T2 IRC Log: 2005-11-08

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 Tue Nov 08 00:00:03 2005
--- Day changed Tue Nov 08 2005
00:00 < mnemoc> rxr: the reason of X LIBC not working is the forced difference between perl/ repo and perl package
00:07 -!- sparc-kly|Univ [n=mubex@66-50-123-218.prtc.net] has joined #t2
00:58 < CIA-12> amery * r13215 /trunk/target/generic/config.in: * improved generic/template to cache .awk files, instead of regenerate them on each loop
01:24 -!- sparc-kly|Univ [n=mubex@66-50-123-218.prtc.net] has quit ["Leaving"]
01:25 < CIA-12> amery * r13216 /trunk/target/generic/config.in: * changed generic to hide sorting prefix on templates, added on r13214. so this number is not considered on config file
01:25 < mnemoc> is 13216 too obscure?
01:25 -!- ojh [n=omer@71-34-239-173.eugn.qwest.net] has joined #t2
01:26 < mnemoc> wb ojh
01:26 < ojh> mnemoc: Hi
01:31 < ojh> I have a small ROCK-PLUG related issue. When I plug in a device that creates a device node in /dev/usb/tts/ the permissions are set to 600 so only root can access the port. ROCK-PLUG needs a script to set the permissions on USB TTY devices. In my cast its the cradle for my Palm Pilot.
01:33 < mnemoc> iirc that goes on /etc/modules.conf
01:34 < mnemoc> /etc/modprobe.conf
01:35 < madtux> o.o
01:56 < mnemoc> isn't a mode=666 there?
01:57 < madtux> u mean mode=mnemoc ?
01:58 < ojh> mnemoc: Odd thing is I don't see /etc/modprobe.conf on my ROCK 2.0.0 system . . .
02:00 < mnemoc> strace -vvfF modprobe foo :)
02:00 < mnemoc> see what file yours reads
02:02 < ojh> mnemoc: query_module("usbserial", QM_INFO, {address=0xe0c38000, size=18012, flags=MOD_RUNNING|MOD_USED_ONCE, usecount=0}, 16) = 0
02:03 < ojh> IC, now that reviewed what strace does . . .
02:46 -!- sparc-kly [n=mubex@64.237.248.53] has joined #t2
02:49 -!- madtux [i=manolo@pf0.hostarica.com] has quit ["Leaving"]
04:35 -!- mubex [n=mubex@64.237.248.53] has joined #t2
05:31 -!- ojh [n=omer@71-34-239-173.eugn.qwest.net] has left #t2 []
05:48 -!- mubex is now known as [sparc-kly]
06:18 -!- ideal [n=idealm_@222.66.46.90] has joined #t2
06:19 < ideal> Congratulations for the release.
07:10 -!- Netsplit herbert.freenode.net <-> irc.freenode.net quits: [sparc-kly], valentin, owl, morfoh, jsaw, SerWou, karasz, _Ragnar_, ideal
07:10 -!- Netsplit over, joins: ideal, [sparc-kly], morfoh, jsaw, karasz, SerWou, owl, _Ragnar_, valentin
07:22 -!- Netsplit herbert.freenode.net <-> irc.freenode.net quits: valentin, [sparc-kly], owl, jsaw, morfoh, SerWou, _Ragnar_, karasz, ideal
07:22 -!- Netsplit over, joins: ideal, [sparc-kly], morfoh, jsaw, karasz, SerWou, owl, _Ragnar_, valentin
07:32 -!- mtr_ [n=Michael@p54AFC38C.dip0.t-ipconnect.de] has joined #t2
07:48 -!- mtr [n=Michael@p54AF8A5B.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
09:30 < jsaw> re
09:31 < jsaw> mnemoc: /etc/rockplug/config/usb.usermap -> last column
09:45 < valentin> moin
09:54 < jsaw> hi valentin
09:55 < jsaw> symlink!
10:06 < valentin> hi jsaw
10:11 < valentin> quite short news on symlink
10:16 -!- geekitus [n=geekitus@83.179.133.149] has joined #t2
10:20 < jsaw> valentin: have a few secs?
10:22 < rxr> hi
10:22 < jsaw> g'morning rxr
10:23 < rxr> valentin: yes - I submitted a 2nd longer one - but after 4 days the picked the first news ...
10:23 < rxr> symlink.ch seems to be in low contributor / holidy mode or so ..
10:23 < rxr> hi ideal - thanks
10:23 < jsaw> I have a question concerning t2 configuration, esp. networking/wlan
10:24 < rxr> jsaw: yep ?
10:24 < jsaw> got interupted...
10:24 < jsaw> I wonder, if we could develop some kind of xml or whatever description,
10:24 < jsaw> that could also be used by knetconf
10:24 < valentin> for what ?
10:24 < jsaw> the problem:
10:25 < jsaw> while the wep interface is standartized, wpa settings are different for all wlan drivers...
10:26 < jsaw> the other possibility that comes to mind, is that we package a "translation" tool per wlan driver (some script...)
10:27 < jsaw> hmmm.... download/mirror/p/psi-0.9.3_zh.qm.cksum-err
10:27 < jsaw> actually download/mirror/p/psi-0.9.3_*.qm.cksum-err
10:28 < jsaw> any ideas for the WPA thing?
10:28 < rxr> well - just wait to get those interfaces streamlined ...
10:29 < jsaw> you think they'll gonna be streamlined any time soon?
10:29 < jsaw> (I think that's a _couple_ of years ahead)
10:32 < rxr> my hope and feeling is that wifi in vanilla kernel get's a lot of work and integration recently
10:32 < rxr> until then we could provide multiple keywords or autoprobing of the interface to use ...
10:32 -!- CIA-12 [n=CIA@70.85.31.102] has quit [Read error: 145 (Connection timed out)]
10:32 < rxr> I would not supply additional state information
10:36 < jsaw> ic
10:37 < jsaw> but that still requires some "translation" mechanism
10:38 < jsaw> I rather hesitate to put everything into "wireless.sh"...
10:40 < rxr> wpa.ssh
10:40 < rxr> -s
10:40 < valentin> jsaw: how would you use those xml records inside rocknet ? xml with bash ...
10:44 < jsaw> the current netconf for wpa is to use the explicit command used to set up the wlan card. That means, I would not use it with bash, only the user interface for setting it up uses the description.
