--- Day changed

--- Log opened Tue Jun 03 00:00:01 2003
03:42 < BB2> morn
04:06 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has joined #uml
05:04 < green> hmmm... it is snowing outside...
05:04 < green> I must be missing something
05:26 < Valen_> Is it possible to edit a UML rootfs file outside the UML itself ?
05:27 < Valen_> I've a "pristine" filesystem image, and a script that sets up new UMLs. I'd love to mount the rootfs, combined with the "changes" file, to put in an IP etc. before booting UML.
05:38 < ichilton> Valen_: not if you use COW.
05:38 < ichilton> Valen_: with a normal rootfs, you can do: mount -o loop rootfs /mnt
05:39 < Valen_> ichilton: ah. I'm using COW. I'm looking at some mad programming with expect then, to startup the UML, login, change the networking files/password, logout, shutdown then.
05:39 < Valen_> The standard linux kernel doesn't have UBD support then ?
05:56 < ichilton> Valen_: I think I recall something on the mlist a while ago about someone wanting to write something to mount cow but afaik it's still not possible
05:56 < ichilton> Valen_: there is something in uml_utilities (uml_moo iirc) which merges the backing file and cow file to produce 1 file which can be mounted with mount, but there is then no way of splitting them again.
05:57 < Valen_> Shame. It'd be nice for automatically pre-configuring COW setups.
05:57 < ichilton> Valen_: if it's just ip's you're worried about, you should be able to pass in a unique mac on the kernel command line and use dhcp to assign the ip, but I persoanlly just boot the new with an unused ip and change it to the one I want to use.
05:59 < Valen_> ichilton: I've a box with 30 IPs, so it's quite a lot of work to do it for each one. How would I come up with MAC addresses for each one ?
06:01 < Valen_> Is the "jail" option useful ? It seems it'd be a little more secure..
06:03 < ichilton> Valen_: not sure
06:04 < ichilton> Valen_: I think I heard jail is slower
06:04 < ichilton> Valen_: and I think skas made it obsolete
06:04 < ichilton> Valen_: normally, the mac is generated from the ip.
06:04 < ichilton> Valen_: maybe you can also pass in an ip on the command line?
06:09 < Valen_> ichilton: maybe.
07:16 < snide> ichilton: i had the idea of writing a COW host driver. but actually i'm too weak on that subject
07:23  * gump is back -( bed )- gone 7 hr 38 min 50 s
07:25 -!- radical [~kernet@202.88.184.39] has joined #uml
07:37 < snide> ichilton: i'm exploring now about hacking the nbd server (since it runs in userspace) and to be able to mount COW as a nbd device (yeah, i know, it's ugly, but it's a workaround.. better than starting an uml and NFS-ing there.)
07:52 < ichilton> snide: ah yeah - I remember now!
07:52 < ichilton> snide: cool
07:52 < ichilton> snide: any "hack" is beeter than nothing at all
08:02 < gump> btw ive got a few m8s who are interested in hgelping with uml
08:02 < gump> were gonna have a go at writing a web interface for umld
08:02 < gump> once i get my box set up =]
08:13 < radical> i have a function called find(..) which is returning a pointer to a struct, the values in the struct when checked inside find(..) are correct, but when received by the calling function.. the values are changed! 
08:14 < radical> but after this callin function is over.. if i try to access the memory addr of the values directly the values are fine.., so seems like i'm not getting the right values in this particular function.. any ideas? 
08:14 < radical> this is under 2.4.20-um4
08:40 -!- avoozl [~jrk@vengeance.et.tudelft.nl] has joined #uml
08:41 < avoozl> what should i do to make tuntap networking work when i run uml as an ordinary user? as root everything works fine, but as a user i get 'Configuring network interfaces: tuntap_open_tramp failed - errno = 22' , even after using tunctl -u myuser -t tap0
08:43  * gump is away -( uni )- at 01:42p -( P:On / L:On )-
08:43 < BB> do you have /dev/net/tun writable by the user?
08:43 < avoozl> yes i changed the owner of it
08:43 < avoozl> just to be sure
08:44 < Getty> avoozl: debian packages?
