T2 IRC Log: 2007-08-26

This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.

« prev | next »

--- Log opened Sun Aug 26 00:00:11 2007
00:16 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
00:16 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
01:35 -!- sepp_ [n=sepp@Z6ddc.z.pppool.de] has joined #t2
01:55 -!- sepp [n=sepp@Z73a2.z.pppool.de] has quit [Read error: 110 (Connection timed out)]
02:41 < happycube> i'm having locale problems with alsa-utils
02:42 < sepp_> locale?
02:52 < happycube> yeah - the t-ru.gmo and t-jp.gmo files vanish
02:52 < happycube> or something - the're not found and it failed
02:52 < happycube> s
03:06 < sepp_> hmmm
03:10 < happycube> yeah it's moving a file around actually
03:24 < happycube> ok it works if you add --disable-nls
03:25 < happycube> i don't know how to submit that as a patch tho
03:27 < sepp_> maybe the strange behaviour should be fixed
03:27 < sepp_> and not just disable nls
03:28 < sepp_> in case someone wants nls support ...
03:43 -!- sepp [n=sepp@Z5410.z.pppool.de] has joined #t2
03:58 -!- sepp_ [n=sepp@Z6ddc.z.pppool.de] has quit [Read error: 110 (Connection timed out)]
05:48 -!- netson99 [n=netson99@210.210.181.128] has joined #t2
06:44 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
06:44 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
06:59 -!- netson99 [n=netson99@210.210.181.128] has left #t2 []
07:00 -!- netson99 [n=netson99@202.152.240.204] has joined #t2
07:07 -!- netson99_ [n=netson99@210.210.181.128] has joined #t2
07:08 -!- netson99_ [n=netson99@210.210.181.128] has left #t2 []
07:08 < happycube> *nod*
07:31 -!- netson99 [n=netson99@202.152.240.204] has quit [Read error: 110 (Connection timed out)]
07:31 -!- netson99_ [n=netson99@210.210.181.128] has joined #t2
07:32 -!- netson99_ is now known as netson99
07:51 -!- netson99_ [n=netson99@210.210.181.128] has joined #t2
07:52 -!- netson99_ [n=netson99@210.210.181.128] has left #t2 []
07:52 -!- netson99 [n=netson99@210.210.181.128] has quit [Read error: 104 (Connection reset by peer)]
09:25 < rxr> re
10:15 -!- mqueiros_ [n=mqueiros@c-217-70-65-203.bragatel.pt] has quit ["Ex-Chat"]
10:41 < CIA-24> susan * r25770 /trunk/package/network/openldap/openldap.desc: * updated openldap (2.3.37 -> 2.3.38)
10:41 < CIA-24> susan * r25771 /trunk/package/security/openct/openct.desc: * updated openct (0.6.12 -> 0.6.13)
10:41 < CIA-24> susan * r25772 /trunk/package/doc/howtos/howtos.desc: * updated howtos (20070818 -> 20070825)
10:41 < CIA-24> susan * r25773 /trunk/package/network/asterisk/asterisk.desc: * updated asterisk (1.4.10.1 -> 1.4.11)
11:20 -!- Cookie [n=Cookie@AToulouse-152-1-68-241.w83-200.abo.wanadoo.fr] has joined #t2
11:20 < Cookie> moin moin
11:21 < rxr> moin moin
11:21 [Users #t2]
11:21 [@ChanServ] [ cyda ] [ happycube] [ mtr ] [ rxr ] [ valentin]
11:21 [ CIA-24 ] [ emte ] [ kic- ] [ pdinoto_ ] [ sepp ]
11:21 [ Cookie ] [ Fatal] [ LMJ ] [ Ragnar|away] [ TobiX]
11:21 -!- Irssi: #t2: Total of 16 nicks [1 ops, 0 halfops, 0 voices, 15 normal]
11:22 < Cookie> <-- LMJ from his girlfriend laptop
11:22 < Cookie> moin rxr
11:22 < rxr> hehehe
11:22 < rxr> greetings .-)
11:23 < Cookie> rxr, I've started a very small OpenSource project kinda look ridiculous, but anyway, i'm looking for a serious bash coder for a couple hours. I haven't found any serious guy on sourceforge... Do you have any ideas where I could find a dude ?
