clean up before, I suppose.
[06:19] <ogra> well, no
[06:19] <ogra> i wrote you something in my last mail ;)
[06:20] <sbalneav> oh, you're talking about what petter did?
[06:20] <ogra> yep
[06:21] <sbalneav> ah, ok, so you're on that case then.
[06:21] <ogra> ltspfsmounter can get a "cleanup" command
[06:21] <ogra> so ldm just needs to start (ugly way) /etc/X11/XSession && ltspfsmounter cleanup
[06:21] <ogra> but along these lines ...
[06:21] <sbalneav> ahhhhhhhhhhhhhhhhh
[06:22] <sbalneav> that's not ugly
[06:22] <sbalneav> that's brilliant
[06:22] <sbalneav> perfect
[06:22] <sbalneav> You're so smart.  When I grow up I want to be just like you! :)
[06:22] <ogra> heh
[06:23] <ogra> i can only give that back :)
[06:32] <cbx33> ogra, :D:D
[06:32] <cbx33> this is going well
[06:32] <ogra> cool !
[06:32] <cbx33> apart from the ACL
[06:32] <cbx33> I can send messages to the system bus, even though I'm not root
[06:33] <cbx33> but I can't own the bus unless I'm root
[06:34] <cbx33> so in essence the server script is working fine and restricted ok
[06:34] <cbx33> however the client (scp) tool can still go about it's business as a normal user
[06:37] <sbalneav> ogra: ok, I'm having a dopey moment.  I've set up the ltsp chroot, and I've got the terminal booted.  done the ltsp-update-sshkeys, but when I log in, it just boots me back to the login.  What have I missed?
[06:39] <ogra> hmm
[06:39] <ogra> set a root pw and look in the ldm.log ? 
[06:39] <ogra> (on the terminal)
[06:40] <ogra> and look in ~/.xsession-errors 
[06:40] <ogra> if you ran the update-sshkeys it should simply work ... is your sshd running? 
[06:40] <sbalneav> ah, there's the proble,
[06:41] <sbalneav> i've got 2 cards in this box, and ltsp-update-sshkeys didn't add an entry for the card I'm running the term on
[06:41] <sbalneav> humm
[06:41] <ogra> thats weird
[06:41] <ogra> shouldnt happen
[06:42] <sbalneav> re-ran it
[06:42] <sbalneav> lets see
[06:42] <ogra> is the interface up ?
[06:43] <ogra> indeed it is ... you just booted from it :)
[06:43] <ogra> silly me
[06:44] <cbx33> ogra, regarding the ACL issue, could it be that we never actually denied send access to the interface before we allowed it from root?
[06:44] <cbx33> so therefore by "default" everyone has send access
[06:45] <cbx33> got it
[06:45] <cbx33> that was it :D
[06:46] <cbx33> we now have working ACL 
[06:46] <cbx33> w00t
[06:49] <ogra> sbalneav, any idea why esd would connect to .ltspfs-socket ? 
[06:49] <ogra> lsof -t gets me esd in the output
[06:53] <sbalneav> no idea at all
[06:53] <sbalneav> as for my logon problem: "Xsession started for user scbal..."
[06:53] <sbalneav> No profile found for user scbal
[06:53] <sbalneav> huh
[06:53] <ogra> sabayon ...
[06:54] <ogra> create a profile or remove sabayon
[06:54] <sbalneav> ohhhh
[06:56] <sbalneav> yep that was it.
[06:56] <sbalneav> that needs.... a better error message :)
[06:56] <ogra> yep
[06:56] <ogra> ldm needs them ;)
[06:56] <sbalneav> ok, lunchtime walkies!
[06:56] <ogra> enjoy
[06:57] <sbalneav> Need to keep my svelte figure :)
[06:57] <ogra> :)
[06:57] <sbalneav> bb in a bit
[07:26] <ogra> hmm, the cdpinger doesnt wotk on uplug anymore either ...
[07:28] <ogra> oh, right, the check_mounted function is gone ...
[07:28] <ogra> so it doesnt clean u anymore ...
[07:28] <ogra> *up
[07:30] <cbx33> ogra, we could use the scp daemon to poll users and make sure that the system message bus listener service is running for each user
[07:38] <ogra> daemon ? 
[07:39] <ogra> please dont make that bigger than it is ... there should be a simple subfunction in scp 
[07:39] <ogra> no daemons ...