08:44 < avoozl> Getty: yes
08:44 < avoozl> Getty: did a rootstrap
08:44 < Getty> avoozl: debian packages are stupid
08:45 < Getty> cp /usr/lib/uml/uml_net /usr/bin/uml_net
08:45 < Getty> and add the user to the umlnet group of course
08:45 < avoozl> i didnt use uml_net
08:45 < Getty> but uml does ;)
08:45 < avoozl> ohh
08:46 < Getty> thats the problem
08:46 < avoozl> cool
08:46 < Getty> its not in the path, uml doesn't find it
08:46 < avoozl> i would have never found that myself :)
08:46 < Getty> yeah
08:46 < Getty> it took me about 6 hours
08:46 < Getty> i always thought i was too stupid for the network shit
08:46 < Getty> i tried all network methods
08:46 < Getty> and so on...
08:46 < Getty> ugly, really
08:46 < Getty> someone should mail the maintainer....
08:46 < avoozl> i almost reverted back to slirp
08:47 < Getty> hehehe
08:47 < Getty> yeah, that was the point where i said... no that can't be the solution
08:49 < avoozl> works like a charm now :)
08:49 < avoozl> thanks
08:49 < Getty> out to lunch
08:49 < Getty> no problem
08:49 < Getty> always a pleasure to help
08:58 -!- ASY [~someone@216.221.51.54] has joined #uml
10:47 -!- radical [~kernet@202.88.184.39] has quit [Remote host closed the connection]
11:04 -!- Lathiat [~Lathiat@203.15.140.23] has quit [Quit: rescreenness]
11:04 -!- Lathiat [~Lathiat@203.15.140.23] has joined #uml
11:26 < Valen_> snide: on the nbd/COW hack...if you need someone to test, let me know.
11:28 < snide> Valen_: i'll need first some time to make something that is in a better state than 'it *almost* works for me' [ ie leaving the almost part ]
11:29 < ASY> does anyone use umld here (or is david on?).  need a little help setting it up...
11:29 < snide> Valen_: bad news is, i don't have much time those days... [ i'm currently working on integrating shfs ]
11:29 < david> ASY: hello
11:29 < ASY> hi david.  pleased to meed you. :)
11:30 < snide> Valen_: but, actually.. i need to see why my tap0 hangs...
11:30 < david> ASY: likewise
11:30 < ASY> been fighting with umld for a little wihle now.  was quite exciting trying to figure everything out by myself :)  
11:30 < david> uh oh :-)
11:31 < ASY> anyhow.  i have umld running. got a basic profile setup, paths in order...  i actually can do create name user profile...
11:31 < david> ASY: impressive :-)
11:32 < snide> david: it seems umld isn't in "iwfm" state anymore
11:32 < ASY> i started dieing in function proxy_uml_build($conn,$umid,$prof,$dist) at which point i realized that there is more then a single system at play.  looks like you are using a separate admin uml to perform various functions...
11:32 < snide> ASY: i think it's to handle COW files
11:33 < ASY> yes
11:33 < Valen_> How useful is umld ?
11:34 < ASY> it looks very good.  but challenging to get it going :)
11:35 < david> ASY: right, you need a seperate UML to create the COW and chroot to dist-upgrade it
11:36 < ASY> however it looks like i have to make this uml manually...  since i can't create one with umld...  (kindof chicken and the egg problem).
11:37 < ASY> i just wanted to ensure what i am doing is correct before i go into a quest of setting this up...
11:37 < david> ASY: right, but there is no other way to setup COW
11:38 < ASY> can you refresh me on what copy on write does exactly?
11:38 < snide> david: soon there should be. (i hope)
11:39 < david> ASY: avoids having a 2Gb file for each UML :-)
11:40 < ASY> that i actually know. it is the means by which i am interested in. am i correct to assume that this basically stores only the data written in the filesystem?