11:24 < rxr> ouh
11:24 < rxr> no idea
11:24 < rxr> maybe some student in your area
11:25 < rxr> usually you fine people to contribute that stuble over your website and such
11:25 < Cookie> nothing :-/
11:25 < rxr> Cookie: ware do you want to start ?
11:25 < Cookie> I'm working on a font manager for linux
11:26 < Cookie> trying to validate the concept with a bash script, it could works
11:27 < Cookie> I mean, it works
11:27 < Cookie> ;)
11:27 < rxr> hm
11:28 < Cookie> Do you know what is the goal of a font management tool for a graphic artist ?
11:28 < rxr> probably to get an overview of the avilable fonts or so ?
11:29 < Cookie> activate only fonts you need when you work for a customer and avoid 100000 fonts at the same time
11:30 < Cookie> if you work for Mr X, you activate the X fonts who buy the customer, make a logo or so and when you are done, you diseable the fonts
11:31 < Cookie> this is used by a lot of print guy
11:31 < Cookie> s*
11:31 < Cookie> but there is not font manager on linux :-/
11:31 < Cookie> I think it's a mistake
11:32 < rxr> yeah - a good font manager would probably make that part way easier
11:32 -!- tri [n=tri@p57ADBFBB.dip0.t-ipconnect.de] has joined #t2
11:33 < Cookie> If you find a bash coder around with free time, you know what to do ;)
11:35 < rxr> I keep in mind
11:35 < rxr> but, are you really sure some shell code can make a good font manager ?
11:35 < tri> moin
11:35 < rxr> hi tri
11:36 < rxr> oh my god: http://www.heise.de/newsticker/meldung/94714
11:38 < Cookie> moin tri
11:38 < Cookie> to validate the basic function : yes, but usually, graphic artists need something more eyes candy later
11:39 < Cookie> I'm thinking PyGTK or so
11:41 < Cookie> here am i : http://tymproject.free.fr/
12:07 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
12:08 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
12:24 -!- tri [n=tri@p57ADBFBB.dip0.t-ipconnect.de] has quit [Remote closed the connection]
12:36 -!- Cookie [n=Cookie@AToulouse-152-1-68-241.w83-200.abo.wanadoo.fr] has quit ["Leaving"]
12:46 -!- tri [n=tri@p57ADBFBB.dip0.t-ipconnect.de] has joined #t2
12:50 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 104 (Connection reset by peer)]
12:50 -!- pdinoto_ [n=pdinoto@200.117.246.103] has joined #t2
12:51 < tri> re
13:18 -!- mqueiros_ [n=mqueiros@c-217-70-65-203.bragatel.pt] has joined #t2
13:42 -!- itsu [n=user@i577BB338.versanet.de] has joined #t2
13:47 < itsu> can anyoune help me on a packe build question ?
14:07 < sepp> yup
14:07 < sepp> just ask
14:07 < rxr> re
14:07 < tri> wb rxr
14:07 < tri> is ha-linux support in t2 ?
14:08 < itsu> how do i build a package without replacing the one on the build host
14:08 < itsu> i just rebuild htop with build-pkg an my host os htop got screwed .
14:09 < itsu> i read the manual but the warning for the package build was right
14:09 < itsu> so how do i accomplish that
14:09 < tri> i'm not sure if i got it right but you can remove a package with mine -r
14:09 < itsu> nono
14:10 < tri> and then do a emerge-pkg an build htop again
14:10 < itsu> i was have build a t2 target on another os
14:10 < itsu> then i saw i forgot some packages , wich are not in the dest-iso
14:10 < itsu> so i would like to rebuild only a single package
14:11 < sepp> you can build packages in the sandbox (for the sandbox) or build them for (into) your running system
14:11 < tri> with emerge-pkg -cfg "target" ?
14:11 < itsu> how do i build a package in the sandbox ?
14:11 < itsu> no i used Build-Pkg -cfg config htop
14:11 < itsu> sorry.
