--- Log opened Mon Dec 06 00:00:19 2004
00:00 < _martin_> no, and I have no time to do no.2 know. I will do it next week
00:01 < _martin_> bash$ less Termine
00:01 < _martin_> Termine fuer Abgabe der Aufgaben in SYP:
00:01 < _martin_> SYP_Ex1: 16.11.04 / 23.11.04
00:01 < _martin_> SYP_Ex2: 7.11.04 / 14.11.04
00:01 < _martin_> $ date
00:01 < _martin_> Mon Dez 6 00:04:08 CET 2004
00:02 < _martin_> ?? Lucius is so ....
00:02 < rxr> yesp
00:03 < rxr> this is what I meant ...
00:03 < rxr> 23:46 < rxr> exercise 2 should be handed to him in mid. of November ... ?!?!
00:05 < _martin_> Aehm, okay. Ich weiss auch nicht weiter, ist einfach nur schrecklich dieser Mensch.
00:05 < rxr> mal sehen was er zu meiner mail sagt ;-)
00:05 < rxr> oh - english even ..
00:11 < rxr> ok - I'll heat up a pizza and then do boring DOS assembly exercies over the night ...
00:11 < rxr> .oO(how my study keeps sucking)
00:20 < rxr> wah!
00:21 < daja77> O_o
00:21 < rxr> the tfh-berlin.de mail server rejects my mail due to unicode escaping in the header? like in my name and the subject?
00:21 < rxr> never saw that broken stuff ...
00:22 < rxr> From: =?UTF-8?B?UmVuw6kgUmViZQ==?=
00:22 < rxr> Subject: SYP =?UTF-8?B?TMO2c3VuZyBBdWZnYWJlIDEgdW5kIDI=?=
00:22 < rxr> damn MTA ...
00:22 < rxr> and/or SPAM filter ...
00:24 < rxr> This nondelivery report was generated by the amavisd-new program
00:24 < rxr> at host mailbox1.tfh-berlin.de.
00:27 < _martin_> I don't understand the tfh-mail-server. Kordecki means, there is a spam filter between mail1.tfh-berlin.de and mail2.tfh-berlin.de. And btw: All filtered mails are stored on a big harddisk with no access from any users
00:28 < mnemoc> rehi
00:28 < rxr> hi mnemoc
00:28 < mnemoc> hi rxr
00:28 < rxr> kordecki is a mindless person anyway
00:29 < rxr> so - immediatly complained back to postmaster@tfh-berlin.de ...
00:29 < rxr> INVALID HEADER (INVALID CHARACTERS OR SPACE GAP)
00:29 < rxr> Non-encoded 8-bit data (char F6 hex) in message header 'X-Spam-Report'
00:29 < rxr> X-Spam-Report: ...stratoren der TFH-Berlin, k\366nnte es\n\tvielle...
00:30 < rxr> ^- shit f*cking hell - that is _not_ the header ...
00:32 < rxr> oh - that is injected by my spamassassin ...
00:33 < rxr> guess I should put some work in to no inject the report - at least not for outgoing mail ...
00:51 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
01:07 < rxr> hi sparc-kly
01:07 < sparc-kly> hi rxr :P
01:10 < jsaw> re
01:10 < jsaw> hi sparc-kly , rxr
01:11 < jsaw> rxr: you said once that distcc could be a bit problematic with t2/rock. did I remember corectly and if yes, why_
01:11 < jsaw> ?
01:16 < rxr> mostly because you have not the resolver config in the chroot ...
01:16 < jsaw> what else?
01:17 < rxr> but when you only use IP addresses for the nodes and add some glue code to pass the HOST_LIST s.th. into the chroot it would work
01:17 < rxr> and/or pass the system recolv.conf into it, too
01:17 < rxr> mostly this config / networking issue ...
01:18 < rxr> do you plan s.th. based on distcc ?
01:19 < jsaw> We have around 10 PC's in the lab at experiments (so not used during the weekend or night most of the time)....
01:20 < jsaw> -> this cries for a minimal distcc livecd...
01:21 < rxr> ;-)
01:21 < mnemoc> jsaw: why not beowulf or mosix?
01:21 < rxr> you would not even need a live-cd the nodes can run the distcc as normal user ...
01:21 < rxr> mnemoc: mosix is painfully slow ... - far too much i/o in this workload ...
01:22 < rxr> and beowolf is application suppoted distribution - so you would need to write the application - and that is what distcc already is
01:23 < mnemoc> so for building only distcc is option
01:23 -!- cyberchills [~nate@c-24-98-58-49.atl.client2.attbi.com] has joined #t2
01:23 < rxr> you can use mosix - but you need even more specialized cluster nodes and it would be slow ...
01:23 < cyberchills> ho all
01:24 < jsaw> these are windows machines and I'm not allowed to install linux...
01:24 < rxr> you can run distcc on any other linux box ... (in theory even on BSD and so on)
01:24 < rxr> hi cyberchills
01:24 < jsaw> dito
01:24 < jsaw> hi mnemoc btw
01:24 < rxr> mnemoc: for distributed compilint the specialized tools for that are far advanced (distcc, icecream, ...) compared to mosix
01:25 < rxr> or you implemenent it yourself like the old ROCK style clustering
01:25 < mnemoc> i don't think clustering is business of the build kit :)
01:25 < rxr> me neither
01:26 < rxr> and the old rock code sucks
01:26 < cyberchills> whois rxr
01:26 < rxr> error prone and very static in the cluster configuration
01:26 < jsaw> cyberchills: ...
01:26 < rxr> e.g. you can not not just plug a laptop in for a few hours and replug it from the build
01:27 < rxr> and clifford ruined it with the src.* to -> build/.../ROCK/src.* migration ..
01:27 < rxr> for the rock cluster you need build on a NFS and this kills performance by magnitudes
01:28 < rxr> clifford claimed build on tmpfs would bring it back - even without much ram and tmpfs swapping out to the local discs ...
01:28 < rxr> but later measurements even from him showed that performance still sucks
01:29 < rxr> so: 2.1 Cluster builds are borked ...
01:29 < rxr> but I do not intend to change this in T2, I would rather rip all that code out and add proper distcc support
01:29 < mnemoc> how does distcc share the stuff? shamred memory?