10:45 < jsaw> I expect the person editting the netconf by hand to know what he's doing... :)
10:46 < jsaw> I think, especially with wlan, the user interaction is cruicial.
10:47 < jsaw> You cannot expect WLAN to be somewhat near static.
10:47 < jsaw> Ethernet is easy with dhcp.
10:47 < jsaw> WLAN sadly is not.
10:49 < rxr> also more and more organizations run some ipsec of w-lan ...
10:49 < rxr> we need config for that as well ...
10:49 < jsaw> I have wlan at home and at the institute. We have three APs, and different nets (one without password on a vpn). Which one do I choose? ....
10:49 < jsaw> I already have some kind of auto-detection.
10:50 < jsaw> But it's specific for my wlan adapter...
10:50 < rxr> valentin: can I call in 10m when I finished glueing the scanner ?
10:51 < jsaw> gotta do some more brain storming... hmmm
11:12 < rxr> ,-)
11:20 -!- CIA-9 [n=CIA@flapjack.navi.cx] has joined #t2
11:22 -!- morfoh [n=jeru@p54BEE60C.dip.t-dialin.net] has quit ["Lost terminal"]
11:35 -!- ideal [n=idealm_@222.66.46.90] has quit [Read error: 110 (Connection timed out)]
11:36 < jsaw> cu later
11:43 < rxr> cu jsaw
11:43 -!- CIA-9 [n=CIA@flapjack.navi.cx] has quit [Remote closed the connection]
11:44 -!- CIA-1 [n=CIA@flapjack.navi.cx] has joined #t2
11:46 -!- ideal [n=idealm_@58.33.53.101] has joined #t2
11:47 [Users #t2]
11:47 [ [sparc-kly]] [ geekitus] [ karasz] [ owl ] [ SerWou ]
11:47 [ _Ragnar_ ] [ ideal ] [ mnemoc] [ rxr ] [ sparc-kly]
11:47 [ CIA-1 ] [ jsaw ] [ mtr_ ] [ sepp_] [ valentin ]
11:47 -!- Irssi: #t2: Total of 15 nicks [0 ops, 0 halfops, 0 voices, 15 normal]
11:49 < karasz> moin all
11:49 < rxr> mon karasz
11:50 < karasz> hello rxr
11:50 < karasz> i was wondering if i shouyld save something from power?
11:53 < rxr> hm
11:53 < rxr> how long will it be around
11:53 < rxr> I could copy the binary archives back on my sever
11:54 < rxr> but - well - we do not really need them since when we have permanent powerpc64 boxes we can quickly rebuild them ...
12:03 -!- [sparc-kly] [n=mubex@64.237.248.53] has quit [Remote closed the connection]
12:08 < CIA-1> rene * r13219 /trunk/package/mail/sylpheed/sylpheed.desc: * updated sylpheed (2.0.3 -> 2.0.4)
12:09 -!- sepp [n=sepp@p213.54.207.14.tisdip.tiscali.de] has joined #t2
12:10 < CIA-1> rene * r13220 /trunk/package/x11/fox/fox.desc: * updated fox (1.4.20 -> 1.4.22)
12:10 < CIA-1> rene * r13218 /trunk/package/audio/xmms-jack/xmms-jack.desc: * updated xmms-jack (0.14 -> 0.15)
12:10 < CIA-1> rene * r13217 /trunk/package/develop/gambas/gambas.desc: * updated gambas (1.0.9 -> 1.0.12)
12:10 < karasz> it will be around til 10 november
12:11 < karasz> rxr do you have use for the file layout of a Debian bootable cd for openpower?
12:12 < karasz> since you said that the boot target needs tickering...
12:14 < CIA-1> rene * r13221 /trunk/package/perl/perl-unicode/perl-unicode.desc: * updated perl-unicode (2.07 -> 2.09)
12:14 < rxr> karasz: yes - that could help
12:14 < karasz> how should i do that?
12:14 < CIA-1> rene * r13222 /trunk/package/perl/perl-html-parser/perl-html-parser.desc: * updated perl-html-parser (3.45 -> 3.46)
12:15 < karasz> make a tar and post it on my web?
12:15 < karasz> or you can grabb it from openpower?
12:17 < rxr> what you prefer
12:17 < rxr> did you verify the CD works on your box ?
12:17 < rxr> I guess I can download, right?
12:17 < rxr> I could test on my rs6k as well
12:18 < rxr> since the T2 CD is only bootable with a tiny manual quirk on the IBM rs6k on my side
12:18 < rxr> is it a ppc or ppc64 cd ?
12:18 < karasz> ppc64
12:18 < karasz> loaded on power
12:18 < karasz> you can check it now
12:19 < CIA-1> rene * r13223 /trunk/target/ (desktop/config.in livecd/default.pkgsel): * added mesa-demos (glxinfo and glxgears) to the livecd and desktop target
12:19 < rxr> hm - if you have a cd writer and cd-r/rw handy we could indeed modify and test the 2.1 bootdisk on ppc64 if you have the spare time
12:21 < karasz> just a min talking on the phone :(
12:22 < rxr> hehe ,-) me too
12:26 < karasz> sadly i don't think that we can test it because the morons from IBM did't sent me the passes for the bios
12:26 < karasz> anyway
12:26 < karasz> i loaded the cd on power but i guess that you will find more usefull the iso, no?
12:27 < karasz> i am copying that to power too
12:29 < karasz> but i have some spare time :D
12:29 < karasz> and some CPU ticks to spare ...
12:29 < karasz> so if you wanna try something just let me know
12:32 -!- sepp_ [n=sepp@p213.54.212.163.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)]
12:35 < karasz> hehe hehe hehe
12:36 < karasz> i am honored....
12:36 < karasz> T2 2.1.0 "Charlotte" released 2005-11-04
12:36 < karasz> my bday ^^
12:36 < karasz> hehe
12:46 < rxr> your baby ?
12:46 < karasz> no, just me :)
12:46 < karasz> i mean the date not the year ofcourse
13:05 < owl> moin
13:07 -!- madtux [i=manolo@pf0.hostarica.com] has joined #t2
13:08 < owl> hi madtux
13:08 < madtux> Hi owlita
13:08 < owl> how are you?
13:09 < madtux> i'm good, yourself?