11:42 < snide> but, actually, i do have another pb... i can't do nbd to localhost ;-(
11:42 < david> ASY: right
11:43 < snide> ASY: COW = Copy On Write
11:44 < ASY> snide: [11:36] <ASY> can you refresh me on what copy on write does exactly?
11:44 < ASY> :)
11:45 < ASY> ok, so I will go setup admin uml.  put it in the config and add it to the umls table manually...
11:46 < ASY> and then i will probably be back here :)
11:46 < snide> ASY: it reads from the backing file... and when it is writing time, it writes into the COW file (and all subsequent read of this sectors are from the COW file)
11:47 < ASY> snide: thanks, i understand.
11:47 < snide> ASY: ;-)
11:48  * snide is going back home
11:48 < snide> bbl
11:48 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has quit [Quit: BitchX: to the rescue!]
11:49 < ASY> david, before i go, another quick question...  if I do "uml test"  I seem to go into the context of this uml as far as command line goes...
11:49 < ASY> how do i move up from the context?  > retr   doesn't seem to have any effect
11:49 < david> ASY: yep
11:51 < david> ASY: yeah, I screwed that up
11:51 < david> ASY: I'll fix that in the next release
11:51 < david> ASY: retr should work, but I rejigged the command parsing code and it's now commented out
11:51 < ASY> oh. and i have another  (i didn't realize i've accumulated so much).  although i was going to look through the code for this... 
11:51 < ASY> auth user  name
11:51 < ASY> pass <password>
11:51 < david> heh
11:52 < ASY> this password comes form where and should be in which form  (or is this a TODO item ?) :)
11:52 < david> ASY: okay, does that not work?
11:52 < david> ASY: it uses posix_getpwnam() first
11:52 < david> ASY: there is also a db table called 'crampass' which you'll find if you grep for it
11:53 < david> ASY: it's a plain text password field which is checked with 'auth user', but can be used with CRAM-MD5
11:53 < david> ASY: look at the code for authentication - It's pretty basic
11:54 < ASY> ok,  let me play a little more around this and I will get back on this.
11:54 < ASY> any plans to setup a CVS rep for this?
11:58 < ASY> well, thanks for help. I have to run grab some food.  ttyl
12:24 -!- BB2 [~chris@217.79.122.210] has quit [Quit: My damn controlling terminal disappeared!]
12:24 -!- BB1 [~chris@217.79.122.210] has joined #uml
12:50 < avoozl> how unsafe is a uml running without the jail option?
12:50 < avoozl> what exactly can 'break out'?
12:50 < caker> avoozl: skas mode achieves better security than even tt with jail mode, i believe
12:51 < caker> avoozl: plus it is faster, etc
12:52 < avoozl> hmm ok... but currently i run a uml with neither of those options, since jail didnt work because modules and hostfs were enabled
12:52 < avoozl> i just wondered in which way it is considered unsafe
12:52 < avoozl> and if i should go trough the trouble of setting it up for skas or jail
12:53 < caker> skas is the way to go
12:53 < avoozl> euh yes ok, i understand that between jail and skas, the latter is best
12:55 < avoozl> oh the skas page explains it well
12:55 < avoozl> thanks for pointing me to skas
12:55 < avoozl> http://user-mode-linux.sourceforge.net/skas.html
12:55 < caker> Yup
12:55 < caker> no problem, avoozl!
13:02 < david> avoozl: jail mode is completly useless
13:02 < david> avoozl: just use skas
13:03 < david> avoozl: jail, in my tests, wasmaybe 3000% slower than standard tracing thread, so it's basically unusable
13:21 < ichilton> hi david 
13:35 -!- Valen_ [~valen@sprocket.hosting365.ie] has quit [Quit: Client Exiting]
13:37 < avoozl> david: im currently building a client and host kernel for skas
13:38 < avoozl> i'll be a while :)
13:40 < avoozl> i seem to get some compile errors
13:40 < avoozl> i fetches uml-patch-2.4.20-5 and applied it on a 2.4.20 vanilla
13:40 < avoozl> during an ordinary make i get : sys_i386.c: In function `do_mmap2':
13:40 < avoozl> sys_i386.c:59: warning: passing arg 1 of `do_mmap_pgoff' from incompatible pointer type
13:40 < avoozl> sys_i386.c:59: warning: passing arg 2 of `do_mmap_pgoff' makes pointer from integer without a cast
13:40 < caker> did you do ARCH=um ?