01:29 < rxr> distcc preprocessed the whole file and sents the preprocessed output over the wire ...
01:30 < rxr> so the node do not need the same headers and such - they just compile the raw C output into an object and sent it back
01:30 < rxr> so the nodes are very leightweight
01:30 < rxr> lightweight even
01:32 < mnemoc> sounds cool
01:33 < rxr> and with distcc you can plug in nodes on the fly - not like the static rock thing ...
01:33 < rxr> and remove even ...
01:33 < rxr> and this is not esily fixable in the rock shell script ...
01:33 < jsaw> so, setting [GK]CC_WRAPPERS_OTHERS + adding -jX to make options and DISTCC_HOSTS is enough
01:33 < jsaw> ?
01:34 < jsaw> oh, it can even use lzo to transmit the data
01:35 < rxr> yep
01:35 < rxr> I do not know it it works via the _OTHERS
01:35 < jsaw> it is done with ccache..
01:35 < rxr> you need to try if distcc has a change to determine if the source is C or C++ ...
01:36 < jsaw> Doom3's Linux build system uses SCons, CC="ccache distcc g++-3.3" JOBS=8 rocks!" Timothee Besset, id Software
01:36 < rxr> the official way to let distcc discover the source type by the program name so you have:
01:36 < rxr> gcc -> distcc
01:37 < rxr> g++ -> distcc
01:37 < rxr> ...
01:37 < rxr> I tried without it - but it will not work when the C++ files are named .c - this is the case in some packages we include ...
01:37 < jsaw> but the example from doom seems to suggest, that you can tell distcc which complier to use
01:37 < rxr> hm - might be
01:38 < rxr> or that is the syntax for the toolchain swtich feature
01:38 < jsaw> man page: distcc [COMPILER OPTIONS]
01:38 < rxr> you can tar up your toolchain and distcc will automatically copy the needed compiler to the host system ...
01:38 < rxr> oh - ok ;-)
01:38 < rxr> then it is that easy
01:40 < jsaw> If the compiler name is an absolute path, it is passed verbatim to the server and the compiler is run from that directory. For example:
01:40 < jsaw> distcc /usr/local/bin/gcc-3.1415 -c hello.c
01:40 < jsaw> If the compiler name is not absolute, or not fully qualified, distccd's PATH is searched. When distcc is run from a masquerade directory, only the base name of the compiler is used. The client's PATH is used only to run the preprocessor and has no effect on the server's path.
01:45 < cyberchills> Sorry to inturrupt but I want to know how I can help w T2
01:46 < rxr> cyberchills: just start using it ;-) report success, failures, suggestions - update your favourite packages ;-) or implement features you need, ... ;-)
01:47 < cyberchills> I am currently working on building a minimal system to properly run the asterisk PBX.
01:55 < jsaw> sounds interesting
01:55 < rxr> seems like a nice task to get intouch with T2 ;-)
01:56 < jsaw> who are you working for, cyberchills? Or is it a private project?
01:56 < nullslack> cyberchills, way to go...i'll be looking forward to it ;)
01:56 < jsaw> hi nullslack
01:56 < nullslack> == 12/05/04 17:43:50 =[5]=> Aborted building package linux26.
01:56 < mnemoc> :(
01:56 < nullslack> hello jsaw
01:56 < rxr> cyberchills: you might like by starting our asterisk package - I saw it a release of it some time ago - but had not time to verify the update
01:57 < nullslack> !> Sub-building xorg ...
01:57 < mnemoc> what sub-building means?
01:58 < jsaw> drm
01:58 < jsaw> see package/x11/xorg/postlinux.conf
01:58 < rxr> mnemoc: ;-) a new t2 feature ;-)
01:59 < jsaw> mnemoc: remember the change to remove linux-src?
01:59 < jsaw> nullslack: more info on the build error
01:59 < cyberchills> gotcha
02:01 < jsaw> ?
02:01 < cyberchills> I will check it out
02:02 < mnemoc> aha
02:03 < rxr> cyberchills: I guess xorg fails do to generic i386 build and not CMPXCHG is available
02:04 < rxr> mnemoc: mail fwd.
02:05 < nullslack> !> File not found: download/mirror/X/X11R6.8.1-src1.tar.bz2
02:05 < nullslack> !> -> Reading package configuration (postlinux.c ..
02:05 < rxr> nullslack: did you run Download -required?
02:06 < nullslack> rxr, yes...it's there in the download dir
02:06 < rxr> ?
02:09 < nullslack> damn!
02:09 < nullslack> -rw-r--r-- 1 root root 30562401 2004-12-05 08:01 X11R6.8.1-src1.tar.gz.cksum-err
02:09 < nullslack> in download dir ;(
02:10 < jsaw> mv X11R6.8.1-src1.tar.gz.cksum-err X11R6.8.1-src1.tar.gz and try ./scripts/Download -required again
02:10 < nullslack> jsaw, ok..will do. thanx
02:10 < jsaw> if it's still wrong, remove the file...
02:11 < nullslack> oh..yeah...i have to remove the cksum errs..it complains about that ;)
02:12 < nullslack> INFO: download from mirror failed, trying original URL.
02:13 < jsaw> rm -f download/Mirror
02:13 < jsaw> okay, gotta go. cu!
02:18 < rxr> cu jsaw
02:25 < cyberchills> gotta jet... give you all updates in a bit
02:26 -!- cyberchills [~nate@c-24-98-58-49.atl.client2.attbi.com] has quit ["using sirc version 2.211+KSIRC/1.3.11"]
02:32 < nullslack> curl: (10) the username and/or the password are incorrect
02:34 < rxr> hm - for which file?
02:35 < nullslack> INFO: download from mirror failed, trying original URL.
02:35 < nullslack> any available mirrors i can force the dowload?
02:36 < nullslack> it's the same csum errors again when downloading from the original url
02:44 < rxr> crazzy freedos has a graphical installer these days ...
02:44 < rxr> nullslack: still - which package is the system trying to download there?
02:47 < rxr> Press OK to run the post-instal script. ;-)
02:48 < nullslack> rxr, post-install? yes...it's downloading the x packages
02:48 < nullslack> Downloading download/mirror/X/X11R6.8.1-src1.tar.gz ...