13:09 < owl> alive i guess. thx. did i meantion that i'm really loving the hi-fi system of my car? ;)
13:10 < madtux> u did not not, but now i know
13:10 < madtux> what kind of car have u?
13:10 < owl> madtux: a golf III joker (with an extra-hi-fi system)
13:11 < owl> you?
13:11 < madtux> audi tt
13:12 < owl> hehehe
13:12 < madtux> ;-)
13:12 < owl> my dad works at audi ;)
13:12 < madtux> really? thats cool
13:12 < owl> yeah. he's doing stuff with 3d-measurement or something...
13:12 < owl> (technical stuff at least)
13:13 < madtux> ask daddy if he can talk at his job for a sponsored "testing car"
13:13 < madtux> :)
13:13 < owl> *lol* test what? t2 in audi-board-computer? ;)
13:13 < madtux> that sounds like a plan
13:13 < madtux> :)
13:14 < madtux> just find an excuse to get an audi :)
13:14 < owl> hehe. i guess that will not work ;)
13:14 < madtux> u'll never know if u don't try :)
13:15 < owl> ;) well. he's just a worker, so... ;)
13:15 < madtux> good point
13:15 < owl> ;)
13:25 < karasz> rxr did you finished your phone call?
13:26 < rxr> yep
13:26 < rxr> and lunch in the meantime ,-)
13:26 < karasz> cool
13:27 < karasz> so what about the boot stuff...
13:30 < rxr> karasz: just leave me a bit more time - right now I work on some aspects of the livecd
13:30 < rxr> can I come back to you in let's say 1 or 2 hours ?
13:31 < karasz> ofcourse
13:31 < karasz> i am not pushy you know me by now ;)
13:46 -!- rxr_ [n=rene@e178143068.adsl.alicedsl.de] has joined #t2
13:46 -!- Topic for #t2: T2 | 2.1.0 RELEASED! | The System Development Environment (SDE) | http://www.t2-project.org/
13:46 -!- Topic set by mnemoc [] [Fri Nov 4 16:55:33 2005]
13:46 [Users #t2]
13:46 [ _Ragnar_] [ ideal ] [ madtux] [ owl ] [ sepp ] [ valentin]
13:46 [ CIA-1 ] [ jsaw ] [ mnemoc] [ rxr ] [ SerWou ]
13:46 [ geekitus] [ karasz] [ mtr_ ] [ rxr_] [ sparc-kly]
13:46 -!- Irssi: #t2: Total of 16 nicks [0 ops, 0 halfops, 0 voices, 16 normal]
13:46 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
13:46 -!- Irssi: Join to #t2 was synced in 11 secs
13:50 < karasz> reboot
13:51 -!- karasz [n=karasz@80.97.102.202] has quit ["I cna ytpe 300 wrods pre mniuet!!!"]
13:56 -!- karasz [n=karasz@80.97.102.202] has joined #t2
13:59 -!- rxr [n=rene@e178160012.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)]
14:20 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: valentin, owl, jsaw, SerWou, madtux, _Ragnar_, karasz, sepp, ideal
14:20 -!- Netsplit over, joins: karasz, madtux, sepp, ideal, jsaw, valentin, _Ragnar_, owl, SerWou
14:27 -!- sparc-kly[g3] [n=tux@66-50-123-218.prtc.net] has joined #t2
14:55 < rxr_> jsaw: mnemoc: anyone here ?
15:00 < CIA-1> sebastian * r13224 /trunk/package/multimedia/jahshaka/ (jahshaka.desc qmake.patch):
15:00 < CIA-1> * workaround for a qmake bug
15:00 < CIA-1> * marked jahshaka NOPARALLEL
15:02 < mnemoc> moin!
15:02 < madtux> mnemoc, !
15:03 < mnemoc> rxr_: did you saw what i did on generic/templates processor?
15:03 < mnemoc> madtux: !!
15:06 < mnemoc> rxr_: i hope to haven't been to obscure
15:10 -!- Baldzius [n=mx@85.206.100.182] has joined #t2
15:10 < Baldzius> moin
15:11 < karasz> moin Baldzius
15:11 < Baldzius> moin karasz
15:23 < jsaw> re
15:23 < rxr_> mnemoc: yes - but I did not fully review it
15:23 < jsaw> rxr_: ?
15:23 < rxr_> mnemoc: I hack the pkgsel processor up right
15:23 < rxr_> now
15:23 -!- You're now known as rxr
15:24 < rxr> for 2.2 I think we should obsolete old-style direct sed filtering
15:24 < rxr> and I wanted to convert the livecd to use $target/pkgsel
15:24 < rxr> though I have several issues with that
15:25 < rxr> how did you both intended those to be used?
15:25 < jsaw> rxr: rev?
15:25 < rxr> e.g. I think mnemoc prefers O * and then enable stuff
15:25 < rxr> however this does not work when you want sane default for architecture specific packages and glibc32 ...
15:26 < rxr> also X base/* enables all sort of fluff due to matching categories
15:26 < rxr> can we agree that the pattern should only match the repository/pkgname and that if we ever need to match other flags introduce other syntax for that ?
15:26 < jsaw> I'd think, removing an entry is fine if the basic strucuture of the target requires it
15:27 < rxr> and what are your proposals for sane defaults for architecture specific stuff and libcCs ?
15:28 < jsaw> Hmmm
15:28 < jsaw> You don't want to build uclibc if the target is glibc based, e.g.
15:29 < rxr> jsaw: nope - I want to select e.g. base/* with the defaults the pacakges have there ...
15:29 < jsaw> you mean, the defaults in $target/pkgsel ?
15:30 < jsaw> I do not yet see the problem...then... .oO(?)
15:31 < jsaw> run $target/pkgsel first, then the user selected stuff... ?
15:31 < rxr> tell me how to write a pkgsel that disables all but base/*
15:31 < rxr> or that enables base/* with it's defautls
15:34 < jsaw> the first one I have to check myself,
15:34 < jsaw> but the second one??? why would you want that?
15:35 < jsaw> if the $target/pkgsel is sane first, you don't need to reenable it...!?
15:36 < rxr> ?
15:36 < rxr> I talk about writing the target pkgsel ...
15:37 < jsaw> so, you mean, in case everything is disabled, then re-enable per repo wise?
15:40 < rxr> for example
15:42 < rxr> for now I restrict the pattern to the repo/pkg, ok ?