14:12 < itsu> ./scripts/Build-Pkg -cfg default htop
14:12 < sepp> that will put it into your running system
14:12 < itsu> and then my host htop got replaced by the built one
14:12 < itsu> i want to avoid exactly that
14:12 < itsu> so i need 2 use emerge-pkg ?
14:13 < sepp> no
14:13 < sepp> you can add things to your target and re-run build-target...
14:13 < sepp> it will not start from 0 ;)
14:14 < itsu> so how do i build in the sandbox ?
14:15 < tri> do Build-Target
14:15 < itsu> but build-target rebuilds only the whole target
14:15 < sepp> Build-Target -cfg foo -job 5-htop
14:15 < itsu> ahh...
14:16 < tri> hm doesn't build-target only build "new / changed" packages and not the whole target again ?
14:17 < sepp> if you add htop and restart build-target it will do what you want
14:17 < itsu> copy that
14:18 < sepp> emerge-pkg is for the running system only
14:18 < itsu> in wich path do i find the added package afterwards
14:18 < sepp> the tarball ?
14:18 < itsu> jep
14:19 < sepp> build/foo*/TOOLCHAIN/packages
14:19 < sepp> or so
14:19 < itsu> cool
14:19 < itsu> thanx 4 support dudes :-)
14:19 < tri> can i do that: "build-target -cfg foo -job 5-htop" also with build-pkg ?
14:20 < sepp> no
14:20 < sepp> i think not
14:20 < tri> why do i have the -cfg option for emerge-pkg if it is only for the running system ?
14:21 < sepp> because you can have several configs
14:21 < rxr> you can still have several configs and use "my-curreng-gnome" system as config :-)
14:22 < tri> ah ok make sence
14:22 < tri> thx :-)
14:24 [Users #t2]
14:24 [@ChanServ] [ emte ] [ itsu] [ mqueiros_] [ Ragnar|away] [ TobiX ]
14:24 [ CIA-24 ] [ Fatal ] [ kic-] [ mtr ] [ rxr ] [ tri ]
14:24 [ cyda ] [ happycube] [ LMJ ] [ pdinoto_ ] [ sepp ] [ valentin]
14:24 -!- Irssi: #t2: Total of 18 nicks [1 ops, 0 halfops, 0 voices, 17 normal]
14:24 < tri> is there http://linux-ha.org/ support for t2 ?
14:26 < sepp> not yet
14:27 < rxr> I think not yet, patches, as usual, welcome :-)
14:34 < itsu> @rxr i posted an issue on t2-devel regarding openvpn
14:43 < rxr> itsu: can you drop some more notes what is wrong ?
14:43 < itsu> lzo dependency is missing
14:43 < itsu> so openvpn does not build
14:43 < itsu> 2 solutions there
14:43 < rxr> hm - the cache lists lzo
14:43 < itsu> tunr openvpn standard config --disable-lzi
14:43 < itsu> or enable lzo
14:44 < rxr> what did you do? build a target or Emerge-Pkg ?
14:44 < itsu> sorry i forgot to mention that i edited the config file selecting openvpn manually
14:44 < itsu> so the lzo package was not enabled
14:45 < itsu> i couldn't find how to select a package manually in the lua-config menu
14:45 < itsu> so maybe you could help me with this
14:47 < rxr> wel - the addon package selection in T2 is more ment for "you know what you do" - so if you just add openvpn then you just get it ...
14:47 < rxr> but if openvpn can be configured to build without lzo it should be automatically disabled
14:47 < itsu> nope
14:47 < itsu> because then then the udp transport is deactivated
14:47 < itsu> thats not desired as for vpn tunnel udp is more efficient
14:48 < itsu> and you need this if u setup a openvpn server with more than just ip<-->ip tunnel
14:48 < itsu> dhcp under openvpn requires udp transport
14:49 < itsu> but the lzo package is needed also with other progis i can't remember right now
14:49 < itsu> anyways would be cool
14:49 < rxr> well, in T2 the package selection controls the basic feature set
14:50 < rxr> so just add lzo ...
14:50 < itsu> i did of course !