02:48 < rxr> I was just joing about freedos ...
02:48 < rxr> ah - still X org damn
02:48 < rxr> I take a look over here ..
02:49 < nullslack> yeah..freedos is cool...last thing i've tried it they don't have this gui thing..now they've gone a bit far now
02:49 < nullslack> and they already have cdrom support already
02:50 < nullslack> oh duplicate ;0
02:50 < rxr> yep
02:51 < rxr> I just need it for some annoying assembly hacking exercies for university :-(((
02:52 < nullslack> hmmm....like clearing the bios password using assembly? ;)
02:52 < rxr> no - just wasting my time with nonsense ... :-(((
02:53 < nullslack> bzip'ing + cksum-test: download/mirror/X/X11R6.8.1-src1.tar.gz
02:53 < nullslack> maybe that's the problem ;(
02:53 < nullslack> Cksum ERROR: download/mirror/X/X11R6.8.1-src1.tar.gz.cksum-err (2061106526)
02:54 < rxr> no - the problem is that you get other files than I when I packaged it ...
02:54 < rxr> nullslack: btw. what does cat download/Mirror yield ?
02:55 < rxr> I review this right now and update the checksums again if needed ...
02:56 < nullslack> http://gsmp.tfh-berlin.de/mirror/t2-source/2.1
02:56 < _martin_> n8 all
02:56 < nullslack> n8 _martin_
02:56 < nullslack> can't i force another mirror?
02:56 < rxr> _martin_: g n8 ;-~)
02:57 < rxr> we do not yet have a more mirrors - and our mirrors all contain of course the same files ...
02:57 < rxr> the problem is that the checksum changed again
02:58 < nullslack> ;) i guess i'll just have to wait...
02:59 < rxr> well you could remove the checksums from the xorg.desc file ...
02:59 < rxr> then they would not be chkecked anymore ...
02:59 < rxr> just remove the number and insert 0 ....
03:00 < nullslack> hmmm...lemme check
03:02 < nullslack> i have xorg.desc open now...which one are u referring?
03:02 < rxr> at the buttom the [D] lines ...
03:02 < rxr> those with ftp.x.org/ ...
03:03 -!- sparc-kly [~mubex@64.237.130.173] has quit [Read error: 104 (Connection reset by peer)]
03:03 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
03:03 < rxr> the numer in the second collumn need to be zeroed ...
03:03 < rxr> glibc-audit 2.3.3-74
03:03 < rxr> ^- mnemoc jsaw maybe this will be nicer to use ;-)
03:03 < nullslack> [D] 3367587284 X11R6.8.1-src1.tar.gz ftp://ftp.x.org/pub/X11R6.8.1/src/
03:04 < nullslack> i'll replace 3367587284 with 0
03:05 < nullslack> oh...got that1
03:05 < rxr> yep
03:05 < rxr> on all seven or so x.org downloads ...
03:07 < rxr> The patch modifies 91 files. Compared to glibc-2.3.2-27.9, the additional code occupies 18 more bytes of .text, and 24 fewer bytes in the .so. On a nano-scopic scale, the typical execution cost is 0 to 3 CPU cycles per affected routine; the estimated median total impact is less than 1 second per machine per day. In the case of *printf(), glibc-audit is faster than glibc because the cleaned-up source helps gcc-3.2 avoid generating atrocious code when initializing printf_spec.info for parse_one_spec() in stdio-common/printf-parse.h.
03:08 -!- CIA-9 [~CIA@to.je.spocco.com] has quit [Excess Flood]
03:09 < rxr> ^- glibc-audit
03:10 < rxr> nullslack: I can not yet access the X.org ftp due to overload there ...
03:12 < nullslack> curl: (10) the username and/or the password are incorrect
03:12 < nullslack> overload?
03:12 -!- CIA-8 [~CIA@to.je.spocco.com] has joined #t2
03:12 < rxr> I get an different one ...
03:13 < rxr> nullslack: ok - I download them from an X.org mirror and push the corrent checksum into the file and the files onto the t2 mirror
03:14 < nullslack> oh..so no need to edit xorg.desc
03:15 < nullslack> snv up ;)
03:15 < nullslack> svn
03:15 < rxr> yeah - not yet - takes some minutes ...
03:15 < rxr> you need to svn revert the xorg.desc - otherwise svn up will get conflicts when it tries to inject my updates checksums ...
03:16 < nullslack> how bout if i delete xorg.desc? and do svn up?
03:16 < rxr> yes - that works too
03:17 < nullslack> thnx...just let me know if i can do svn up already ;)
03:19 < nullslack> rxr, btw, are the tar.gz's the same as the maintainer package them? i mean the same tar.gz's or bzip2's from the maintainers download site. no renaming or repackaging whatsoever
03:22 < rxr> yes
03:22 < rxr> those files are completely unmodified - when we modifiy it there is a .patch in the package directory ...
03:22 < rxr> xorg has some ...
03:24 < nullslack> pretty much to my liking...the reason is that the builder doesn't depend on the buildkit in terms of packages...they could go directly through the maintainers site..am i right?
03:25 < rxr> yes, sure
03:25 < rxr> and we try to patch as less as possible ...
03:25 < rxr> so that a simple version modification is enough to update most pacakges
03:25 < rxr> packages even
03:25 < nullslack> and they could extend the buildkit to there liking without worrying about the packages ;)
03:25 < rxr> yep
03:27 < nullslack> lets say...the t2 developers don't want to integrate the change of a builder...they could just do it themselves with much ease
03:27 < rxr> yes - but I hope this will nearly never happen ...
03:27 < rxr> except for embedded stuff where some company does not want the changes to be merged back ...
03:27 < nullslack> yeah...only for rare purposes only
04:41 < CIA-8> rene * r4961 /trunk/package/x11/xorg/xorg.desc: * reverted parts of r4890 - I commited crap checksums :-(
04:41 < rxr> xorg commit done
04:43 < rxr> syncing mirrors ...
04:43 < rxr> takes some time ...