15:43 < jsaw> so only explict pkg addition and removal? no removal or enabling of complete repos?
15:45 < rxr> yes reps of course
15:45 < rxr> but currently the pattern matches the category as well
15:45 < rxr> which sucks at least for me
15:45 < jsaw> hmm, ic
15:46 -!- geekitus [n=geekitus@83.179.133.149] has quit [Read error: 110 (Connection timed out)]
15:47 < jsaw> okay with me for now. Later, however, I'd like to have exactly the opposite, of course, with a finer grained category.
15:48 < jsaw> btw,
15:48 < jsaw> if you're at it, change the generated rule (from pkgsel_parse) to sth. like this:
15:48 < rxr> it is already changed a lot ..
15:49 < jsaw> ( $5 ~ /base\/[^\/]*/ ) Â{ $1='X' }
15:49 < jsaw> then, you can also do
15:49 < jsaw> ( $5 !~ ...
15:49 < jsaw> which deselects everything but...
15:52 < rxr> can you paste the diff you intend ?
15:54 < jsaw> hmmm
15:54 < jsaw> do you give some hours to re-understand the logic of pkgsel_parse?
15:55 < jsaw> - address="" ; first="( " ; others="&& "
15:55 < jsaw> no, forget this
15:57 < jsaw> do we, at any time, need this complicated pkgsel_parse?
15:57 < jsaw> did anybody ever see some pattern with "&&" generated?
15:58 -!- sparc-kly[g3] [n=tux@66-50-123-218.prtc.net] has quit [Read error: 110 (Connection timed out)]
16:00 < rxr> jsaw: nope
16:00 < rxr> jsaw: I simplify them right now
16:00 < rxr> jsaw: if we ever want more complex stuff let's do it then
16:00 < jsaw> good :)
16:00 < rxr> jsaw: did you intended to allow negative matches?
16:00 < rxr> like O all bug base/* ?
16:01 < jsaw> you were the one that asked for it...don't you?
16:01 < rxr> yes
16:01 < rxr> but your sentence indicated you intended s.th. like that in the past ...
16:03 < jsaw> no, I just glanced at the awk manual...
16:04 < jsaw> mnemoc: opinions?
16:05 < jsaw> btw, there are new trac entries for 2.1.0-stable
16:06 < rxr> from whom ?
16:06 < rxr> btw. we need trac email notifications ...
16:07 < jsaw> two end-user
16:07 < jsaw> (that just downloaded 2.1.0-stable)
16:08 < jsaw> we should also check if some of the errors have been resolved...
16:09 < jsaw> gotta go, cu later
16:10 < rxr> cu
16:11 < rxr> jsaw: still here?
16:12 < rxr> I think I'll implement inverted matches
16:12 < rxr> O !base/* =
16:12 < rxr> ? even
16:12 < rxr> O !base/* ?
16:13 < karasz> rxr do you plan to be here in the evening too?
16:16 < karasz> i have to go now, but i will be back
16:17 < karasz> in the mean time rxr if you need something from power help yourself
16:18 < karasz> i put an iso in your home, it is a debian boot with some knoppix live cd for pSeries and OpenPower
16:18 < karasz> you can find the cd too in media
16:18 < karasz> cu, i will be back in aprox 1,5h
16:19 < rxr> yes - I'll be often around
16:36 -!- Baldzius [n=mx@85.206.100.182] has quit [Remote closed the connection]
17:01 -!- sepp_ [n=sepp@p213.54.218.177.tisdip.tiscali.de] has joined #t2
17:01 -!- sepp [n=sepp@p213.54.207.14.tisdip.tiscali.de] has quit [Read error: 104 (Connection reset by peer)]
17:11 -!- g2 [n=g2@cpe-066-057-008-035.nc.res.rr.com] has joined #t2
17:15 -!- g2 [n=g2@cpe-066-057-008-035.nc.res.rr.com] has quit [Client Quit]
17:20 < rxr> hey - after rewriting pkgsel_parse and tweaking and tweaking of the pkgsel of livecd I'm down to 25 differences to the previous sed list ..
17:36 < rxr> ok - well - now we have a 30 minute info talk about T2 on the linux world expo ,-)
17:47 -!- MisterN [n=MrN@p549CF8A1.dip.t-dialin.net] has joined #t2
17:48 < MisterN> hi
17:55 < rxr> hi
17:56 < MisterN> rxr: how far from good is the new livecd?
17:59 -!- CIA-1 [n=CIA@flapjack.navi.cx] has quit [Excess Flood]
18:00 -!- CIA-14 [n=CIA@flapjack.navi.cx] has joined #t2
18:00 -!- ideal [n=idealm_@58.33.53.101] has quit [Remote closed the connection]
18:04 < rxr> MisterN: not far ,-)
18:05 < MisterN> rxr: while my main reason is hotplug++, i want to try this thing *g*
18:21 < karasz> re moin
18:26 < rxr> puh - finally ...
18:26 < rxr> +X rc
18:26 < rxr> +X gtk+12
18:26 < rxr> +X gtk+
18:26 < rxr> +O libdvdnav
18:26 < rxr> +O autotrace
18:26 < rxr> +X bashcompletion
18:26 < rxr> the difference of old style, ugly package selection on the live cd and the new style pkgsel with a lot of pkgse_* function adaptions to make useful selection possible ...
18:31 < karasz> rxr i am back, if you need me whissle ;)
18:32 < rxr> well, ..
18:32 < rxr> karasz: I worked on the stupid pkgsel stuff for too long
18:32 < rxr> karasz: I fear I do not have too much time left today for powerpc64
18:32 < karasz> but it is turning out good isn't it?
18:32 < rxr> karasz: is tomorrow ok as well or should we work on it today ?
18:32 < rxr> karasz: yep
18:32 < karasz> tomorrow is ok
18:32 < rxr> though I hope mnemoc does not complain I rewrote the pkgsel_parse nearly completely ...
18:33 < rxr> karasz: ok - because I wanted to go to the swimming hall with susan today ..
18:33 < karasz> usualy mnemoc does not like that ;)
18:33 < karasz> have fun
18:33 < CIA-14> rene * r13225 /trunk/scripts/Build-Target: * fixed Build-Target fuser on log files
18:34 < karasz> morons from IBM lost the password :D
18:34 < karasz> hmmm go figure
18:36 < rxr> ?
