--- Log opened Fri Oct 08 00:00:03 2004
00:22 < CIA-9> mnemoc * r4219 /trunk/package/develop/ (4 files in 4 dirs): * Mono, mod_mono, mono-xsp and mono-doc updated to 1.0.2
01:44 < jsaw> re
01:44 < jsaw> ping mnemoc
01:45 < mnemoc> jsaw:
01:46 < jsaw> have a look at my patch for the pkgsel improvement!
01:46 < jsaw> (sent to ml)
01:46 < mnemoc> ok
01:48 < mnemoc> jsaw: X foo will only enable package foo? and repos will need /?
01:48 < jsaw> yes
01:48 * mnemoc reading the patch
01:49 < mnemoc> is it tested?
01:49 < jsaw> I tried the given example
01:50 < jsaw> and the respective packages (check via egrep "perl|python") have correct X/O.
01:51 < jsaw> however I did not try the ! pattern, wait...
01:51 < mnemoc> commit it :p i guess my target is the only which does it's own pkgsel parsing
01:51 < mnemoc> ! pattern ?
01:54 < jsaw> ohoh, that is flawed anyway...
01:54 < jsaw> reading that part, sth. like "O perl/* !perl" should be possible
01:54 < mnemoc> oh
01:55 < jsaw> but it obviously does not work with my patch :(
01:55 < mnemoc> jsaw: i wanted to have pkgsel parser as a function, what do you think?
01:55 < jsaw> yes. much better, therefore this is only sent to the list as example and not committed.
01:56 < mnemoc> i know, CIA-9 has said nothing :p
01:57 < jsaw> so how should the pkgsel function look like?
01:58 < mnemoc> stdin a rule, stdout awk code
01:58 < mnemoc> or the rule as two arguments
01:59 < mnemoc> while read action pattern X; do
01:59 < mnemoc> pkgsel_parse "$action" "$pattern"
02:00 < mnemoc> done < $pkgsel_file >> $awk_file
02:00 < mnemoc> comments?
02:01 < jsaw> ok, so while ... done < <( if [ "$*" ]; then echo "$*"; else cat; fi)
02:02 * mnemoc got lost with the usage of $* there
02:02 < jsaw> either take args from stdin or argument list
02:03 < mnemoc> hehe
02:03 < mnemoc> nice trick
02:03 < rxr> re
02:04 < mnemoc> rxr: just on time
02:05 < mnemoc> rxr: jsaw is proposing a fix to pkgsel parser to fix the problem i mailed
02:05 < jsaw> hehe
02:05 < mnemoc> s/a fix/an improvement/1
02:05 < rxr> yelect perl by "perl perl" ;-)
02:05 < rxr> repository and package match ... :-(
02:05 < rxr> I know it sucks ...
02:05 < jsaw> read the mail first
02:05 < rxr> the 3.0 dependency node selection will be cooler ...
02:05 < jsaw> I mean, the patch
02:05 * rxr just scrolling back
02:06 < mnemoc> rxr: targets need something fast as pkgsel, users need something cool as dependency node selection
02:06 < rxr> valentin and I have been @ some cafe listening to www.falk-huebnder.de Jazz ;-)
02:06 < mnemoc> jazz is good for relaxing
02:07 < rxr> no targets also want node selections and just built that dependencies - at least destop and server wants it ...
02:07 < rxr> and C++ dependencies will be fast ...
02:07 < rxr> 4s of so on a 800Mhz iBook to scan all files (most of the time) and generate the graph in emory
02:07 < rxr> just the bash awk implementation sucks majorly ...
02:07 < rxr> and you also do not need to run it on each keypress in the config ...
02:08 < rxr> valentin: minimallist setup shoudl be fixed to inject Reply-To ...
02:10 < rxr> no mailman because it is overfeatures, C code run as cgi (urgs) and does not fit into our securtiy policies including jailing configuration ...
02:10 < rxr> we will never run mailman on our servers - sorry
02:11 < rxr> the pkgset is intended to be able to selct by repository ...
02:11 * rxr now taking a look into the mail client
02:20 < jsaw> ok, but what's about fixing/improving pkgsel right now (I mean before C++ works)?
02:21 < mnemoc> C++ = 3.0, and we want a good 2.1, doesn't we?
02:22 * mnemoc thinks it would be nice to have a CIA monitoring the Mailing list :p
02:25 < rxr> yes - sure - I just answer the mails
02:25 < rxr> mnemoc: ?
02:25 < rxr> you are quite irc centric these day, aren't you?
02:26 < mnemoc> :D
02:26 * mnemoc was thinking in a CIA for submaster too :p
02:32 < CIA-9> mnemoc * r4220 /trunk/package/mnemoc/runit/parse-config: * runit, fix a pkgcheck usage on parse-config
02:33 < mnemoc> rxr: what happened to svgalib and linux-src?
02:33 < rxr> mnemoc: your r4219 comment is invalid
02:33 < rxr> you just implanted the checksums - no version update ...
02:33 < mnemoc> uhm? ... let me see
02:34 < mnemoc> i forgot checksum,
02:35 < rxr> yes - correct! You even stripped the checksum away! - Me was blinded ...