13:40 < avoozl> am i building the wrong target or something?
13:41 < avoozl> thats what i feared :)
13:41 < caker> uml-patch will make a UML kernel, then in a seperate clean tree, apply the skas patch to make a host kernel
13:42 < avoozl> caker: that's what i thought, i just forgot to set any ARCH vars.. let me dig up the howto again
13:42  * caker is afk for a bit
13:43 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has joined #uml
13:52 < avoozl> ok the only thing left now is: vmlinux.o(.text+0xd0e6c): In function `rtnetlink_rcv':
13:53 < avoozl> : undefined reference to `rtnetlink_rcv_skb'
13:53  * avoozl will toggle some random related options in menuconfig
13:55 < green> avoozl: well, you compile with gcc 3.3m so you need to patch the kernel a bit
13:56 < avoozl> green: oww :) i wasn't aware
13:57 < green> avoozl: look for rtnetlink_rcv_skb somewhere in net/*, it is devined as extern inline, replace extern with static
13:57 < avoozl> ohhh yeah
13:57 < avoozl> i recall
13:57 < avoozl> i completely forgot
13:57 < avoozl> i had to do that for my other kernel as well, but it has been a while
13:58 < green> hehe
13:58 -!- shanti [~shanti@157.182.194.85] has joined #uml
14:03 < caker> time for a motorcycle ride - I've been staring at a screen for days
14:03 < green> ΘΕΘΕ
14:04 < green> hm. whong charset
14:04 < BB1> heh
14:04 < caker> um yeah :)
14:04 < BB1> u trying to type 44 ;)
14:07  * green ordered new box that will also hold my uml, because the old one had postgres running too slow. And now I have figured how to make postgres run fast. Sigh
14:07 < caker> oh man
14:07  * BB1 ordered new box today too :)
14:08  * caker spent 10k out of pocket on gear this year
14:08  * BB1 spent £3K in one day
14:08 < caker> my bank statements are scary
14:08 < BB1> you still look at yours? 
14:08  * green bought a new car and got in a car accident in a month.
14:08 < caker> no hehhee
14:08 < caker> oh crap
14:08 < BB1> caker me either ;)
14:09 < BB1> green doh :(
14:09 < avoozl> everything except tuntap is working now :)
14:09 < green> and got in second one in another three months and the car cannot be repaired anymore. At least I am alive
14:09 < caker> bad weather up there or something?
14:10 < green> got asleep for the first time, and bad weather on the other one.
14:10 < avoozl> something rather with Configuring network interfaces: tuntap_open_tramp : didn't receive a message
14:10 < avoozl> must be something i forgot in the kernel config i suppose
14:13 < avoozl> (note that i didnt reboot the host kernel with skas yet, but i thought that it should work like this already as long as i dont use skas)
14:17 < avoozl> hm im back to the same problem i had a few hours ago, it works fine as root, and fails as a user, but this time uml_net is owned by the right group
14:18 < green> what about set-uid bit?
14:18 < avoozl> it doesn't have any
14:19 < avoozl> i made the tun0 device owned by the user that is running uml, that should do it right?
14:19 < green> no
14:19 < avoozl> uhm but it works with my older uml kernel
14:19 < green> you need to be root to configure interface so either you configure interface as root by hand or set umlnet to be setuid root
14:20 < avoozl> i configured it by hand
14:20 < green> if you configured it by hand, then you do not need to call uml_net at all
14:20 < avoozl> my old debian-shipped uml kernel works, both as a user and as root
14:20 < avoozl> my new compiled kernel refuses to work as a user, but does work as root
14:20 < green> and have you upgraded uml_tools too?