18:37 < karasz> they lost the bios password
18:37 < CIA-14> rene * r13226 /trunk/scripts/config-functions.in:
18:37 < CIA-14> * rewrote most parts of pkgsel_parse: do only match on repository/package
18:37 < CIA-14> and not use categories. So base/* only matches on the base repository
18:37 < CIA-14> and not the many base/* categorized packages. Allow inverted matches
18:37 < CIA-14> via ! (O !base/*) and to get the packages' default state (e.g.
18:37 < CIA-14> depending on the architecture). In my opinon this makes pkgsel way
18:37 < CIA-14> more useful to use.
18:38 < CIA-14> rene * r13227 /trunk/target/livecd/ (default.pkgsel pkgsel):
18:38 < CIA-14> * converted the livecd target from old-style custom sed package
18:38 < CIA-14> filtering to pkgsel map
18:38 < karasz> i guess that in the end it turns out well cause today i have planned some db2 group tickering
18:38 < CIA-14> rene * r13228 /trunk/target/livecd/rootfs/sbin/init.d/system: * fixed livecd init.d/system overlay's #! ...
18:39 < karasz> so maybe it is even better tomorrow to fiddle with boot
18:39 < karasz> rxr i read some docu abt that debian and i didn't kiked it very much....
18:40 < karasz> http://karasz.pastebin.ca/28009
18:40 < karasz> ^^ doesn't sound too encouraging...
18:42 < karasz> guys and girls (owl ;) ) i am interested in a little chit-chat about kernels
18:42 < karasz> anyone inyterested?
18:45 < rxr> karasz: what details do you wanna chit-chat ?
18:48 < karasz> hmm basic stuff i guess
18:48 < CIA-14> rene * r13229 /trunk/target/livecd/config.in: * removed former explicit pkgfilter form the livecd/config.in
18:48 < CIA-14> rene * r13230 /trunk/target/livecd/build_image.sh:
18:48 < CIA-14> * added proof of concept filter to remove certain .a files from the
18:48 < CIA-14> livecd (space optimization)
18:48 < karasz> like making a monolithic kernel does present speed advantages..
18:48 < karasz> or not
18:48 < rxr> nope
18:50 < karasz> so it is just against loadable kernel module based attacks?
18:55 < rxr> yeah - maybe - but if you have root / CAP_... rights you can inject code in other ways - e.g. via /dev/mem ,-)
18:55 < mnemoc> rehi
18:55 < rxr> hi mnemoc
18:55 < rxr> I hope you like my livecd/pkgsel and the associated pkgsel_parse rewrite ...
18:55 < rxr> mail to the list in some seconds ...
18:56 < karasz> hi mnemoc
18:56 < karasz> and then rxr what are the advantages of monolith?
18:58 < _Ragnar_> karasz: most scriptkiddies won't know how to patch /dev/mem tho ... and if you compile a custom kernel, chance of something going wrong with that are higher, so it's still worth to take the easy way away from the scriptkiddies
18:58 < mnemoc> rxr: i really don't like 13226
18:58 < mnemoc> moin rxr, karasz, _Ragnar_
18:58 < rxr> mnemoc: why ?
18:58 < mnemoc> categories thing
18:59 < rxr> mnemoc: do you use it ?
18:59 < _Ragnar_> moin
18:59 < rxr> aside that our categories stink and I still wait for your proposal how to improve them ,-)
18:59 < mnemoc> rxr: not because because current categorization sucks
18:59 < mnemoc> rxr: not now because current categorization sucks
18:59 < mnemoc> but we have to solve that
19:00 < mnemoc> and pkgsel shall supoort them
19:00 < rxr> mnemoc: ok but I strongly suggest an alternative keyword
19:00 < mnemoc> no need
19:00 < rxr> since it will always suck when you want base/* and get random category selections mixed in
19:00 < rxr> s/keyword/syntax/ ...
19:00 < mnemoc> we have four things
19:00 < mnemoc> packages, repos, categories and flags
19:01 < rxr> e.g. [XO=-] category/....
19:01 < rxr> yes - just prefix what namespace you want when it is not repository/package ,-)
19:01 < rxr> I hope you find my = useful ...
19:01 < mnemoc> if token ends with / is a repo
19:01 < mnemoc> base/
19:01 < rxr> well - right now you need base/* though ...
19:02 < mnemoc> why * ??
19:02 < rxr> just so ...
19:02 < mnemoc> listen please :)
19:02 < rxr> one can use multimedia/lib*
19:02 < rxr> ok - I listen ,-)
19:03 < mnemoc> let's do things simple
19:03 < mnemoc> if token ens with / it's a repo
19:03 < mnemoc> if not, it can be anything except a repo
19:03 < mnemoc> with * expanding to [^ ]*
19:04 < mnemoc> if you want multimedia libs, lets do a category for them
19:04 < mnemoc> multimedia/* <--- all multimedia categories
19:04 < mnemoc> multimedia/libs <--- all multimedia libraries
19:04 < mnemoc> as categories
19:05 < mnemoc> desktop/* desktop/kde/* desktop/*/core
19:05 < mnemoc> categories
19:05 < mnemoc> perl* <--- package
19:05 < mnemoc> perl/ <--- repo
19:06 < rxr> I would not limit the selections to a repo only repo/lib* should be fine
19:06 < mnemoc> not all libs start with lib
19:06 < mnemoc> don't limit it
19:06 < rxr> I also would not match anything if it has no slash - imagine "rc" it might match a lot of stuff aside the "rc" shell you want
19:07 < rxr> mnemoc: if you want a category you can still write category/multimeda/lib if that is what you want
19:07 < mnemoc> ' $pattern '
19:07 < mnemoc> not '$patern'
19:07 < rxr> the inexactness of the pkgsel_parse of t-1h is exactly what made my afternoon a pain ...
19:08 < rxr> when I write "rc" I want rc - and not anything that is accidently in the way ...
19:08 < mnemoc> you are mapping pkgsel you your ugly sed matches which are intrinsicly wrong because you depend on repo or package name
19:09 < mnemoc> ' rc ' will only match rc
19:09 < CIA-14> rene * r13231 /trunk/package/filesystem/devfsd/ (devfsd.desc devfsd.init):
19:09 < CIA-14> * do not build devfsd by default anymore and do not enable it during
19:09 < CIA-14> bootup automatically
19:09 < mnemoc> multimedia/lib* will put libidibum which is not a library on you package selection
19:10 < rxr> ok - we are only talking about a minor syntactic detail what to match ...