04:56 < rxr> debugging code with the dump DOS debug is "fun"
04:57 < rxr> damn - not the night for assembly
05:17 -!- rxr_ [~rene@p213.54.223.101.tisdip.tiscali.de] has joined #t2
05:17 -!- Topic for #t2: T2 | the system development environment | http://www.exactcode.de/t2 | C++ people around, too
05:17 -!- Topic set by rxr [] [Mon Nov 29 02:10:27 2004]
05:17 [Users #t2]
05:17 [ _martin_] [ CIA-8 ] [ mnemoc ] [ rxr ] [ valentin]
05:17 [ _Ragnar_] [ daja77] [ nullslack] [ rxr_ ]
05:17 [ af_ ] [ jsaw ] [ nzg ] [ sparc-kly]
05:17 -!- Irssi: #t2: Total of 13 nicks [0 ops, 0 halfops, 0 voices, 13 normal]
05:17 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
05:17 -!- Irssi: Join to #t2 was synced in 15 secs
05:22 < rxr_> re
05:24 < rxr_> mirror syncing
05:24 < rxr_> so if you still do not get the file from the x.org master site the files will be on the mirror, soon
05:28 < nullslack> rxr_, thnx...i'll check it out now
05:33 -!- rxr [~rene@p213.54.234.148.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)]
05:36 -!- You're now known as rxr
05:41 < nullslack> rxr, then just run Build-Target again..right? nothing to be done with Config
05:43 < rxr> yep - nothing ...
05:43 < rxr> do you have the files now ?
05:43 < rxr> ah - they are also on the mirror now:
05:43 < rxr> wrote 236 bytes read 54659501 bytes 52785.84 bytes/sec
05:44 < nullslack> rxr, the first one is being check for the cksum
05:44 < rxr> /dev/ad0s1a 28564396 26609084 -329838 101% /
05:44 < rxr> .oO
05:44 < nullslack> 101%?
05:46 < rxr> it is a freebsd at our uni - don't ask - I'm neither in freebsd nor am I an admin of the box
05:46 < rxr> maybe freebsd displays 100% as user accessable space and the 5% reserved for root are >100% then - who knows ..
05:46 < rxr> or sofware bug - or disk corruption due to hardware noise ...
05:46 < rxr> I ask the admin today - when I'm @ uni ...
05:51 < nullslack> success!!!
05:51 < nullslack> Downloading download/mirror/X/X11R6.8.1-src2.tar.gz ...
05:51 < nullslack> ;)
05:55 < rxr> ;-)
05:55 < rxr> I was a fool injecting those crappy checksums days ago ... :-(
05:56 < nullslack> where did u get it actually?
05:59 < nullslack> == 13:08:37 =[5]=> Building base/linux26 [2.6.9 2.1.0-beta].
05:59 < nullslack> ;)
06:00 < rxr> no idea what I checksummed there - maybe some text that came from the site telling me to use a mirror *g*
06:36 < jsaw> rxr: sleep cycle totally off...?
06:37 < rxr> sure
06:37 < rxr> I still have some pressure due to boring uni stuff I never was motivated for ...
06:37 < jsaw> hehe
06:37 < rxr> so I need to get assembly hacking, and some other fluff done this semester ... :-(
06:38 < rxr> aside from my contract work that needs an beta version until next monday ...
06:38 < rxr> btw. sleep, what was that?
06:38 < rxr> oh - and than this atmega proejct I neeed to do for Microcontroler (a lecture I do just for fun and have no time runaway ...)
06:39 < jsaw> don't ask me. sleep? never heard.
06:39 < rxr> but I rather do a bit more now and get this boring study done ...
06:40 < rxr> oops - just used vi keystrokes in the FreeDOS edit ;-)
06:40 * jsaw hands rxr some coffein^Wteein pills.
06:40 < rxr> pills?
06:41 < rxr> did you discarded drinking fluids as you did your sleep?
06:42 < jsaw> no. but you cannot call coffee a drink... (in the sense of needed fluids)
06:42 < jsaw> I only used one coffein pill in the last week before I finished my diploma thesis.
06:43 < jsaw> Now, I'm actually used to sleep deprivation... (It's called Marlene *g*)
06:43 < rxr> .oO (maybe I should invest into those, too)
06:44 < jsaw> guess a good tea will make it. Still. But when you haven't slept for 2 days in row... then maybe it'll help...
06:44 < rxr> hey I have slept yesterday (over parts of the day - and far too much - damn body ;-)
06:51 < rxr> damn - some key kombination kills the damn freedos edit
06:52 < rxr> some strange fn + ctrl + whatever I managed to acidently hit a few times this night
06:55 -!- rxr_ [~rene@p213.54.223.101.tisdip.tiscali.de] has joined #t2
06:55 -!- Topic for #t2: T2 | the system development environment | http://www.exactcode.de/t2 | C++ people around, too
06:55 -!- Topic set by rxr [] [Mon Nov 29 02:10:27 2004]
06:55 [Users #t2]
06:55 [ rxr] [ rxr_]
06:55 -!- Irssi: #t2: Total of 2 nicks [0 ops, 0 halfops, 0 voices, 2 normal]
06:55 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
06:55 -!- rxr [~rene@p213.54.223.101.tisdip.tiscali.de] has quit [Read error: 54 (Connection reset by peer)]
06:55 -!- Irssi: Join to #t2 was synced in 15 secs
06:55 [Users #t2]
06:55 [ rxr_]
06:55 -!- Irssi: #t2: Total of 1 nicks [0 ops, 0 halfops, 0 voices, 1 normal]
06:55 -!- You're now known as rxr
06:57 -!- CIA-8 [~CIA@to.je.spocco.com] has joined #t2
06:57 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
06:57 -!- _martin_ [~martin@brln-d9ba01de.pool.mediaWays.net] has joined #t2
06:57 -!- mnemoc [~amery@200.75.27.58] has joined #t2
06:57 -!- nullslack [user34@203.131.109.118] has joined #t2
06:57 -!- af_ [~af@ip-148-227.sn1.eutelia.it] has joined #t2
06:57 -!- nzg [~tschmidt@s15173431.rootmaster.info] has joined #t2
06:57 -!- _Ragnar_ [loki@abarrach.franken.de] has joined #t2
06:57 -!- daja77 [[hfOsKdTkh@odoaker.hrz.tu-chemnitz.de] has joined #t2
06:57 -!- jsaw [~jsaw@volans.mpimf-heidelberg.mpg.de] has joined #t2
06:57 -!- valentin [~valentin@port-212-202-170-162.dynamic.qsc.de] has joined #t2
06:58 < rxr> ounm
06:59 < rxr> what was that? IRC reset ?