14:22 < avoozl> i don't know how to find out the current versions, but debian states the package as Version: 20030312-1
14:22 < green> no, I mean, have not you changed the uml_tools with the kernel? because may be old uml_net installation had setuid bit and new does not
14:22 < avoozl> oh no i did not
14:23 < avoozl> i only built a new kernel
14:23 < green> hm, then this is strange
14:23 < avoozl> i'll try the older kernel again
14:24 < avoozl> yup old kernel works
14:24 < avoozl> maybe i overlooked some configuration options, but i think i enabled everything that was needed for tuntap
14:24 < avoozl> http://130.161.36.92/avoozl/.config
14:27 < avoozl> could this be related to the 'host filesystem' option?
14:28 < avoozl> hm no that's not it
14:29 < avoozl> off for a shower, bbl
14:31 -!- radical [~kernet@202.88.184.39] has joined #uml
14:44 < snide> green: do u know why tap0 is hanging ?
14:46 < snide> btw, how can i know if it's the host that hangs tap0 or the uml ?
14:46 < snide> since, it happens w/ uml_switch and uml-tuntap
14:48 < green> snide: what do you mean by "hanging"?
14:49 < snide> green: hanging as doing ' for(;;){}; ' ;-)
14:49 < snide> green: while(1) { } would be more understandable
14:49  * green barely can imagine interface doing 'while(1) { }'
14:50 < snide> green: i don't know if it's the interface, but actually, it's the userspace process that munches 100% cpu, not the host kernel
14:51 < green> so the problem is in process.
14:52 < snide> green: yup... but the thing is, it seems to be independent of the type of eth0 (tuntap & daemon are both hanging)
14:53 < snide> green: and it's not the uml_switch that hangs, it's ./linux
14:54 < snide> and, the worst part is that both 2.5.69-1 & 2.4.20-4 are 'infected' ;-(
14:54 < green> hm.
14:54  * snide starts to think that his 'host' is not so clean ;-)
14:54 < green> well, attache to it with gdb and look who is doing bad thing
14:54 < green> and why
14:55 < green> also check that your kernel headers with which you compile your uml match to the host kernel
14:55 < snide> green: i do have a call stack (from 'sysrq p'), but i didn't manage to use it, since 2.5 doesn't support modules
14:56 < snide> green: kernel matching ? what if host is 2.4 and uml is 2.5... what do u mean ?
14:56 < green> no
14:56 < green> I mean that if you have host kernel version 2.4.19 and /usr/include/linux from 2.4.3, that might be a problem
14:57 < snide> green: ah... no, /usr/include/linux is a symlink
14:57  * green does not exactly remember when the tuntap thing was changed in 2.4 time
14:57 < snide> hmmm... *checking*
14:57 < green>  /usr/include/linux is not a symlink if you are using glibc.
14:58 < green> and it should not be
14:58 < snide> green: i made it a symlink... so, should i copy them ?
14:59 < green> no.
14:59 < snide> green: what's the prob about the symlink ?
14:59 < green> glibc itself expects that /usr/include/linux is the kernel headers from the kernel that glibc was compiled against
15:00 < green> well, but this is not relewant to your problem then
15:03 < snide> green: ah, no... it's not a symlink anymore, it's 2.4.18 [ host is 2.4.19 ]
15:04 < green> then it should be fine, I think
15:05 < snide> green: ;-)
15:32 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has quit [Quit: BitchX-1.0c19 -- just do it.]
15:50 < BB1> any of you Matthew Bloch ?
15:57 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has joined #uml
15:57 < snide> re
15:57 < BB1> re
15:58 < snide> btw, it seems that sf.net doesn't allow to Dthe last 
15:58 < snide> d/l the last uml_utils
15:58 < snide> ;-(
16:01 < snide> does any1 has the latest uml_utils
16:01 < BB1> are they on the mirrors?
16:01 < snide> seems not. [ may be, but i didn't manage to find them ]
16:01 < BB1> what ver are you looking for?