19:10 < mnemoc> (just a sample)
19:10 < rxr> we can decide that when we have useful categories ...
19:10 < mnemoc> it's not minor because you changed how repositories are matched
19:10 < rxr> mnemoc: ok - but audio/*jack* is valid, no?
19:10 < rxr> unless you want a jack category ,-)
19:10 < mnemoc> *jack*
19:10 < mnemoc> why audio?
19:11 < rxr> well - even jack, was just an example ..
19:11 < mnemoc> if a bunch of packages from a repo are needed together lets do a category for them
19:11 < rxr> mnemoc: let's decide in which tiny differing syntactic way to handle categories vs. repositories when we have sensible cats ...
19:12 < mnemoc> when we start using categories we will get categories
19:12 < mnemoc> you change was a big change
19:12 < mnemoc> your*
19:12 < rxr> mnemoc: do you really thing we want to maintain a set of to be maintained categories when one target just wants to select 5 programs it needs ...
19:12 < mnemoc> we can improve without changing current semantic
19:12 < mnemoc> if target wants the package, add the package explicitly
19:13 < rxr> a change was needed base/ would have matched base/ categories as well ...
19:13 < rxr> also do you see the need for the = ?
19:13 < mnemoc> if they are really tied, lets tie them
19:13 < mnemoc> i like !
19:13 < mnemoc> what is = ?
19:13 < rxr> take the default state for that match
19:14 < rxr> mnemoc: read my mail on the list
19:14 < mnemoc> = reset to what .desc say?
19:15 < rxr> nope
19:15 < rxr> the Create-PkgList generated state of course
19:15 < rxr> that is per architecture mostly
19:16 < rxr> how do you want to select arch specific stuff in the pkgsel otherwise ?
19:16 < rxr> x86/ will enable x86 stuff on non x86 arches ...
19:19 < rxr> which is of course also what .desc says for non arch packages ...
19:19 < mnemoc> the whole repo, or using categorization, or always explicit, or by flag.... so finally they will be disabled due to arch restrictions on .desc
19:20 < rxr> ?
19:20 < rxr> they will not be disabled later on
19:20 < rxr> since arch restrctions comes first and then it goes thru the pkgsel ...
19:20 < mnemoc> X x86/ will add all x86, but if it's for ppc they wont exist
19:20 < rxr> ??
19:20 < rxr> sure they exist as = and X x86/ turned them on ...
19:20 < rxr> (and still turns)
19:21 < rxr> just that you use = x86/ now ...
19:21 < mnemoc> if arch = sparc, x86/ packages will never be on packages file because Create-PkgList will ignore them
19:22 < rxr> oh - you are right, did not noticed that ,-)
19:22 < rxr> but still - it honors .desc
19:22 < mnemoc> so X BOOT is enough
19:22 < mnemoc> on generation
19:22 < mnemoc> after that you should only loop over 'config/$config/packages'
19:23 < mnemoc> i don't see the point for '='
19:24 < rxr> i see it ,-) don't use it then ,-)
19:24 < mnemoc> for foo-% we will need a macro (to add new lines on packages files)... i was thinking in '+' as oposite of '-'
19:25 < mnemoc> good :)
19:26 < mnemoc> but on patterns i think all we need is 'repo/' whole repo, and if don't end with / match any field except the repo
19:27 < mnemoc> and categories can do the rest
19:27 < mnemoc> pakcage name is not under our control
19:27 < mnemoc> repositories are only one so too much stuff enters on them
19:28 < mnemoc> so the only thing we can really contorl is categorization
19:28 < mnemoc> desktop/kde/office
19:28 < mnemoc> desktop/kde/*
19:29 < mnemoc> desktop/*/core
19:29 < mnemoc> games/rpg
19:29 < mnemoc> lets see what categories do we need for our current targets
19:30 < mnemoc> minimal, minimal-xorg, minimal-desktop, desktop, livecd, archivista
19:30 < mnemoc> explicit pkgsel is bad because it miss new deps
19:31 < rxr> yeah - that is another thing
19:31 < mnemoc> repo or package name matching is bad because miss relocations and add new packages which noone expected to be added
19:31 < mnemoc> so we have to maintain categories
19:31 < rxr> let's not put too much more work on it - because in 2.3 or so I want to select by edges of the dep graph
19:31 < rxr> like X koffice ooo xterm (or so) and have all selected that is needed to build it ...
19:31 < mnemoc> O packages wont ever be on your dep graph
19:32 < rxr> just eleminating most of current pkgsel ..
19:32 < mnemoc> to do that we need to discriminate between 'levels' of dependency
19:32 < rxr> mnemoc: those can still be added using the other mechanims we have now and talked about
19:32 < mnemoc> X vim adding X?
19:32 < mnemoc> xorg*
19:33 < rxr> X vim does only added libs and proto packages required to build vim
19:33 < mnemoc> auto-deps yes, but only the minimal set
19:33 < rxr> if you want a text only write O xorg/ at the bottom, no?
19:33 < rxr> yeah - or the other way round ...
19:34 < mnemoc> but autodeps is a whole new field :)
19:34 < mnemoc> i was discussing against pkgsel semantic :p
19:34 < mnemoc> the change on pkgsel semantic*
19:35 < mnemoc> i thinkg mixing repo/pkg is harmfull and unnecesary
19:36 < mnemoc> repo and package names are almost an accident, and shall be avoided as much as posible on sematic processing
19:37 < rxr> mnemoc: why it is harmful - when I want and find this useful: multimedia/lib* why you want me to stop using it ?
19:37 < mnemoc> because that should be used for categories
19:37 < rxr> I wasted a lot of hours on the first category reorganization in ROCK times
19:37 < rxr> I'll never do this again
19:37 < rxr> and until someone rewrites all categories to be useful I want to select packages this way
19:38 < mnemoc> mnemosyne depend on pkgsels, i use lot of pkgsel 'bricks' if semantic changes i would have to fork it
19:38 < rxr> mnemoc: can you look at livecd/pkgsel and rewrite it the way it should be and produce the same package files as it does right now ?