07:00 < jsaw> no idea
07:03 < rxr> so - damn dos assembly does some BCD convertions ...
07:03 < rxr> so - time for a fast breakfast and then off into uni :-(
07:03 < jsaw> have fun...
07:03 < rxr> oh - damn - there is no convenient way to get the data out of qemu - darn ...
07:04 < rxr> except loop mounting my disk image ...
07:07 < rxr> I should run qemu with -fda /dev/scsi/...usbstick/partX
07:07 < rxr> just that I have no usb stick ...
07:09 * jsaw loves tracking latex style file errors... (forgotten % at end of line)
07:09 < rxr> yesh - tracking latex errors can be fun ...
07:09 < rxr> how is your research doing?
07:11 < jsaw> last night didn't work out. I still have to do some damned simple images of dummies and slices for the paper...
07:11 < jsaw> but...
07:11 < jsaw> maybe today.
07:13 < rxr> Track 01: Total bytes read/written: 401408/614400 (300 sectors).
07:13 < rxr> ^- assembly files
07:13 < rxr> because those damn windows 2000 boxes have no net access due to too many virus ...
07:16 < rxr> when I finished my muesli I'm off ...
07:16 -!- _martin_ [~martin@brln-d9ba01de.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
07:18 < jsaw> *grin* not net access...
07:20 < rxr> so - toothbrushing ....
07:20 < rxr> 3
07:21 < rxr> 2
07:21 < rxr> 1
07:21 < rxr> cu
07:22 < jsaw> cu
07:37 -!- mnemoc_ [~amery@200.75.27.45] has joined #t2
07:45 -!- mnemoc [~amery@200.75.27.58] has quit [Read error: 110 (Connection timed out)]
07:45 -!- mnemoc_ is now known as mnemoc
08:23 -!- mtr [~michael@Hba1a.h.pppool.de] has joined #t2
08:49 -!- _martin_ [~martin@217.186.6.94] has joined #t2
08:58 < valentin> moin
09:02 < nullslack> moin valentin
10:36 < rxr> re
10:36 < rxr> moin valentin
10:37 < rxr> so - got two assembly exercies handed away
11:11 < nullslack> rxr, how can i cancel the current build, if i want to build it later?
11:12 < rxr> Ctrl-C
11:13 < nullslack> hmmm...and t2 will rebuild it for me, without problems?
11:16 < rxr> yep
11:16 < rxr> if you do not power down the box use Ctrl-Z to suspend the process ...
11:16 < rxr> and "fg" to get it back running
11:16 < rxr> if you power down just power down or hit Ctrl-C before ...
11:17 < rxr> but the whole package will be rebuild - e.g. if mozilla was nearly finished - but was not yet finished - the wohle mozilla build will start from scratch
11:18 < nullslack> rxr, maybe you could include some explanations on what happens on each stages of the build, so the builder can get a big picture of what's going on ;)
11:18 < nullslack> isn't a package built in stages?
11:18 < nullslack> in the handbook, i mean
11:20 < rxr> nope - not the packages
11:20 < rxr> the stages are for:
11:21 < rxr> 0: build a new toolchain to get a working or even cross compiler
11:21 < rxr> 1: cross build the basic stuff
11:21 < rxr> 2: rebuild toolchain to be sure it is functional
11:21 < rxr> 5: build normal packages
11:21 < rxr> 9: rebuild all
11:22 < rxr> the rest is in between is for fun and normally not used
11:22 < rxr> (prev. we only had 5 stages in rock - and clifford meant he need to exapnd it just for fun)
11:23 < rxr> in t2 we consider to switch to "verbose" stages that would only be:
11:23 < rxr> TOOLCHAIN
11:23 < nullslack> oh...6 7 and 8? does minimal+xorg has a stage 9?
11:23 < rxr> BOOTSTRAP
11:23 < nullslack> rxr, yeah, that's much better
11:23 < rxr> NORMAL
11:23 < rxr> REBUILD
11:23 < rxr> (or so)
11:23 < rxr> the stage 9 can be selected in the config
11:23 < rxr> somwhere there is an option "Rebuild all packages in stage 9" or so ..
11:24 < rxr> you can selectt there whatever you want ...
11:24 < nullslack> so by default it is disabled ;)
11:24 < rxr> only my desktop target forces the stage9 off
11:24 < rxr> nope - default is on :-(
11:24 < nullslack> damn!
11:25 < nullslack> so rebuild is just for double checking>
11:26 < rxr> jups ...
11:26 < rxr> for paranoid people from the past ..
11:26 < CIA-8> rene * r4962 /trunk/scripts/config.in: * disabled the rebuild stage by default
11:26 < rxr> ;-)
11:26 < nullslack> CTR-C; svn up; Build-Target
11:26 < nullslack> ;)
11:27 < rxr> no
11:27 < nullslack> Config -oldconfig; Build-Target
11:27 < rxr> a) I'm not sure if that will redisable it for you
11:28 < nullslack> ;) how bout CTR-C; disable stage 9 in config; Build-Target?
11:28 < rxr> b) depending on what you want to do with your minimal you will miss the binary packages since they will only be tared up in the last stage
11:29 < rxr> so you right now have no binary packages - they would be created in stage 9
11:29 -!- sparc-kly [~mubex@64.237.130.173] has quit ["Leaving"]
11:29 < rxr> when you now disable stage 9 you will miss the binary package for those packages already built in stage 5 ...
11:29 < rxr> but when you only wanna chroot into it or so (e.g. no CD install) you can life with taht ...
11:30 < nullslack> hmmm...i would really chroot to it...until i've re-compiled my kernel to enable loop support
11:32 < nullslack> but...i guess i'll just wait for it to finish ;) i'm just eager to use what i built ;)
11:33 < nullslack> btw, suggestion, maybe we could use the cpu optimizations that the kernel config uses...i.e. they have PIII/Celeron, i586, i686, etc,,,
11:51 < nullslack> rxr, from the handbook, chapter 10 - "The mayor improvement"
11:54 < rxr> e
11:54 < rxr> re
12:00 < nullslack> re?