14:50 < rxr> but we should change openvpn to force disabling of that feature so that openvpn builds if lzo is not wanted
14:50 < itsu> i just couldn't find out how to select a single package in the lua menu
14:51 < rxr> I would not name it lua menu
14:51 < itsu> sorry
14:51 < rxr> it just contains some lua internals to make it a little faster
14:51 < rxr> (it was even slower before)
14:51 < rxr> you enable expert option, and the custom package selectino
14:51 < itsu> u are right of course
14:51 < rxr> and then the rules are the same in in the pkgsel files for target:
14:51 < rxr> X lzo
14:51 < rxr> X openvpn
14:51 < itsu> oh...
14:51 < itsu> i missed that one
14:52 < itsu> ;-)
14:52 < itsu> have u got another minute for a question
14:52 < itsu> ?
14:52 < rxr> sure
14:52 < rxr> just always ask :-)
14:53 [Users #t2]
14:53 [@ChanServ] [ emte ] [ itsu] [ mqueiros_] [ Ragnar|away] [ TobiX ]
14:53 [ CIA-24 ] [ Fatal ] [ kic-] [ mtr ] [ rxr ] [ tri ]
14:53 [ cyda ] [ happycube] [ LMJ ] [ pdinoto_ ] [ sepp ] [ valentin]
14:53 -!- Irssi: #t2: Total of 18 nicks [1 ops, 0 halfops, 0 voices, 17 normal]
14:54 < itsu> a while ago we talked about a thinclient target involving a initram only
14:54 < itsu> i could not figure out how to accomplish this
14:54 < itsu> maybe you could write down a small howto
14:54 < itsu> since i use t2 for many purposes
14:55 < itsu> usb-stick boot works fine , only thing i am missing is a pxe target
14:55 < itsu> which just needs a initram-fs only
14:57 < CIA-24> rene * r25774 /trunk/package/network/openvpn/openvpn.conf: * fixed openvpn to disable lzo support if not lzo is not installed
14:57 < rxr> itsu: either you need to write some lines of code in the build.sh of your target, to extend the cpio of the ininitramfs
14:58 < rxr> or we need to add a new generic, "initramfs" output image option
14:58 < rxr> the later might indeed be pretty useful for a lot other people
14:58 < itsu> i guess i am not skilled enough for the first option
14:58 < rxr> itsu: in how many days do you need it ?
14:58 < itsu> it's not urgent .
14:59 < itsu> maybe somewhere in september/october ?
14:59 < itsu> it would help me boost t2's popularity a lot !!!
14:59 < rxr> ok, added to my t2 TODO basket (http://www.t2-project.org/packages/basket.html)
14:59 < itsu> since i personally love t2
14:59 < itsu> cool....
14:59 < rxr> I try to do that ASAP
15:00 < itsu> no hurry rxr,
15:00 < itsu> greetz fly out 2 berlin :-)
15:00 < itsu> cya around guys
15:00 < rxr> if you want to try it yourself you can take a look at the livecd output, it already uses some T2 expand_initrd function to inject more files to the initrd
15:00 < itsu> okay i'll give it a try - definetly
15:00 < rxr> itsu: cu - greetings back :-)
15:00 < itsu> quit
15:01 -!- itsu [n=user@i577BB338.versanet.de] has left #t2 []
15:02 < CIA-24> rene * r25775 /trunk/package/graphic/gimp/gimp.desc: * updated gimp (2.2.17 -> 2.4.0-rc1)
15:02 < mqueiros_> rxr: I already had thought about adding the custom package selection to the "newbie guide" or the "one week with T2"...