19:38 < rxr> mnemoc: did you use stuff that is no longer working ?
19:38 < rxr> s/did/do/
19:38 < rxr> do yo uselect on category ?
19:38 < mnemoc> what stuff?
19:39 < rxr> stuff that worked t-2h ago and I removed ?
19:39 < mnemoc> phone
19:40 < rxr> ok /me swimming hall
19:41 < rxr> back on the terminal in 2h or so ...
19:42 < mnemoc> re
19:43 < mnemoc> i have been jumping around the whole day
19:43 < mnemoc> i ahven't touched anything
19:43 < mnemoc> i think multimedia/lib* is harmfull because makes imposible to match categories in any future
19:44 < rxr> category/foo/bar
19:44 < rxr> mnemoc: did my changes broke any of your pkgsel rules ?
19:45 < mnemoc> i haven't touch my repos since ooo update.... two days ago
19:45 < rxr> well - ok
19:45 < rxr> i take a look at your current rules
19:45 * rxr really gone now
19:45 < rxr> cu
19:45 < mnemoc> that's not the point
19:45 < mnemoc> ok, cu rxr
19:46 < karasz> have fun rxr
19:46 < mnemoc> i have only explict packages now
19:46 < mnemoc> but a feature that makes imposible a grow we must have, it's a damange and not a feature, even if it's conformatable today
19:56 < karasz> for an athlon xp what iso is recomended pmmx?
20:02 < mnemoc> yes
20:03 < madtux> sparc64
20:04 * madtux hides
20:13 * karasz finds madtux :D
20:14 * mnemoc wonders what karasz does to madtux now
20:15 < madtux> lol
20:16 * sepp_ unhines and say hello
20:16 -!- sepp_ is now known as sepp
20:16 < sepp> unhides*
20:16 < madtux> hello unhided sepp
20:17 < mnemoc> watch out, karasz is free
20:17 < sepp> :D
20:17 < sepp> its trapping time?
20:18 < mnemoc> karasz just did very nasty things to poor madtux
20:18 < madtux> 0_o
20:18 < mnemoc> look at his poor face now :(
20:18 < madtux> mnemoc, you sick little BOFH
20:18 < madtux> :)
20:18 < mnemoc> i'm not a bofh
20:19 < mnemoc> i'm just a poor coder
20:19 < madtux> BCFH then
20:19 < mnemoc> :)
20:23 < karasz> mnemoc you ruin my reputation... :)
20:23 < mnemoc> :)
20:23 * karasz thinks `oh wait I didn't have one in the first place!?`
20:24 < valentin> re
20:24 < mnemoc> madtux is a cute teddy boy :) that wont ruin your reputation
20:24 < madtux> karasz, don't mind... he did not know that.. so have fun making him feel guilty
20:24 < mnemoc> :)
20:24 < madtux> teddy boy??! OMG.
20:24 < karasz> nah i'm not that kind :)
20:24 < madtux> valentin, hello.
20:25 < mnemoc> yellow, blue, brown.... brown is ground right?
20:25 < karasz> mnemoc do not play with electric stuff
20:25 < karasz> we still need you :)
20:27 * mnemoc has nothing to test except his tonge :\
20:27 < karasz> 0_0
20:28 * karasz fingers crossed ;)
20:28 < karasz> mnemoc are you still there?
20:28 < mnemoc> yes, i have done thing yet
20:28 < karasz> god
20:29 < karasz> you really scared me
20:29 < mnemoc> just cuting cables with a knife
20:29 < karasz> testing sharpness with your tongue?
20:29 < mnemoc> nah ;)
20:29 < karasz> ;)
20:30 < karasz> damn my net is slow... :(
20:30 < madtux> well mnemoc your tongue is not needed for coding its alright to test the knife's sharpness with rather than ur fingers
20:34 < mnemoc> the yellow is ground
20:38 < valentin> mnemoc: did you succed in recovering some documents ?
20:38 < mnemoc> just a few
20:38 < valentin> :(
20:39 < mnemoc> "shit happens"
20:41 < mnemoc> weirdly i recovered subdirs, but not documents at the root of my mistake
20:49 < valentin> huh - which hash is beeing used for /etc/passwd passwords ?
20:50 < mnemoc> crypto :p
20:50 < mnemoc> crypt*
20:50 < mnemoc> + fixed 'salt'
20:51 < valentin> i just found out that my university account only uses the first 8 letters of my passwd and ignores appended stuff
20:51 < _Ragnar_> fixed salt? that kinda defeats the purpose
20:53 < mnemoc> valentin: i think it's more like login/passwd cutting it, that encription only considering first 8
20:54 < valentin> damn - and i wasted my time typing all 10 for two years now...
21:50 -!- jsaw [n=jsaw@volans.mpimf-heidelberg.mpg.de] has quit [Read error: 104 (Connection reset by peer)]
21:51 -!- jsaw [n=jsaw@volans.mpimf-heidelberg.mpg.de] has joined #t2
21:54 < rxr> re
21:54 < rxr> md5
21:55 < mnemoc> :)
21:55 < rxr> md5 on modern systems
21:55 < rxr> in rock since 199x ...
21:56 < mnemoc> even without shadow?
21:57 < rxr> mnemoc: with pam it depends what back end it is anyway ...
21:57 < mnemoc> i think that's not valentin's case :)
21:58 < mnemoc> password trimmed to 8 chars
22:00 < mnemoc> rxr: did you try to build ooo?
22:00 < mnemoc> i'm not sure if it builds for gcc4
22:01 < mnemoc> java part builds with java 1.4 but not 1.5 due to 'strictness' issues
22:01 < rxr> mnemoc: not yet
22:02 < rxr> I wanted to build ooo when the rest of the livecd is streamlined
22:02 < rxr> sun is great
22:02 < mnemoc> i love sun almost as much as microsoft
22:03 < MisterN> who needs java in ooo? i thought it was needed for the database stuff only :D
22:03 < mnemoc> help needs java
22:04 < MisterN> *checks*
22:04 < rxr> mnemoc: even the hlpe ?
22:04 < MisterN> nah help runs fine with java disabled
22:05 < mnemoc> at least xmlhelp module needs java and bdb :p
22:05 < mnemoc> i don't use ooo
22:06 < rxr> MisterN: dit you built from source ?
