--- Log opened Wed Sep 29 00:00:57 2004
00:31 < rxr> re
00:31 < rxr> mnemoc: yes - the mozilla plugin and similar stuff can be per-user configured in the mozilla or simillar prefs.
00:35 < rxr> martin_: sure - there always has beeen the dog slow blackdown java - we also used it sometimes on PPC, although half the speed of comparable x86 java implementations ...
00:46 < mnemoc> rxr: do you know how?
00:55 < rxr> hm - in the preferences always has been some plugin path selector ...
00:55 < rxr> although I seldom used it ...
00:58 < rxr> btw. I have some beer in me - so if I commit s.th. righ tnow
00:58 < rxr> better double check it ...
00:58 < rxr> but I'll be in bed in some seconds anyway
01:00 < mnemoc> =)
01:01 < mnemoc> i couldn't find such stuff on firefox :\
01:01 < rxr> hm :-(
01:02 * rxr starting to crawl into bed - starting with toothcrushing ...
01:06 < rxr> hm - strange gnome-session heisenbug ...
01:06 < rxr> anyway error count decresed by one by doing nothing (except rebuilding a package ...)
01:10 < mnemoc> if i build ogg-vorbis, should i re-build mplayer?
01:28 < jsaw_> ldd /usr/bin/mplayer |grep ogg
01:28 < jsaw_> libogg.so.0 => /usr/lib/libogg.so.0 (0x40050000)
01:29 < jsaw_> (if ogg's ABI did not change, no.)
01:31 < mnemoc> i have no ogg currently, then i guess i shall rebuild
01:31 < jsaw_> yes
01:32 < mnemoc> there is any tool, life 'file' to see what decoders do i need for a file?
01:32 < jsaw_> no idea ...
01:32 < jsaw_> file's not enough?
01:33 < mnemoc> this time is was.... i'm just wondering :p
01:33 < mnemoc> Failed to open /dev/misc/rtc: Permission denied (mplayer should be setuid root or /dev/misc/rtc should be readable by the user.) <-- this shall be fixed on devfsd.conf?
01:37 < jsaw_> ???
01:38 < jsaw_> neither rtc is readable nor mplayer setsuid... still it works ...???
01:38 < mnemoc> yep
01:38 < jsaw_> here (TM)
01:39 < mnemoc> i get _lot_ of file not found and stuff when starting mplayer :p
01:39 < mnemoc> btw, this is the very first time that something different than a beep gets out of my laptop' speaker
01:40 < mnemoc> and this laptop is the first machine i have with audio device :p
01:42 < jsaw_> what about xine?
01:44 < mnemoc> ./scripts/Emerge-Pkg xine
01:45 < jsaw_> :)
01:45 < mnemoc> i'm a complete newbie :p
01:46 < mnemoc> i first tried xmms but i aborted because it wanted gtk+14
01:46 < mnemoc> then i tried mplayer
01:46 < mnemoc> now you 'show' me xine, and i'll test it :p
01:47 < jsaw_> there's also a GUI for mplayer: gmplayer (name looks like it's done with gtk...)
01:47 < mnemoc> if it's gtk2 there is no problem :p
01:48 < mnemoc> but i dislike the idea of having two versions of the same library on living on the same machine
01:49 < jsaw_> :( .cache says gtk+12 ...
01:49 < mnemoc> :\
09:19 < martin_> moin
09:28 < martin_> rxr: screen/BitchX have
09:29 < martin_> rxr: screen/BitchX have illegible your writing about java.
09:30 < martin_> rxr: there is blackdown Java in /usr/local. Wow. But only version 1.3.1, too
09:41 < martin_> rxr: There will only IBM Java 1.4.x for PowerPC. But I will not register me at IBM. Hmmm
10:51 -!- rxr_ [~rene@p213.54.200.254.tisdip.tiscali.de] has joined #t2
10:51 -!- Topic for #t2: t2 - 2.1-final 'tilll 30th October, 2004 (!?)
10:51 -!- Topic set by rxr [] [Mon Sep 27 16:45:14 2004]
10:51 [Users #t2]
10:51 [ jsaw_] [ martin_] [ mnemoc] [ rxr] [ rxr_] [ valentin]
10:51 -!- Irssi: #t2: Total of 6 nicks [0 ops, 0 halfops, 0 voices, 6 normal]
10:51 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
10:51 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup
10:51 -!- Irssi: Join to #t2 was synced in 15 secs
10:57 -!- rxr [~rene@p213.54.208.162.tisdip.tiscali.de] has quit [Nick collision from services.]