15:33 -!- write_erase [n=olivier@vaio.aixmarseille.com] has joined #t2
15:39 -!- itsu [n=user@i577BB338.versanet.de] has joined #t2
16:17 -!- pdinoto_ [n=pdinoto@200.117.246.103] has quit [Read error: 113 (No route to host)]
16:22 < CIA-24> rene * r25776 /trunk/package/multimedia/dvbd/dvbd.desc:
16:22 < CIA-24> * added a note about the config file name of dvbd as I forget it
16:22 < CIA-24> all the time and need to unpack the source again and again :-)
16:28 -!- andr011 [n=root@bzq-219-167-68.static.bezeqint.net] has joined #t2
16:28 -!- andr011 [n=root@bzq-219-167-68.static.bezeqint.net] has left #t2 []
16:37 -!- itsu [n=user@i577BB338.versanet.de] has left #t2 []
16:44 < tri> http://www.rasterman.com/ :-)
16:44 < tri> would be nice for my t2 workstation
17:00 -!- tri [n=tri@p57ADBFBB.dip0.t-ipconnect.de] has left #t2 []
17:06 < CIA-24> rene * r25777 /trunk/target/ (rtbt/ t2rtbt/): * renamed t2rtbt to just rtbt
17:20 -!- mqueiros_ [n=mqueiros@c-217-70-65-203.bragatel.pt] has quit [Connection reset by peer]
17:22 -!- appletizer [n=oreos@77-99-46-114.cable.ubr04.hawk.blueyonder.co.uk] has joined #t2
17:22 < appletizer> mm...
17:22 < appletizer> has anybody seen jsaw?
17:23 < sepp> no
17:23 < appletizer> thanks sepp
17:23 < sepp> not seen for months
17:28 -!- itsu [n=user@i577BB338.versanet.de] has joined #t2
17:28 < itsu> need some more help
17:28 < itsu> i am having a clean svn tree of t2
17:28 -!- mqueiros_ [n=mqueiros@c-217-70-65-203.bragatel.pt] has joined #t2
17:28 < itsu> and a pre-selected config
17:28 -!- appletizer [n=oreos@77-99-46-114.cable.ubr04.hawk.blueyonder.co.uk] has left #t2 []
17:29 < itsu> if i want 2 build a single package with Build-Targe -cfg default -job 5-example
17:30 < itsu> what steps are required prior to that for the toolchain ?
17:30 < rxr> re
17:31 < itsu> @rxr ? re
17:31 < rxr> what exactly do you want to build ?
17:31 < rxr> if you build a full target system just run Build-Target without any -job specifier
17:31 < itsu> unfortunately i did cleanup -full
17:32 < itsu> so do i have to do a full rebuild first ?
17:32 < rxr> yes, probably - as any major thing will needs a dozent of dependencies
17:32 < rxr> Build-Target as to bootstrap the full thing again
17:33 < itsu> so when i update the svn tree is there the same problem
17:33 < itsu> afk 4 30min guys - need to hurry
17:37 < mqueiros_> itsu: I believe that if you update the source tree when doing a Buil-Target only the updated packages will be rebuilted.
17:41 < rxr> itsu: update is no problem
17:42 < rxr> you can selectively re-schedule to build new stuff with Create-ErrList -newremove
17:42 < rxr> and a Build-Target, it will then only build new stuff
17:42 < rxr> unless you Clean you do not commonly rebuild everything
17:45 [Users #t2]
17:45 [@ChanServ] [ emte ] [ itsu] [ mqueiros_ ] [ rxr ] [ valentin ]
17:45 [ CIA-24 ] [ Fatal ] [ kic-] [ mtr ] [ sepp ] [ write_erase]
17:45 [ cyda ] [ happycube] [ LMJ ] [ Ragnar|away] [ TobiX]
17:45 -!- Irssi: #t2: Total of 17 nicks [1 ops, 0 halfops, 0 voices, 16 normal]
17:49 < rxr> itsu: I started the inird stuff locally
17:49 < rxr> maybe I have something to commit later tonight
17:49 < rxr> as soon as the kernel was built I sleep my notebook and go home to continue there
17:53 -!- DeeKey [n=denis@84-50-21-182-dsl.kjj.estpak.ee] has joined #t2
17:53 < DeeKey> hi!