12:01 < rxr> re(turned) from a talk here ...
12:01 < rxr> but now I leave the uni - need food and sleep
12:01 < rxr> cu all
12:02 < nullslack> cu rxr
13:51 -!- CIA-8 [~CIA@to.je.spocco.com] has quit [Remote closed the connection]
13:53 < mnemoc> moin
14:08 -!- _martin_ [~martin@217.186.6.94] has quit [Read error: 104 (Connection reset by peer)]
14:11 < nullslack> moin mnemoc
14:13 < mnemoc> moin nullslack
14:15 < nullslack> mnemoc, can we mimic the way the kernel ,config displays cpu optimizations? it's more readable that way and less confusing. ;)
14:17 -!- CIA-9 [~CIA@to.je.spocco.com] has joined #t2
14:19 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
14:22 < mnemoc> nullslack: i didn't understood :( can you repeat please? :)
14:26 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: mnemoc
14:28 -!- Netsplit over, joins: mnemoc
14:30 < nullslack> mnemoc, i mean...when u configured the kernel, there are options like Celeron, i586, i686. in t2 those are missing...literally ;)
14:30 -!- kensai [~kensai@64.237.129.108] has joined #t2
14:31 < mnemoc> nullslack: oh
14:35 -!- kensai [~kensai@64.237.129.108] has quit ["Leaving"]
14:39 < nullslack> == 12/06/04 21:40:27 =[5]=> Aborted building package dhcp.
14:39 < nullslack> In file included from raw.c:62:
14:40 < nullslack> make[2]: *** [raw.o] Error 1
14:41 < nullslack> ROCK/src.dhcp.1102340401.18649.1694616223/dhcp-3.0pl2/includes/dhcpd.h:306: error: mode `byte' applied to inappropriate type
14:43 < nullslack> mnemoc, ;(
14:45 < mtr> nullslack: this regression is caused by gcc-3.4.3
14:45 < nullslack> mtr, so what should i do?
14:46 < mtr> there is patch in rock submaster, you will find it when google'ing for "submaster dhcp gcc343"
14:46 < mtr> I have not tried or tested it, but it should fix the compiler error
14:47 < mnemoc> nullslack: it still on top of Config
14:47 < nullslack> mnemoc, the patch?
14:47 < mnemoc> nullslack: the optimizations
14:49 < nullslack> mnemoc, yes...but there's no Celeron, i586, and i686...literally ;)
14:50 < nullslack> mtr, rock submaster?
14:50 < mnemoc> only rock use submaster :)
14:51 < mnemoc> nullslack: look into architecture/x86/config.in
14:51 < mtr> nullslack: you may pull the patch from the rock website, without submaster :)
14:52 < mnemoc> mtr: if you apply it into T2 add a comment on top telling how wrote it
14:52 < nullslack> mtr, ok...so how can i apply the patch?
14:52 < mnemoc> who*
14:52 < mnemoc> .patch files con $confdir are autoappliedd
14:52 < mnemoc> grab the patch on submaster
14:52 < mnemoc> patch -p0 it
14:52 < mnemoc> and paste here the output :)
14:53 < mtr> mnemoc: the author is already mentioned on top of the patch: Andreas V. Meier
14:53 < nullslack> mnemoc, i've looked into it...and technically it's there...just the literal options in the Config are missing ;)
14:53 < mnemoc> nullslack: add it :)
14:54 < mnemoc> nullslack: but add it completely
14:54 < mnemoc> nullslack: with all the conversions needed
14:56 < nullslack> mnemoc, i'll try...i have little knowledge about scripting ;(
14:56 < nullslack> mnemoc, patch the dhcp package?
15:04 < mnemoc> the patch will create a patchfile or edit .conf, but i guess create a patchfile
15:04 < mnemoc> considering t2's reorganization a patch -p0 will not apply
15:05 < mnemoc> cd enter dhcp confdir
15:05 < mnemoc> s/cd //
15:05 < mnemoc> and patch -p3 inside it
15:05 < mnemoc> but lsdiff will tell you what the patch touches
15:08 -!- kensai [~kensai@64.237.129.108] has joined #t2
15:11 < mnemoc> nullslack: ?
15:22 < nullslack> mnemoc, have u tried it?
15:22 < mnemoc> nope :)
15:22 < nullslack> --- package/base/dhcp/gcc343-fix.patch (revision 0)
15:22 < nullslack> +diff -ruN dhcp-3.0.2rc2/includes/dhcpd.h dhcp-3.0.2rc2-new/includes/dhcpd.h
15:23 < nullslack> +--- dhcp-3.0.2rc2/includes/dhcpd.h 2004-10-01 20:43:21.000000000 +0200
15:23 < nullslack> ++++ dhcp-3.0.2rc2-new/includes/dhcpd.h 2004-11-17 20:15:59.414035328 +0100
15:24 < nullslack> t2 is dhcp-3.0pl2
15:24 < mnemoc> update it :)
15:26 < mnemoc> have you tried to see if that gcc343-fix.patch applies to dhcp-3.0pl2?
15:28 < nullslack> brb...i'll try ;)
15:30 < mtr> I am sorry if my hint causes so much patch-confusion
15:31 -!- kensai [~kensai@64.237.129.108] has quit ["Leaving"]
15:32 < mnemoc> mtr: why sorry? nullslack wants to understand T2, these 'problems' are vital to do it
15:33 < mtr> nullslack, if you do not need a dhcp server package you may ignore the error and use dhcpcd as client
15:35 < mnemoc> mtr: do you know any good dhcpd?
15:36 -!- kensai [~kensai@64.237.129.108] has joined #t2
15:42 < mtr> mnemoc: (sorry for the delay, got a phone call)
15:43 < mtr> if i have to choose a dhcpd i would use the isc.org dhcpd ... ;-)
15:43 < mnemoc> :\
15:43 < mnemoc> i'm using ISC's, but i'm kind of alergic
15:44 < mtr> alergic ?
15:50 -!- kensai [~kensai@64.237.129.108] has quit ["Leaving"]
15:50 < mnemoc> the only fact of being using a product of ISC produce me skin problems
16:03 -!- _martin_ [~martin@brln-d9ba1eb8.pool.mediaWays.net] has joined #t2
16:06 < mtr> mnemoc: hmm, did I miss something about isc? /me trying to remember what I've read about them ...