[07:39] <cbx33> I meant the scp message bus server
[07:40] <ogra> that should be scp itself 
[07:40] <ogra> no separate rogram
[07:40] <cbx33> oh right
[07:40] <ogra> and keep it as simple as possible 
[07:41] <cbx33> ogra, then
[07:41] <ogra> no two way communication or anything ... thats fine for version 2
[07:41] <cbx33> please explain in the spec
[07:41] <cbx33> the difference between server and client
[07:41] <ogra> server = scp 
[07:41] <ogra> client = the listener in the users session
[07:42] <ogra> i thought that was explained there 
[07:42] <cbx33> read the execution of the application
[07:42] <cbx33> that paints a completely different picture
[07:42] <cbx33> half the work I've done is not required
[07:42] <cbx33> if we're implementing in the way you have just said
[07:43] <ogra> A Student Control Panel listener service file with the namespace 'com.ubuntu.StudentControlPanel' will be installed in /usr/share/dbus-1/service.d/ which will listen for messages from Student Control Panel. 
[07:43] <cbx33> read further down
[07:43] <ogra> doesnt talk about a "student control panel server"
[07:43] <cbx33> further down it gives that impression
[07:43] <ogra> but about student control panel
[07:45] <ogra> oh, right Client here = SCP 
[07:45] <ogra> i'll add the word schematic to the caption of that piece ...
[07:45] <cbx33> so client and server are the same thing?
[07:45] <ogra> no
[07:45] <cbx33> then why does it have to communicate with itself over dbus?
[07:46] <cbx33> ok what is server
[07:46] <ogra> SCP communicates with a tool in the users session
[07:46] <ogra> what you are referring to is a plain schematic picture 
[07:46] <ogra> i could have written A and B instead of Server and Client
[07:46] <cbx33> yeh I understand that
[07:46] <ogra> or X and Y
[07:47] <ogra> its just to show the dataflow
[07:47] <cbx33> but it implies that the A talks to B
[07:47] <ogra> nothing you should base an implementation on
[07:47] <cbx33> and the C picks up the messages from the bu
[07:47] <ogra> right
[07:47] <ogra> they are two different apps
[07:47] <cbx33> so who are A and B
[07:48] <ogra> A is SCP
[07:48] <ogra> B is the tool running in the users session
[07:48] <cbx33> and C?
[07:48] <ogra> ??
[07:48] <ogra> there are only two ends on a sausage ... what do you mean with C ?
[07:50] <cbx33> ok....
[07:50] <cbx33> line by line....
[07:50] <cbx33> Client (SCP) sends the message
[07:50] <cbx33> to where?
[07:50] <ogra> sigh
[07:50] <ogra> dbus
[07:51] <ogra> on the system bus
[07:51] <cbx33> Server (who is this) verifies the message
[07:51] <cbx33> unless by this you mean server as in LTSP
[07:51] <ogra> no
[07:51] <cbx33> the dbus server
[07:51] <ogra> forget about the term server
[07:51] <cbx33> ogra, I get it now....
[07:51] <cbx33> that entire section
[07:52] <cbx33> is the authentication
[07:52] <cbx33> o my word
[07:52] <cbx33> I can;t beleive I was so THICK !
[07:52] <cbx33> ogra my apologies.....I totally mis understood taht section
[07:54] <cbx33> can you ever forgive that stupidity
[07:59] <ogra> even if we might have a remote executable SCP at some point for remote management of multiple SCP servers, there will only be one instance running on the respective server 
[07:59] <ogra> s/muliple SCP servers/multiple LTSP servers/
[07:59] <cbx33> my case was two teachers in different class rooms
[07:59] <cbx33> wanting to control students
[08:00] <cbx33> both rooms powered by the same ltsp server
[08:01] <ogra> that cant work ...
[08:01] <ogra> unless you have userlists and databases who is in which room
[08:01] <cbx33> or databases of which machines are in which room
[08:01] <cbx33> which is how our system at school works currently
[08:02] <ogra> how would you know
[08:02] <ogra> its 100% dynamically .... all you could use here would be the MAC address of the client
[08:02] <cbx33> we use a commercial package at the moment
[08:02] <cbx33> called ranger remote
[08:02] <ogra> thats the only thing thats not dynamically
[08:02] <cbx33> yes it uses macs
[08:02] <ogra> and its way beyond SCP at the current state 
[08:02] <cbx33> ok
[08:03] <ogra> i really need to go on coding now
[08:03] <cbx33> once again I apologise for making you pull your hair out
[08:03] <ogra> nah, it was very badly written, yu are right 
[08:03] <cbx33> sorry again
[08:51] <sbalneav> ogra: Got time for a quick techie discussion?