10:57 -!- You're now known as rxr
12:42 -!- martin_ is now known as martin
12:43 < martin> rxrÖ you read my commend about java?
12:48 < rxr> yes
12:48 < rxr> valentin: can you comment on Java (on PPC) you are our expert in that area ...
12:56 < rxr> we have 1286 packages ...
12:58 < rxr> what about two new repositories "textproc" (textprocessing) and printing?
12:58 < rxr> the first for things like libxml* libxslt, latex, ... the later for lprng, cups, ghostscript(?) ... ?
13:01 < rxr> I'm open to receive proposals how to sort the stuff currently haning in extra ...
13:01 < rxr> s/haning/hanging/ - too bad when the irssi urxvt is covered and I press enter without seeing what I typed ... ;-)
13:11 < valentin> martin: i can give you a copy of ibm's java
13:11 < valentin> but it is 3 times slower than the sun implementation :(
13:12 < valentin> oh, i think i am not alowed to copy to unregistered users :(
13:12 < valentin> so i cant give you...
13:13 < rxr> there is sun java von powerpc that is 3 times faster then the Sun one?
13:14 < rxr> ibm one even ...
13:17 < valentin> no, but the sun x86 java for linux is noticable faster
13:23 < rxr> it might be more the powerpc implementation that sucks majorly - isn't the IBM one faster on x86?
13:24 < rxr> valentin: any comments regarding the category comments above?
13:25 * rxr starting a new rebuild on the athlon
13:26 < rxr> having the athlon-xp back here as devel machine really rocks ;-)
13:29 < valentin> rxr: the two categories are ok, but perhaps there is a better name than textproc
13:29 < rxr> bsd's use textproc ...
13:30 * martin is away ...
13:30 < valentin> bye
13:30 < rxr> cu martin
13:30 < valentin> ok, then lets use textproc
13:30 < rxr> http://www.openbsd.org/cgi-bin/cvsweb/ports/textproc/
13:31 < rxr> they have xpdf in textproc ...
13:32 < rxr> hm - the openbsd ports are quite chaotic - the more I look
13:33 < rxr> acroread is in printing ... - where xpdf is in textproc ...
13:33 < rxr> print even
13:35 < valentin> hm, since xpdf is used by latex people only, this makes a bit sense
13:36 < rxr> xpdf? I use it for every pdf - nothing to do with latex
13:36 < valentin> oh - i read xdvi :)
13:36 < rxr> and worte xpdf ;-)
13:36 < valentin> i see
13:37 < rxr> wrote even ...
13:37 < rxr> if you have some spare minutes to move some more yet unsorted packages
13:37 < rxr> I'm puffed out moving the packages a bit ...
13:37 < valentin> which ones ?
13:37 < rxr> there are 3 or so named repos left + the extra stuff
13:37 < valentin> the ones in the "special" named dirs..
13:38 < rxr> special?
13:40 < jsaw_> what's about an "office" or "publishing" directory?
13:41 < rxr> I already thought about those, too ...
13:41 < rxr> we have cad - so maybe name the publishing dtp ...
13:43 < valentin> rxr: ack
13:43 < jsaw_> the question the other way round: where shall I put openoffice.org?
13:44 < rxr> I think a "office" repository woudl fit good, or are the better ideas?
13:46 < valentin> damn, how many kanji training tools do we have in t2 ?
13:46 < valentin> just found the 4th or 5th one ....
13:46 < valentin> manga freaks ...
13:46 < rxr> stuff like scribus are not really "office" in terms of spreadsheet and word processing - so that would fit better in a dtp rep.
13:46 < rxr> valentin: yeah ...
13:47 < rxr> put them in dict, I think?
13:47 < rxr> I moved some other into dict at least
13:47 < jsaw_> that's why I ask. scribus is more like publishing, but still a bit graphics... (damn, it's hard to sort this)
13:47 < jsaw_> I do not envy you :)
13:48 < valentin> rxr: oh - good
13:48 < rxr> in my view graphic is for stuff that does mostly graphic manipulation - so a dtp catefory for the stuff that is clearly for "publishing" fit
13:48 < rxr> and office for the other every day letter and sheet producing stuff ...
13:49 < valentin> rxr: did you do a huge commit just now ??