16:06 < mnemoc> nothing really :)
16:06 < mtr> !> `server/dhcpd.conf' -> `/usr/doc/dhcp/dhcpd.conf'
16:06 < mtr> !> No 5-dhcp.log file.
16:06 < mtr> -> $root/var/adm/logs/5-dhcp.out -> 5-dhcp.err
16:06 < mtr> == 12/06/04 14:55:35 =[5]=> Aborted building package dhcp.
16:07 < mtr> hmm, 3.0.2rc2 build, but no logfile :(
16:08 < mtr> nullslack: did the patch apply to 3.0pl2?
16:11 < mnemoc> mtr: src.dhcp.*/ERROR-LOG
16:12 < sparc-kly> moin all
16:14 < mnemoc> moin sparc-kly
16:19 < jsaw> mtr: did you type any command (scripts/*) while dhcp was compiling?
16:20 < mtr> jsaw: no, I did not
16:22 < mtr> i have no idea, but scripts/Cleanup warns me about "Non-Standard Perm at ..." for nearly every directory
16:22 < jsaw> o_O
16:22 < jsaw> is $root/var/adm/logs writable?
16:23 < mtr> yes: drwxr-xr-x 2 root root 6744 2004-12-06 16:17 var/adm/logs
16:24 < jsaw> better check with "touch $root/var/adm/logs/test"
16:24 < mtr> touch works, too
16:26 < mtr> and after the compilation, there is the 5-dhcp.err file in var/adm/logs, and ERROR-LOG is also present
16:29 < CIA-9> mnemoc * r4963 /trunk/package/database/postgresql-autodoc/ (. postgresql-autodoc.conf postgresql-autodoc.desc): * Adds postgresql-autodoc, and utility to autodocument PostgreSQL databases
16:32 < mnemoc> mtr: i got that problem with old svn-s with created ewvery file with sticky bit enabled
16:34 < mtr> my svn --version: svn, Version 1.1.0 (Release Candidate 2)
16:35 < mnemoc> can you stat a directory Cleanup flag as 'non standard perms'?
16:48 -!- madtux [~mike@200.91.101.97] has joined #t2
16:48 < madtux> moin
16:52 < mnemoc> wb beachtux
16:52 < madtux> danke
17:36 < jsaw> heading home, cu later.
18:01 -!- madtux [~mike@200.91.101.97] has quit ["leaving"]
18:17 -!- madtux [~mike@200.91.101.97] has joined #t2
18:17 < madtux> re
18:22 < CIA-9> mnemoc * r4964 /trunk/package/database/postgresql/ (postgresql.conf postgresql.desc):
18:22 < CIA-9> * Updates postgresql (8.0.0beta4->8.0.0beta5)
18:22 < CIA-9> * Adds contrib to postgresql's docdir
18:46 < mtr> For anyone interested in the "No $xpkg.log file" error message:
18:46 < mtr> it was caused by an error during postmake evaluation
18:47 < mtr> dhcp.conf: cp -av doc/* client/dhclient.conf server/dhcpd.conf $docdir
18:48 < mnemoc> mtr: enabling XTRACE on config is very usefull to find that stuff
18:48 < mtr> as there is a doc/ja_JP.eucJP subdir, this is omitted, and cp returns 1
18:49 < mnemoc> cp -avR ?
18:50 < mtr> mnemoc: the original cmd was "cp -v ...", i changed it to "cp -av ..." which works now
18:50 < mnemoc> commit it :)
18:51 < mtr> yes, I would, but I do not have svn write access
18:52 < mtr> for now i send it to the ml
19:01 * madtux running selinux pre-commit build
19:02 < mtr> the dhcp doc files are japanes man pages, shouldn't they go into /usr/share/man/jp_JP ?
19:03 < mnemoc> mtr: move them on premake
19:03 < mnemoc> moveto the place where dhcp has it's man pages
19:04 < mnemoc> so that cp doc is kept clean
19:04 < mtr> yep, ok
19:47 < CIA-9> mnemoc * r4965 /trunk/package/x86/ (5 files in 5 dirs):
19:47 < CIA-9> Miguel Angel Bolanos
19:47 < CIA-9> * Updated acpid (1.0.3 > 1.0.4)
19:47 < CIA-9> * Updated bin86 (0.16.14 > 0.16.16)
19:47 < CIA-9> * Updated lilo (22.5.9 > 22.6.1)
19:47 < CIA-9> * Updated memtest86 (1.27 > 1.40)
19:47 < CIA-9> * Updated microcode_ctl (1.10 > 1.11)
19:49 < mnemoc> mtr: mike has mailed the updated for dhcp
19:49 -!- sparc-kly [~mubex@64.237.130.173] has quit [Read error: 110 (Connection timed out)]
19:50 < mnemoc> err... it was you!
19:50 < mnemoc> :)
19:50 < CIA-9> mnemoc * r4966 /trunk/package/x86/grub/ (gcc34.patch grub.desc):
19:50 < CIA-9> Miguel Angel Bolanos
19:50 < CIA-9> * Updated grub 0.94 > 0.95
19:54 < CIA-9> mnemoc * r4967 /trunk/package/extra/pcre/pcre.desc:
19:54 < CIA-9> Michael Tross
19:54 < CIA-9> * update pcre (4.3 -> 5.0)
19:56 < CIA-9> mnemoc * r4968 /trunk/package/network/dhcp/ (dhcp.conf dhcp.desc gcc343.patch):
19:56 < CIA-9> Michael Tross
19:56 < CIA-9> * update dhcp (3.0pl2 -> 3.0.2rc2)
19:56 < CIA-9> * add Andreas V. Meier's dhcp/gcc-3.4.3 patch
19:56 < CIA-9> * fix postmake cp of japanese man pages for dhcp
20:01 < madtux> mnemoc, thanks for the patches apply
20:01 < madtux> mtr, good work on dhcp
20:04 < madtux> mnemoc, ... you did not apply the part on my grub updated removing the gcc34 part for grub... if you don't do this grub build brake because this patch no longer needs to be applied
20:06 < madtux> err nevermind you did it :)
20:19 < mnemoc> :D
20:23 < madtux> ok i'm starting to get a headache.. i will standby selinux stuff for a little and do something else to get distracted
20:24 < mnemoc> what about porting router to pkgsel ? :)
20:24 < mnemoc> i'll install svk now
20:24 < mnemoc> handling many branches in sync with a far far far away server is sloooooow
20:25 -!- _martin_ [~martin@brln-d9ba1eb8.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20:27 < madtux> porting router to pkgsel.. let me make my target stable again then i will let you try to convince me :)
20:27 < mnemoc> ok :)
20:28 < madtux> good luck :)
20:35 < mnemoc> my first branch on svk will be t2-2610
20:35 < madtux> ... because ...