[08:54] <ogra> sure
[08:55] <ogra> let me make a fresh coffee first :)
[08:55] <sbalneav> ok, while you're making coffee, I'll type: pitti wants us to make sure when we unmount the directory, it's ours...
[08:56] <sbalneav> However, I can't do a stat() call, because EVEN IF I DROP PRIVS!, fuse stops us because the binaries setuid, so I can't stat the directory.
[08:56] <sbalneav> However, what I CAN do, is check /proc/mounts for the mountpoint:
[08:57] <sbalneav> which will show somehing like uid=1000 in the options field.
[08:57] <sbalneav> So, before unmounting, I look for the dir in proc mounts, see that it's got a bindmount owned by me, and it should be safe, yes?
[08:57] <sbalneav> That make sense?
[08:58] <sbalneav> Pitti's not in -devel, otherwise i'd ask him
[08:59] <ogra> he'll be around later, we have distro meeting at 1am (local time)
[08:59] <sbalneav> 1am?
[08:59] <sbalneav> Wow
[08:59] <ogra> 23:00 UTC
[09:00] <sbalneav> I'm going to assume that's valid, and code it for now.  
[09:00] <sbalneav> since I KNOW that the stat() call won't work. :)
[09:00] <ogra> ok
[09:00] <sbalneav> It's this or don't do the test :)
[09:01] <ogra> yeah
[09:01] <ogra> i'D rather dont ) 
[09:01] <ogra> :)
[09:01] <ogra> it starts getting a lot of bloat here
[09:02] <pygi> ogra, libburn burns ^_^
[09:03] <sbalneav> Don't think pitti will like that.  And we want to stay in his good books :)
[09:03] <ogra> sbalneav, btw whyts /tmp/foople for ? :)
[09:04] <sbalneav> argh.  Debugging stuff left in.
[09:04] <ogra> lets use ~/.xsession-errors :)
[09:04] <sbalneav> anything I send you while developing might have wibbly bits in it :)
[09:04] <ogra> probably even for cdpinger etc *g* we could send it through the socket :P
[09:05] <sbalneav> I name all my debugging files/variable/etc "foo*", then after I've got everything working, I do a "grep foo *" on my source, and clean them all up :)
[09:05] <ogra> ok, looks like i have cdpinger running fully again here with the usb cdrom ... it doesnt cry if i unplug etc ...
[09:06] <ogra> ah, nice :)
[09:06] <ogra> ogra@edubuntu:~/devel/ltspfs/ltsp-localdev-scott$ fusermount --help
[09:06] <ogra> bash: /usr/bin/fusermount: Permission denied
[09:06] <ogra> grmbl
[09:07] <ogra> I ONLY WANT YOUR HELPSCREEN YOU DARN THING !
[09:07] <ogra> sorry :)
[09:08] <ogra> sad, there is no --user option for the unmount :)
[09:08] <sbalneav> yeah, fuse is HYPER secure
[09:09] <ogra> would be nice to just fusermount -uzq --user $USER 
[09:09] <ogra> to unmount everything 
[09:09] <sbalneav> ogra: ah, so that's the hacked up cdpinger that I gave you last night being launched from udev?
[09:09] <ogra> nope
[09:09] <ogra> cdpinger itself had some serious bugs 
[09:10] <ogra> i rewrote the lsof part completely ...
[09:10] <ogra> fiddled around with the unmount .... (it started to unmount over and over when i unplugged ...
[09:10] <ogra> )
[09:11] <ogra> and re-added the check for the link and the closing of the filehandle ...
[09:13] <sbalneav> awsome. mail me please please please please please please pretty please
[09:14] <ogra> suure
[09:17] <ogra> sbalneav, sent
[09:17] <sbalneav> thx
[09:17] <sbalneav> are you starting it from udev, or still just statically?
[09:18] <ogra> still statically 
[09:18] <ogra> with one default one started with cdrom as argument 
[09:18] <ogra> and after that i'm running the loop for multiple CDs you added 
[09:23] <cbx33> ok ogra I'm getting there again
[09:23] <sbalneav> Cool, lookin HOT
[09:23] <cbx33> my listener is just not receiving the signals
[09:23] <cbx33> but it will receive HAL signals
[09:23] <cbx33> :
[09:23] <cbx33> S
[09:23] <ogra> sbalneav, oh, the check_mount function can be dropped again
[09:24] <ogra> i did that onyl for s test
[09:24] <ogra> *some
[09:25] <ogra> oh, i didnt even realize you write directly to fstab :)
[09:25] <ogra> to sad thats not possible with sed
[09:25] <ogra> thats bothering be in the remove_fstab_entry script
[09:26] <ogra> *me
[09:26] <sbalneav> yeah, I just figured why call an extra program when we're already python.