17:54 -!- itsu [n=user@i577BB338.versanet.de] has quit [Read error: 110 (Connection timed out)]
17:57 < CIA-24> denisk * r25778 /branches/rtai/package/base/busybox/ (busybox.conf busybox.desc): busybox ->1.7.0
17:58 < CIA-24> denisk * r25779 /branches/rtai/package/base/busybox/ (busybox-1.6.1-adduser.patch busybox-1.6.1-init.patch): busybox: deleted old patches
18:02 < CIA-24> denisk * r25780 /branches/rtai/package/contrib/rtai/pkg_linux26_post.conf: rtai fixes
18:03 -!- its1 [n=user@i577B8F16.versanet.de] has joined #t2
18:03 < its1> back again
18:12 < its1> (itsu reincarnated): @rxr
18:12 < its1> sorry for leaving shortly
18:13 < its1> so what about updating svn and rebuilding packages ?
18:13 < its1> do i have to rebuild the whole target after a svn update ?
18:13 < mqueiros_> itsl: no you don't.
18:13 < its1> how do i rebuild the new packages ?
18:13 < mqueiros_> only the updated packages will be build.
18:14 < its1> cool.
18:14 < mqueiros_> itsl: if you re-run Build-Target it will automatically check for updated packages...
18:58 -!- its1 [n=user@i577B8F16.versanet.de] has left #t2 []
19:02 -!- mqueiros_ [n=mqueiros@c-217-70-65-203.bragatel.pt] has quit [Read error: 104 (Connection reset by peer)]
19:03 -!- mqueiros__ [n=mqueiros@c-217-70-65-203.bragatel.pt] has joined #t2
20:19 -!- itsu [n=user@i577B8F16.versanet.de] has joined #t2
20:19 < itsu> does anyno know if initramfs support is default enabled in t2 ?
20:22 < DeeKey> u mean in kernel?
20:24 < itsu> exact-ly
20:26 < DeeKey> i think you can add this optin if you need it
20:26 < DeeKey> create file in your target build directory named
20:27 < DeeKey> linux.config and add line
20:27 < DeeKey> X option ---> to enable the option
20:27 < DeeKey> O option to disable
20:27 < DeeKey> example:
20:27 < DeeKey> X initramfs
20:30 < itsu> just digged the default kernel config file
20:30 < itsu> initramfs is already enabled
20:30 < DeeKey> even so !
20:30 < itsu> what do you mean : even so ?
20:30 < DeeKey> great, but i dont need it... I use my custom kernel config
20:31 < itsu> ah, okay didn't know
20:31 < itsu> thanx mate
20:31 < DeeKey> BTW what are you compiling?
20:31 < DeeKey> what target?
20:31 < itsu> im working on a initramfs for a pxe image
20:32 < itsu> when booting t2 via pxe initramfs is needed cause there is no hd to mount
20:33 < DeeKey> In my build I am doing the same things...
20:33 < DeeKey> I have an old discless PC and I boot it via PXE. My file system is mounted via NFS
20:34 < itsu> nfs is not an option in my case ...
20:34 < itsu> anyone having already a target / config for my purpose ?
20:35 < DeeKey> hmm.. You want to run your file system in ram?
20:36 < itsu> yes
20:36 < itsu> maybe i should consider squashfs instead
20:36 < itsu> cause its already in t2
20:36 < itsu> i think...
20:37 < DeeKey> please leave a copy of your target in T2 ... I think later I could need such a target...
20:37 < itsu> yes i already talked to rene about it
20:37 < DeeKey> great!
20:38 < itsu> he is willing to make such a target for thinclients .
20:38 -!- sepp [n=sepp@Z5410.z.pppool.de] has quit ["leaving"]
20:38 < DeeKey> yeah! Routers also need this!
20:38 < itsu> the estimated timeframe will be somewhere semptember/october
20:38 < itsu> unless i get it working myself before
20:38 < itsu> but,
20:38 < itsu> puppylinux already has such an pxe image
20:39 < itsu> i heard
20:39 < DeeKey> there is WRT, freesco and other router projects which are run in RAM
20:39 < itsu> have to dig puppylinux website
20:39 < itsu> yes there are , but i'll stick with t2
20:39 < DeeKey> have fun!