16:02 < BB1> 20030529 i can see
16:02 < snide> uml_utilities_20030529.tar.bz2 
16:02 < BB1> http://uk.mirror.camelbackup.com/uml/uml_utilities_20030529.tar.bz2
16:02 < snide> thx
16:02 < BB1> that is if my server can cope
16:03 < snide> why doesn't sf.et work ? [ yes, your works ;-pp ]
16:03 < BB1> i dunno, i never download from sf.net unless i have to
16:04 < snide> BB1: where do u have it from then ? ;-p
16:04 < BB1> how do you mean ?
16:05 < snide> where do u rsync from ? ;)
16:05 < BB1> i dont jdike puts it there :P
16:05 < snide> u'r not the owner of uk.mirror.camelbackup.com ?
16:05 < snide> BB1: oh ;p
16:06 -!- snide [~snide@AMontsouris-108-1-4-210.w193-252.abo.wanadoo.fr] has quit [Quit: [BX] Homer Simpson uses BitchX. D'OH! D'OH! D'OH!]
16:08 < avoozl> re
16:09  * avoozl will have another go at finding the difference between these two uml kernels
16:22 < avoozl> somehow, networking works good for the stock debian kernel, but for my newly compiled kernel is only works if i run it as root, not if i run it as a user.. very odd.. rights on the tap device and rights on uml_net are right.. are there any options that i could have missed?
16:27 < caker> chmod 777 /dev/net/tun
16:27 < Getty> someone having network problem with uml?
16:27 < Getty> (joined bounced just a 10 seconds ago)
16:30 < avoozl> caker: the rights are set correctly
16:31 < caker> you're using uml_net aren't you?
16:31 < avoozl> not to setup the stuff
16:31 < avoozl> i do that by hand
16:31 < caker> good :-)
16:31 < avoozl> i dont like the script
16:31 < avoozl> but it does work with one kernel, both as a user and as root
16:31 < avoozl> i just can't get my new compiled kernel to work as a user, it only works with networking as root
16:32 < caker> how are you creating the tap device?
16:32 < avoozl> tunctl -u timo -t tap0
16:32 < avoozl> ifconfig tap0 192.168.11.1 netmask 255.255.255.0 up
16:32 < avoozl> where timo is the user that will run it
16:32 < caker> i mean create the actual device
16:32 < avoozl> i've also set the user/grp manually
16:32 < caker> that doesnt matter
16:32 < caker> you need to tunctl username tapdevice
16:32 < avoozl> crw-rw----    1 timo     timo      10, 200 Jun  3 13:15 /dev/net/tun
16:33 < avoozl> but this tun stuff wouldn't explain that it works with one kernel and not with the other right?
16:34 < caker> UML kernel or host kernel?
16:34 < avoozl> uml kernel
16:34 < avoozl> i haven't touched the host kernel yet
16:34 < caker> and you compiled in tuntap?
16:34 < avoozl> yes
16:34 < avoozl> http://130.161.36.92/avoozl/.config
16:34 < caker> do you get an error message when you ifconfig inside uml?
16:34 < avoozl> let me paste that
16:35 < avoozl> Configuring network interfaces: tuntap_open_tramp : didn't receive a message
16:35 < avoozl> setreuid to root failed : : Operation not permitted
16:35 < avoozl> tuntap_open_tramp failed - errno = 22
16:35 < avoozl> SIOCSIFFLAGS: Invalid argument
16:35 < caker> and how are you launching ./linux (specifically the eth0 arg)
16:35 < avoozl> DISPLAY=:0.1 ./linux eth0=tuntap,,,192.168.11.1
16:35 < avoozl> the same for both uml kernels
16:35 < caker> since you specify the IP on the command line, it attempts to run uml_net
16:36 < avoozl> owww
16:36 < caker> and uml_net prob is set SUID
16:36 < caker> rather, is NOT set SUID
16:36 < avoozl> indeed it isnt
16:36 < caker> and when it runs as normal user, it fails to execute ifconfig/routes,etc
16:36 < caker> on the host
16:36 < avoozl> hm
16:36 < avoozl> still how does that explain that it works with one uml kernel and not with the other?
16:37  * caker shrugs
16:37 < caker> perhaps you have more than one uml_net ?