13:49 < rxr> perfect sorting is also not that important - we can re-move some of them later on, we currently just need to find out if the cetegory in favour of names sorting works out as expected ...
13:49 < rxr> jep
13:49 < rxr> oh - sorry - you are moving - damn
13:50 < jsaw_> but "publishing" has also it's problems: what's about latex? I use latex for professional publishing...
13:50 < rxr> I hope you do not get that many conflicts ...
13:50 < rxr> valentin: I finished now - so feel free to do your stuff now ;-)
13:50 < valentin> don't thinks there were conflicts
13:50 < rxr> good ;-)
13:51 < rxr> jsaw_: yes, as usual with the categories they are not a perfect sort - we should sort them into the category where it is used "normally" for
13:51 < jsaw_> anyway, when I have a compiled t2 I'll try work on OOo again and commiit office/openoffice.org
13:51 < rxr> please do not name it .org - better just openoffice
13:51 < jsaw_> I have to.
13:52 < rxr> what?
13:52 < jsaw_> there's a problem concerning law, openoffice is a registered trade mark.
13:52 < valentin> huh ?
13:52 < jsaw_> but it's not registered by the OOo people.
13:52 < rxr> ic - then better no . i the name - would openoffice-org work out? or maybe just ooo or OOo ... ,-~)
13:53 < jsaw_> ooo is good, imo.
13:54 < rxr> ok ;-)
13:54 < rxr> jsaw_: how is your t2 build going?
13:55 < jsaw_> just did a cleanup -full, the binutils update caused problems (usually I just remove [012]-*, but I was not that far...).
13:55 < valentin> where should i put libowfat ?
13:55 < jsaw_> Guess the new change, to not rebuild the tools will solve this in the future.
13:55 < rxr> hm
13:56 < rxr> valentin: did you got updates for your SDSL line? svn is even faster then usual today ...
13:56 < rxr> (hm - or my slow line fot faster ;-)
13:56 < valentin> well, a year ago they said it's 256kbit/s and it was ~500
13:57 < jsaw_> btw, the net-tools problem was solved by kernel header patch (the if.h thing iirc)
13:57 < valentin> now they say its 512 and who knows
13:57 < valentin> we have much crap packages imo
13:57 < rxr> yes - that is why I already removed some of them ..
13:57 < rxr> what did you just found?
13:57 < valentin> like onethousand bad spam filters, iptables frontends etz
13:58 < valentin> just look into package/sirkull
13:58 < rxr> move to strange ones into contrib ...
13:58 < valentin> i'll sort them away, but as soon as they do not compile anymore i would vote for deletion
13:58 < rxr> that "marks" them as "to be get rid and non-vital" ...
14:00 < valentin> libowfat ?
14:00 < rxr> valentin: I take a look in a sec
14:01 < valentin> ok
14:01 < rxr> just had to fix Check-System
14:01 < rxr> you saw that is a fefe lib?
14:02 < valentin> jep
14:02 < valentin> but i want to know where to put it
14:02 < rxr> we should not trow it away - maybe we need a rep. for such "support libs" - for now I moved most of those into develop
14:02 < valentin> i did not want to throw it away - i wanted to know where to move it
14:02 < valentin> so -> develop
14:03 < rxr> yes yes - I was not sure if it falls into your "ugly packages" category ...
14:03 < rxr> I just enabled the e16 packages for the desktop target ...
14:04 < rxr> me off to aldi - gettig new "direkt gerpessten blutorangen saft" ...
14:04 < valentin> ok, cu
14:26 < rxr> re
14:27 < rxr> cool - they still had the suice ... yesterday it as so empty I thought that they'll not have in in autumn any longer
14:27 < rxr> btw, something more important:
14:28 < rxr> I just noticed a paper on hour house door inviting to the Friedenau "Talks" (Gespraeche) done by the SPD I think
14:29 < rxr> the paper also listed the next dates and at December the 17th the meeting will be (os is at least planed) with Brigitte Zypris - maybe we should attent and bug them regarding software patents ...
14:58 -!- mnemoc_ [~amery@200.75.27.54] has joined #t2
14:59 < rxr> hi mnemoc_
14:59 < rxr> btw subversion 1.1.0 got released
14:59 < rxr> shoudl I just bump our subversion package to that new stable series?
15:01 < rxr> ouhm - there is so much fixed, better, new, faster, ... I think I should just update it now, right?
15:03 < rxr> *lol*: "I thought this could have
15:03 < rxr> been a specific setting for subversion.