[09:26] <sbalneav> figured it might be faster.
[09:26] <ogra> (sed needs to write a .swp file to /etc for it :/)
[09:27] <ogra> (and mv doesnt work either because t needs to remove the file first)
[09:27] <ogra> heh ...
[09:28] <ogra> we could make cdpinger a python module :P
[09:28] <ogra> then i could just use the functions from it in the add/remove_fstab_entry scripts
[09:30] <ogra> hmm ...
[09:30] <ogra> hmm ...
[09:30] <ogra> hmm hmm
[09:31] <ogra> what about just starting/killing cdpinger from these scripts ?
[09:31] <ogra> for all devices
[09:32] <ogra> not ... that'll eat to much mem
[09:33] <sbalneav> yeah, I like it how it is for memory sticks: there's nothing hanging around, the connection's plumbed, and then it's done.
[09:33] <ogra> but in fact we do a lot redundant stuff
[09:33] <ogra> all the functions we have in cdpinger are used in the udev scripts in shell ...
[09:34] <sbalneav> brb, work ping
[09:46] <ogra> sed -e "/^\/dev\/$1 /d" /etc/fstab >/tmp/fstab
[09:46] <ogra> cp /tmp/fstab /etc/
[09:46] <ogra> rm /tmp/fstab
[09:46] <ogra> thats better :)
[09:46] <ogra> still not perfect but well
[09:49] <sbalneav> gad, the nerve of these people, expecting me to work on LEGALAID related matters during working hours!  Don't they know I've got Ubuntu stuff to take care of? :)
[09:49] <ogra> heh
[11:42] <cbx33> pygi, know a good way in python to return the current username?
[11:45] <rodarvus> cbx33, 
[11:45] <rodarvus> >>> import os
[11:45] <rodarvus> >>> os.getlogin()
[11:45] <rodarvus> 'rodarvus'
[11:45] <rodarvus> >>>
[11:45] <cbx33> oooh thanks rodarvus 
[11:45] <rodarvus> :)
[11:45] <LaserJock> yeah, that os module is nifty
[11:45] <rodarvus> I wish all questions could be as easily answered as this one ;)
[11:46] <cbx33> I looked at the docs
[11:46] <cbx33> but couldn't find one
[11:46] <rodarvus> cbx33, the interactive interpreter is the *best* doc you can find in python
[11:46] <rodarvus> dir (module) does wonders
[11:46] <LaserJock> ok, I've got like 2 hrs to come up with a talk about Launchpad :-)
[11:46] <LaserJock> hmmmm
[11:51] <cbx33> EASY
[11:53] <LaserJock> hehe
[11:53] <LaserJock> yeah right
[11:54] <rodarvus> how long is your talk going to be?
[11:54] <rodarvus> what aspects of LP are you going to talk about?
[11:54] <LaserJock> well, 1hr
[11:54] <rodarvus> LaunchPad is *huge* stuff :)
[11:55] <LaserJock> of course
[11:55] <LaserJock> I was supposed to be a backup because I'm giving another talk
[11:55] <LaserJock> but nobody wanted to touch LP ;-)
[11:55] <LaserJock> but the conference is pretty informal
[11:55] <rodarvus> LaserJock, you're going to talk about LP itself, Malone, Rosetta, Soyuz, Blueprint, or something else?
[11:55] <LaserJock> so I think I'll talk about Malone, Rosetta, Soyuz, Blueprint
[11:56] <LaserJock> briefly show how to get around a bit
[11:56] <LaserJock> maybe a little triaging or something
[11:56] <LaserJock> and then maybe have more of a discussion
[11:56] <LaserJock> about peoples experiences with it or something
[11:57] <LaserJock> that's the best I can figure out
[11:58] <sbalneav> Hello all
[11:59] <LaserJock> hi scotty
[12:10] <cbx33> ogra, ok
[12:10] <cbx33> kill is now performed via dbus
[12:10] <cbx33> I've done it
[12:10] <cbx33> 70% of the work done
[12:10] <cbx33> just need to do execute