16:37 < caker> or are launching linux differently
16:37 < caker> Jeff does tricky things like searching paths so maybe it's finding the wrong (non-suid) one
16:38 < avoozl> god that is it
16:38 < avoozl> there were two uml_net's
16:38 < avoozl> one with a g+s
16:38 < avoozl> the other without
16:38 < caker> I'll send you a bill :-p
16:38 < avoozl> thanks again
16:39 < caker> hehe no sweat! go fix :-)
16:40 < avoozl> after this i can reboot my newly build host kernel and see if that skas thing works as well :)
16:40 < avoozl> yup making uml_net suid helped
16:57 < avoozl> is there any way to test wether skas works? any specific message i should pay attention to?
16:58 < avoozl> Checking for the skas3 patch in the host...found
16:58 < avoozl> Checking for /proc/mm...found
16:58 < avoozl> is that enough? :)
16:58 < BB1> yes
16:59 < avoozl> cool :)
17:10 -!- shanti [~shanti@157.182.194.85] has quit [Quit: Client Exiting]
17:18 -!- avoozl is now known as avoozzzl
18:12 -!- eddie-moz [~knoppix@157-188.dsl.scc.net] has quit [Remote host closed the connection]
18:24 < avoozzzl> one more small question, if i do a reboot of the uml client system, the shutdown goes ok, the system begins to restart, but then it hangs at:
18:24 < avoozzzl> Initializing stdio console driver
18:24 < avoozzzl> Netdevice 0 
18:25 < avoozzzl> at this point i kill linux and restart it by hand, and then it works fine
18:25 < caker> avoozzzl try appending con=null con0=fd:0,fd:1 to your ./linux command line
18:26 < avoozzzl> i will
18:30 < avoozzzl> caker: still the same
18:50 < avoozzzl> oh well, thanks for all the help, great channel here. goodnite
18:50 -!- avoozzzl [~jrk@vengeance.et.tudelft.nl] has left #uml []
19:38 -!- rainstaiN [mistik1@lame.isps.cant.prevent.selfserve.us] has quit [Ping timeout: 492 seconds]
19:53 -!- radical [~kernet@202.88.184.39] has quit [Quit: Client Exiting]
20:56 < mistik1> david: you about
20:57 < mistik1> david or caker please take a look here 
20:57 < mistik1> bhttp://pastebot.geeksinthehood.net:8888/1
20:57 < mistik1> http://pastebot.geeksinthehood.net:8888/1
20:58 < david> hello
20:59 < mistik1> hey man
20:59 < mistik1> you have inspired me :)
20:59 < mistik1> please take a look at that and tell me what you think
21:03 < caker> mistik1: pretty cool!
21:03 < mistik1> thanks man
21:04 < mistik1> If you guys have any concepts at all you would like to add beat me up please
21:04 < mistik1> mistik1 at ltsp dot org
21:04 < caker> mistik1: did you get the email i sent today?
21:05 < mistik1> i've not checked since I came in
21:05 < mistik1> yep. its here
21:05 < caker> mistik1: cool
21:06 < mistik1> I'll be checking that out
21:07 < caker> thanks
21:07 < mistik1> your stuff is awesome to say the least
21:07 < caker> wow thanks
21:07 < caker> Been working on it for a long time, so I'm anxious to see how well it does
21:08 < mistik1> I'm sure you'll be fine once everything is proven stable
21:08 < mistik1> its a great concept, one actually that I myself have been after for some time.
21:08 < mistik1> not sure why I didn't get around to UML earlier
21:09 < caker> UML is the shit
21:09 < mistik1> indeed
21:15 < david> mistik1: looks good to me :-)
21:17 < mistik1> hehe
21:17 < mistik1> cool
21:54 -!- rainstaiN [mistik1@guk.really.does.ann0y.us] has joined #uml
23:09 -!- ASY- [~someone@216.221.51.54] has joined #uml
23:09 -!- ASY [~someone@216.221.51.54] has quit [Read error: Connection reset by peer]
23:09 -!- ASY- is now known as ASY
--- Log closed Wed Jun 04 00:00:01 2003