15:03 < rxr> Windows-to-Windows communication is really **slow** without disabling
15:03 < rxr> Nagle algorithm.
15:03 < rxr> As I said somewhere else, it is faster to interpose a Linux Apache proxy between
15:03 < rxr> Windows machine rather than making direct communication.""
15:11 -!- mnemoc [~amery@200.75.27.58] has quit [Read error: 232 (Connection reset by peer)]
15:44 < rxr> mnemoc_: I just update monot to 1.0.2 in T2 ..
16:22 < rxr> who evented this code line in Build-Pkg:
16:22 < rxr> makeopt="$makeopt $x=\${$x/${root//\//\\/}/}"
16:29 < rxr> hm - seem to present quite some time already ...
16:29 < rxr> damn
16:29 < rxr> and svn could have a nicer "show me which revisions touched this line" command ...
16:40 < rxr> noone here anymore?
16:40 < rxr> is modutils also installing outside basedir in rock linux?
16:41 < rxr> I see no reason why it should be fixed there - all the code lines that matter are "functional equally" ...
16:43 < rxr> I modified Build-Pkg - if modutils does not install outside $basedir anymore I'll restart my build if the "quite big change" does cause other packages to break ...
16:47 < rxr> me restarting the build - my Build-Pkg fix, fixed modutils ...
16:47 < rxr> ^- /
16:49 < rxr> ah - I guess s.th. - the old bash2 might not evaluated it correctly - so the code was unfunctional
16:49 < rxr> and not it evaluates and produces crap ...
17:09 < mnemoc_> rxr: an update to mono implies update to gtk# and mod_mono too
17:10 < mnemoc_> btw, mono>=1 is not 100% friendly to apps built with mono<1
17:10 < rxr> ok - wanted to do this when my build reaches them - but I can do them right now, too
17:10 < mnemoc_> moin!
17:10 < rxr> moin ! mnemoc_
17:10 * mnemoc_ _must_ find time to maintain his damn packages!
17:11 < rxr> just use T2 ;-)
17:11 < rxr> as you see I already update them for you ;-)
17:12 < rxr> and when valentins new C++ .desc parsing and download stuff finsihes we will run an fully automated update finder and regression tester ;-)
17:13 < mnemoc_> i want to improve 2.0 download script to work with both structures, with symlinks
17:13 < mnemoc_> then i will move to t2
17:17 < rxr> cool ;-)
17:30 < rxr> so - me preparing some lunch and such - idling a bit in the background ...
17:47 * rxr now slowing down the ref. build by watching a bit tv ;-)
17:52 < rxr> hm the http://www.go-mono.org site is quite broken
17:52 < rxr> search defect and so is every 4th link ...
18:09 < mnemoc_> http://www.go-mono.com/archive/1.0.2
18:39 < rxr> yes - I already udpate all - I just wanted to browse the site a bit ...
18:40 < mnemoc_> =)
18:48 < rxr> I think I'm going to commit this after a few more tests:
18:48 < rxr> --- Build-Pkg (revision 4107)
18:48 < rxr> +++ Build-Pkg (working copy)
18:48 < rxr> @@ -370,27 +370,18 @@
18:48 < rxr> flistrfilter="ldconfig\..*: .*|.*: /var/adm/.*|.*/pkgconfig/.*\.pc"
18:48 < rxr> flistdel="etc/ld.so.cache|var/tmp/.*|usr/tmp/.*|var/adm/logs/.*|.*\\.old"
18:48 < rxr>
18:48 < rxr> +makeopt='prefix=$root/$prefix CC="$CC" CXX="$CXX"'
18:48 < rxr> if [ $stagelevel -le 1 ]
18:48 < rxr> then
18:48 < rxr> - makeopt='CC="$CC" CXX="$CXX" CC_FOR_BUILD="$BUILDCC"'
18:48 < rxr> + makeopt="$makeopt"' CC_FOR_BUILD="$BUILDCC"'
18:48 < rxr> makeopt="$makeopt"' BUILDCC="$BUILDCC" BUILD_CC="$BUILD_CC"'
18:48 < rxr> makeopt="$makeopt"' HOSTCC="$HOSTCC" HOST_CC="$HOST_CC"'
18:48 < rxr> makeopt="$makeopt"' STRIP="$STRIP" AR="$AR" LD="$LD"'
18:48 < rxr> makeopt="$makeopt"' RANLIB="$RANLIB" NM="$NM"'
18:48 < rxr> -
18:48 < rxr> - makeinstopt="$makeopt"' prefix=$root/$prefix install'
18:48 < rxr> -
18:48 < rxr> - for x in prefix bindir sbindir libdir datadir \