22:06 < rxr> did even
22:06 < mnemoc> MisterN: are you using t2's ?
22:06 < MisterN> no from binary :)
22:06 < MisterN> and i have no t2 thingy
22:06 < MisterN> :)
22:06 < mnemoc> most ooo deps are only at build time
22:06 < MisterN> oh ok
22:07 < rxr> MisterN: you run Gentoo right
22:07 < rxr> ? even
22:07 < MisterN> rxr: yes i do
22:07 < rxr> and you Emerged ooo-binary or what it is named ?
22:07 < MisterN> openoffice-bin
22:08 < MisterN> yes
22:08 < mnemoc> built on their farm or by sun/ximian?
22:08 < MisterN> by sun/ximian
22:09 < mnemoc> do they have ooo as source package?
22:09 < MisterN> yes
22:09 < MisterN> but not for my platform
22:10 < MisterN> (the -bin is not for my platform, either :D)
22:10 < mnemoc> which?
22:10 < CIA-14> rene * r13232 /trunk/target/archivista/rootfs/home/archivista/install.sh:
22:10 < CIA-14> * fixed archivista/install.sh for the case where no update is possible
22:10 < CIA-14> by initializing a variable and added som quoting to debug output for
22:10 < CIA-14> easier communication with customers on the phone ...
22:11 < MisterN> my platform: amd64
22:12 < mnemoc> rxr: is noisy running continues now?
22:15 * mnemoc just notice rene is using my 'mnemoc' screen at noisy :p
22:19 < mnemoc> rxr: http://t2.geeks.cl/regressions/stable/
22:37 < rxr> mnemoc: misl said so
22:37 < rxr> mnemoc: ?
22:37 < rxr> mnemoc: nope - I never touched your screen there ...
22:38 < rxr> mnemoc: do you sell a tape to that company now?
22:38 * rxr dinner
22:40 < mnemoc> rxr: i'll have to prepare a bacula facility :)
22:41 * mnemoc to pick up his girls
22:41 < mnemoc> back in 1h
22:59 < MisterN> n8
22:59 -!- MisterN [n=MrN@p549CF8A1.dip.t-dialin.net] has quit ["Wieso gibt es keine guten Quitmessages? Jaja, diese ist auch doof."]
23:06 < rxr> mnemoc: have you mailed mr. archivista sss
23:06 < rxr> ? even
23:06 < rxr> n8 - cu early
23:26 < karasz> mnemoc around?
23:26 < madtux> he is away for a little.. he should be back shortly
23:26 < karasz> i just noticed that...
23:27 < madtux> karasz, even though, if you need help, perhaps i can be helpful ?
23:27 < karasz> thanks for the offering but i just wanted to talk to him abt postgresql 8.1 being released
23:28 < karasz> he is the mantainer of postgresql package, isn't he?
23:29 < madtux> u just need to ask the .desc file :)
23:30 < karasz> i can not check now but i think that now in T2 is 8.0.3 beta or some...
23:30 < madtux> let me see
23:30 < karasz> now 8.1 is released as "stable"
23:31 < madtux> current package is "Stable 8.0.4"
23:31 < karasz> 2.1.0?
23:31 < karasz> or trunk?
23:31 < madtux> trunk
23:32 < madtux> http://svn.exactcode.de/t2/trunk/package/database/postgresql/postgresql.desc
23:32 < karasz> my opinion is that trunk should have 8.1.0 and 2.1.0 should remain as is because is released
23:32 < karasz> but than again that is just my opinion...
23:33 < karasz> mnemoc should decide
23:34 < karasz> i meant that is mnemoc's decision (guess i am tired :( )
23:34 < madtux> i agree with ur opinion about 2.1.0 ... unless mnemoc has made branch to keep 2.1.0 updated
23:34 < karasz> but trunk is not suposed to be bleeding edge?
23:35 < karasz> anyway at least the very last "stable" releases?
23:35 < madtux> but if u do not have write access to svn please feel free to send an update patch to the mailing list.. but please make sure to build and test the package before submiting the update
23:35 < karasz> i know, i know
23:35 < karasz> other that that mnemoc will rip my head off :)
23:36 < madtux> no... rene will :)
23:36 < karasz> and no i do not have svn write because i am not a developer
23:36 < karasz> i am just lingering around trying to learn something
23:37 < karasz> kinda' apprentice stuff
23:37 < madtux> well starting with update patches can be a good beginign to learn more about t2 internals.. and as well to start making a path to become a future developer
23:37 < madtux> :)
23:38 < madtux> i have write access, but i rather have u send the update patch so that u can learn about package structure and updates, and so that mnemoc gets to see it
23:38 < karasz> now i do not have a T2 machine... :( i had to make an urgent server for a customer
23:38 < madtux> i see. so what are u currently running?
23:38 < karasz> when i had such a machine i was a little more active, even added konversation to T2
23:39 < madtux> NOTE: if the answer is windows, please just say .. something useless for the matter
23:39 < karasz> which i am very proud of even if it is really nothing
23:39 < karasz> i didn't understood the last line...
23:40 < karasz> ???
23:42 < karasz> am I lagging here?
23:42 * karasz wonders...
23:46 < karasz> hmmm my connection could't be that bad...myabe i scared madtux off?
23:47 < karasz> yes i am definitely tired
23:47 < karasz> my spelling is horrific
23:49 -!- morfoh [n=jeru@p54BEF712.dip.t-dialin.net] has joined #t2
23:50 < morfoh> moin
23:50 < karasz> hehe moin morfoh
23:51 < morfoh> hehe moin karasz :)
23:51 < morfoh> karasz: still banging your head ? ;)
23:51 < karasz> nope
23:51 < madtux> rats in the channel i smell
23:51 < karasz> i have a very strong head an week walls here :D
23:52 < morfoh> madtux: I just smell a "pyromaniac" ;)
23:53 * karasz googles around for a smell enabled irc client ....0_0
23:54 < morfoh> madtux: but where is the guy who paints pigs ? :D
23:54 < madtux> haha
23:55 < morfoh> karasz: you'll need a special service provided by "hostarica" for the "smell-feature"
23:56 < morfoh> not to mention a ill brain :)
23:56 * morfoh hides
23:57 < morfoh> does somebody know the status of mnemoc's "adventure" to rescue the lost data ?
--- Log closed Wed Nov 09 00:00:53 2005