20:37 < mnemoc> uhm?
20:37 < madtux> why t2-2610 ?
20:39 < mnemoc> to keep version control of changes needed to build againt lx2.6.10?
20:39 < madtux> ah ok
21:18 -!- hannes_ [ucbiv@rzstud3.stud.uni-karlsruhe.de] has joined #t2
21:18 < hannes_> hi
21:49 -!- _martin_ [~martin@brln-d9ba1eb8.pool.mediaWays.net] has joined #t2
21:53 < mnemoc> hhttp://khason.biz/blog/2004/12/why-microsoft-can-blow-off-with-c.html
22:07 < madtux> *g*
22:15 < madtux> mike@selene:~/Projects> ls t2/patches/*.diff | wc -l
22:15 < madtux> 20
22:15 < madtux> mike@selene:~/Projects>
22:16 < madtux> mmm... tomorrow's patch submission will be fun
22:17 < hannes_> madtux: selinux patches? =)
22:17 < mnemoc> uhm
22:18 < mnemoc> madtux: can you put them where i could download them?
22:18 < madtux> hannes_, thats just updates, pending selinux patches and several fixes realted tot he updates
22:18 < madtux> and finally router taget fixes
22:19 < madtux> mnemoc, i will submit them tomorrow, besides this are not selinux patches, i'm not ginished withmy selinux stuff yet
22:19 < mnemoc> madtux: don't be horn, let me download them.... grab them from mail is a pain
22:19 < madtux> mnemoc, I'm NOT alan cox, i mean i'm not a patch machine
22:20 < madtux> madtux, i must test this updates before i actually submit them
22:21 < madtux> mnemoc, i'm working on having generic-minimal and bootdisks to build flawless
22:21 < mnemoc> which linux-headers?
22:21 < madtux> i'm not sure if i will suceed doing it with router yet
22:22 < madtux> mnemoc, i'm NOT touching linux-header untill 2.6.10-FINAL is out
22:22 < mnemoc> i got your point
22:22 < madtux> :)
22:23 < madtux> it should be out pretty so it makes no sence to go through all the pain making things work as we move to 2.6.9
22:23 < madtux> 2.6.10 will be much cleaner
22:23 < madtux> and besides our current linux-header is working so leave it that way for now :)
22:24 < mnemoc> it's doesn't work if you want them patches in the same way that your running kernel
22:24 < madtux> :)
22:28 < hannes_> gn8!
22:29 < mnemoc> n8 hannes_
22:37 < madtux> gute nacht
23:06 < CIA-9> mnemoc * r4969 /trunk/package/x86/syslinux/syslinux.desc:
23:06 < CIA-9> Miguel Bolanos
23:06 < CIA-9> * sync [V] of syslinux with it's file
23:06 < madtux> gracias
23:06 < rxr> moin
23:06 < madtux> good day rene
23:07 < rxr> hi madtux
23:08 < madtux> i asked mnemoc just about 1 minute ago if he knew where u were :)
23:08 < madtux> how are things?
23:17 -!- sparc-kly [~mubex@64.237.130.173] has joined #t2
23:19 < rxr> mnemoc: thanks for the patch apply work ;-)
23:19 < mnemoc> it was easy
23:19 < madtux> i promisse you much more fun tomorrow mnemoc :)
23:20 < rxr> hannes_: ?
23:20 < rxr> hi hannes_ ...
23:20 [Users #t2]
23:20 [ _martin_] [ CIA-9 ] [ jsaw ] [ mtr ] [ rxr ]
23:20 [ _Ragnar_] [ daja77 ] [ madtux] [ nullslack] [ sparc-kly]
23:20 [ af_ ] [ hannes_] [ mnemoc] [ nzg ] [ valentin ]
23:20 -!- Irssi: #t2: Total of 15 nicks [0 ops, 0 halfops, 0 voices, 15 normal]
23:26 < mnemoc> madtux: i hope good working fun, not broken fun :)
23:28 < madtux> i listened to ur inpatient $foo it would be broken .. but since i'm a carefull boy i will test my stuff before submitting :)
23:29 < madtux> if i*
23:29 < mnemoc> :)
23:29 < mnemoc> i'm just teasing you
23:29 < mnemoc> i'm not in a hurry :)
23:31 < mnemoc> svk mirror is damn slow!
23:32 < mnemoc> 2seconds per revision
23:39 < rxr> hm
23:40 < mnemoc> what that hm means?
23:40 < madtux> it means u are probably blaming it on poor svk when its really ur fault
23:40 < mnemoc> :(
23:41 < mnemoc> why my fault? :(
23:41 < madtux> because it has to be somebodies fault
23:41 < madtux> :P
23:41 < mnemoc> why not yours?
23:41 < madtux> because i'm not the one using svk
23:41 < madtux> :)
23:41 < mnemoc> :D
23:42 < madtux> btw do you think that you can have that svk package for t2 ready this week? :)
23:42 < mnemoc> i'll not package it
23:42 < madtux> oh come on!
23:42 < madtux> :)
23:43 < mnemoc> i wont add 43 perl-* packages for it
23:43 < mnemoc> sorry
23:43 < madtux> please?
23:43 < madtux> :)
23:43 < mnemoc> no!
23:43 < madtux> :(
23:43 < mnemoc> :)
23:44 < madtux> thats mean :(
23:44 < madtux> ah well i'm going home now
23:44 < madtux> have a good everyone
23:45 < mnemoc> see you later mike
23:45 < madtux> yeah :)
23:48 -!- madtux [~mike@200.91.101.97] has left #t2 ["Leaving"]
--- Log closed Tue Dec 07 00:00:20 2004