20:39 < itsu> for many reasons
20:39 < itsu> yes mate i will
20:39 < itsu> thanx for chatting
20:40 -!- sepp [n=user@Z4066.z.pppool.de] has joined #t2
20:53 < Fatal> the projects listed on the SoC page, are all of those in progress? I'm mostly interested in the build as non root one
20:57 < LMJ> moin
20:58 < itsu> moin
20:58 < itsu> :-)
21:00 < sepp> :-)
21:01 < sepp> afaik no one works on non-root builds
21:09 < Ragnar|away> imho, building in a vmware or vserver is a lot easier approach then patching every package to build without root
21:15 < Fatal> I've already been bitten once when a package plays with stuff as root and break non t2 related files
21:18 < Ragnar|away> yea ... so build in a vserver
21:20 < Fatal> I've never been much for covering my hears and sing solutions
21:20 < Ragnar|away> hm?
21:21 < Fatal> guess it didn't survive the translation, a more english one would be to bury my head in the sand
21:22 < Ragnar|away> there will always be some packages that are poorly coded. there may even be packages that maliciously mess up stuff
21:24 < Ragnar|away> the latter will try to get around being build/installed as user by trying to exploit loopholes on the system.
21:24 < Fatal> we're going around in circles, you advocate wasting resources on emulation layers, I tend to do my work as user and forbid the other people working on the same project any root access to build images, give them root, they can mess stuff up w/o realizing it etc
21:24 < Ragnar|away> vserver isn't an 'emulation layer'
21:25 < Fatal> semantics
21:25 < Fatal> virtualization is bad aswell :)
21:25 < Ragnar|away> and that's your religion?:)
21:26 < Fatal> and vserver is yours
21:27 < Ragnar|away> not quite, I just use a dedicated build system
21:27 < Fatal> if you're happy with that workaround/protection I'm happy for you
21:27 < Fatal> same here, but I got idiot coworkers, which I sadly can't do anything about
21:28 < Ragnar|away> well then you sound like a good prospect to work on the build-as-non-root ;)
21:28 < Fatal> but as I haven't migrated our builds over to t2 yet, this is not a problem yet
21:28 < Fatal> which is why I asked if anyone knew about it
21:29 < Ragnar|away> I haven't messed with it, but the main problem I see is preserving permissions the package sets (which it can't do as non-root)
21:29 < Fatal> I would if I could, but doing this job under cygwin is way out of my interest, which rxr has said is an requirement
21:29 < Ragnar|away> o_o
21:29 < sepp> most of the things are shell scripts, they wont get 500% slower with vserver or something
21:30 < Ragnar|away> .oO(why would that be a requirement?)
21:30 < Fatal> I believe some customers of t2 build their stuff under cygwin, I didn't press the issue
21:30 < Ragnar|away> vserver runs nearly 100% cpu speed. it's not an emulation, it just modifies the kernel to implement VM boundaries
21:31 < Ragnar|away> *shivers at the cygwin*
21:31 < Fatal> for linux systems preserving permissions is trivial, the current scripts I use run everything under fakeroot and retain state even between runs
21:32 < Ragnar|away> fakeroot won
21:32 < Fatal> so porting t2 to a fakeroot, fakechroot etc is rather easy, I just wouldn't be able to make t2 work full out on supported platforms other than linux
21:32 < Ragnar|away> will likely conflict with the flist wrapper which is also LD_PRELOADED
21:35 < Ragnar|away> but if you merge them (since fakeroot seems to do almost the same like the flist wrapper anyway) that would be a quite feasible and easy to implement thing
21:37 < Ragnar|away> only problem would be statically linked files
21:38 < Fatal> why?
21:38 < Ragnar|away> LD_PRELOAD doesn't work on statically linked apps
21:44 -!- itsu [n=user@i577B8F16.versanet.de] has left #t2 []
23:42 < happycube> argh... xslt dosen't build because there's no libxml2, and the latter dosen't seem to cross compile
23:46 < Ragnar|away> >.<
23:47 < mqueiros__> happycube: Isn't libxml2 named just libxml in T2 ?
23:48 < mqueiros__> some time ago I had the same thought... that there was no libxml2. But after checking I got the impression that it was libxml or libxml1...
23:50 < happycube> that one dosen't build
23:50 < happycube> it has no idea about cross compiling :P
--- Log closed Mon Aug 27 00:00:12 2007