18:48 < rxr> - infodir mandir sysconfdir localstatedir \
18:48 < rxr> - includedir
18:48 < rxr> - do
18:48 < rxr> - makeopt="$makeopt $x=\${$x/${root//\//\\/}/}"
18:48 < rxr> - done
18:48 < rxr> else
18:48 < rxr> - makeopt='prefix=/$prefix CC="$CC" CXX="$CXX"'
18:48 < rxr> - makeinstopt='prefix=/$prefix CC="$CC" CXX="$CXX" install'
18:48 < rxr> flistdel="$flistdel|`echo $base | sed s,^/,,`/.*"
18:48 < rxr> fi
18:48 < rxr> +makeinstopt="$makeopt"' install'
18:48 < rxr>
18:48 < rxr> prepare="" ; hook_add prepare 5 'eval "$prepare"'
18:48 < rxr> prepatch="" ; hook_add prepatch 5 'eval "$prepatch"'
18:48 < rxr> (please review ;-)
18:51 < mnemoc_> you may loose flexibility only giving prefix to $MAKE
18:52 < rxr> it still goes to both variables
18:53 < rxr> I mainly removed the for loog which substitutes the $root (/usr/src/t2-trunk/build/.../ from the variables
18:53 < rxr> this breaks e.g. modutils, because it then installs into /sbin and freinds due to the removed $root
18:54 < rxr> I have no idea why the rock people have not yet noticed this - and in fact I have no idea why $root is stripped at that place there
18:54 < rxr> it can only break things - what should it fix?
18:54 < rxr> the remaining makeopt and makeinstopt content will be the same - I only factored the common parts out ...
18:54 < mnemoc_> remember pseudo native crap.... RL will always be built in chroot
18:55 < mnemoc_> then $root is senseless on stage>=3
18:55 < rxr> yes - this code was only used for the first cross stages - and there it matters - and $root got stripped away
18:55 < rxr> I see not a single reason for this - except that it breaks makefiles that use that variables and then install into the host system ...
18:56 < rxr> so far my build goes just normal - no regressions and it fixes modutils ...
18:56 < mnemoc_> =)
18:56 < rxr> which fomerly produced "files outside basedir" due to that ...
18:57 < rxr> I wonder why it does not happen in RL, maybe the outside basedir check is broken there again?
18:57 < rxr> maybe it broken and noone fixes it ...
18:57 < mnemoc_> be carefull with hardcoded locations inside binaries referencing your $root
18:57 < rxr> yes - I know that - but ROCK makes no attempt implementing that
18:57 < rxr> we need to implement this later - but that is what this DESTDIR often is for
18:57 < mnemoc_> yeah
18:58 < rxr> in ROCK the DESTDIR is never used - I guess when clifford first hacked his scripts in 1998 he did not noticed it
18:58 < rxr> for 2.1 I want to keep this as it is - to not unbreak so much code and hardcoded assumptions for 3.0 with cross all as user we need to do this right of course ...
18:59 < rxr> btw, I noticed a cross compiled bash does not work fully on sparc64 (when cross compiled from sparc(32) - my wild guess is that it is due to false archtest* content for sparc ...
19:00 < rxr> I wonder why the archtest.c file got removed in the past - when there is so much crap in misc/archive in ROCK ... :-()
19:00 < rxr> anyway - we will have some fun w/ cross buliding all when 3.0 times come ;-)
19:01 < mnemoc_> 3.0 is a good concept
19:01 < mnemoc_> RL stinks because that
19:01 < mnemoc_> 2.1 is too contaminated
19:01 < rxr> yep
19:22 < rxr> committed
19:42 -!- martin [~martin@brln-d9ba00b1.pool.mediaWays.net] has quit [Read error: 238 (Connection timed out)]
21:02 < rxr> so - my latest rebuild - even with my Build-Pkg cleanup - looks very good ...
21:02 < rxr> crunching in stage5 now ...
21:02 < mnemoc_> :D
21:12 -!- martin [~martin@brln-d9ba275f.pool.mediaWays.net] has joined #t2
21:12 -!- mnemoc_ is now known as mnemoc
21:12 < rxr> hi martin
--- Log closed Thu Sep 30 00:00:11 2004