02:35 < rxr> :-(
02:35 < mnemoc> i saw mono on 1.0.1, do i need glases
02:36 < mnemoc> 1.0.1 -> 1.0.2 should have been the update
02:36 < rxr> mnemoc: would you please revert that changset?
02:36 < rxr> merge -r4219:4218 ... ;-)
02:37 < mnemoc> oh
02:37 < mnemoc> it was on 1.0.2 o_O
02:39 < mnemoc> i guess i need to commit after merge
02:41 < CIA-9> mnemoc * r4221 /trunk/package/develop/ (4 files in 4 dirs): * revert 4219, mono was _already_ updated sorry.
02:42 < rxr> yes - merge does modify the local repository only ...
02:42 < rxr> commit needed ...
02:42 < rxr> ok - all pkgsel questions answered?
02:43 < mnemoc> i dislike X perl perl because it adds dependency to repo's name
02:43 < rxr> yes - it was a hack / work around ..
02:43 < rxr> I like jsaw's idea
02:43 < rxr> and propose to converr the config.in's to use the same pkgsel ...
02:43 < mnemoc> but the parser gets more complex, then a function is needed
02:44 < mnemoc> "to use the same pkgsel"? wha do you mean?
02:44 < rxr> config.in:
02:44 < rxr> pkgsel X base/*
02:44 < rxr> pkgset X perl
02:44 < rxr> ...
02:44 < rxr> instead of the manual sed's
02:45 < mnemoc> rxr: i have a pkgsel on my target
02:45 < mnemoc> target/*/pkgsel maybe be cleaner
02:46 < mnemoc> if you take a look into target/mnemosyne/ i have a add-on structure based on pkgsel files
02:46 < mnemoc> that's why i wanted a function on scripts/config.func or like
02:47 < rxr> yeah - looks nice
02:47 < mnemoc> add-ons can be .all (forced) or .ask (optional)
02:47 < rxr> ok - so use a pkgsel per target in 2.1?
02:47 < mnemoc> if enabled it can add config and enable a dir
02:48 < mnemoc> rxr: yes
02:48 < mnemoc> rxr: but config.in will remain working, so no damage is produced
02:48 < rxr> ok - let's better exchange patches before screwing up trunk ..
02:48 < rxr> yep
02:49 < rxr> mnemoc: could you review jsaw's patch and convert one target to pkgsel as proof of concept and send it as patch for review before commit?
02:49 < mnemoc> jsaw was not completly satisfied with his patch
02:50 < rxr> ouhm - then maybe improve it or so
02:50 < rxr> I can take a look after sleep ...
02:50 < jsaw> I'm not 100% sure about the ! logic.
02:51 < jsaw> otherwise it seems fine. I'll send a patch now that adds a "pkgsel_parse" function to scripts/functions in a couple of seconds
02:51 < rxr> it is 2:50 here, I have alcohol in my blood and do not want to shift my sleep cycle too much for the next get-up-early university week ...
02:51 < mnemoc> i trust jsaw's pkgsel_parse
02:51 < jsaw> alcohol in your blood or some rest of blood in ethanol?
02:51 < rxr> ouch - svn up on my iBook yields more lines of updates files then urxvt scrollback :-(
02:51 < mnemoc> hehe
02:52 < rxr> jsaw: ok - more likely the later ...
02:52 < jsaw> :))
02:52 < mnemoc> :D
02:58 < CIA-9> rene * r4222 /trunk/package/ (24 files in 24 dirs): * more rene sorting
03:00 < CIA-9> rene * r4223 /trunk/package/ (10 files in 10 dirs): * and yet more rene sorting
03:03 < CIA-9> rene * r4224 /trunk/package/ (filesystem/gtoaster/ rene/): * moved gtoaster into filesystem (where xcdroast already resists)
03:03 < rxr> so - only mnemoc as named repository is left?
03:04 < rxr> http://cia.navi.cx/
03:04 < rxr> ^- t2 at gcc level ;-)
03:05 < rxr> oh - ok - well this is already for today - so no suprise ..
03:06 < rxr> so - me in bed - n8 all
03:06 < rxr> cu in the morning ;-)
03:06 < jsaw> sleep well
03:10 < rxr> ah - one more thought about scrollback mnemoc
03:10 < rxr> linux*-src are removed to save disc space and build time and such
03:10 < rxr> and to make 3rd party kernel modules of packages saner
03:11 < rxr> I'm on the way to implement this for the linux packags
03:11 < rxr> at some point of the linux package (likely at the end) postlinux.conf files in other packags are searched (like postsysfiles)
03:12 < rxr> it will iterator over the found ones, and build those packages with the postlinuc.conf files as "code"
03:13 < rxr> this way we get the kernel modules for all kernels (no only some artificial "default" one), all modules of "enabled" packages are built everytime you update your kernel without manually rebuilding those packages
03:13 < rxr> plus some other minor benefits here and there ...
03:13 < rxr> maybe you can dump some thoughts about that here while I sleep ;-)
03:13 < rxr> cu
03:28 < mnemoc> re
03:41 < CIA-9> jsaw * r4225 /trunk/package/office/: * create package/office directory
03:44 < mnemoc> jsaw: openoffice?
03:47 < jsaw> still takes some work but, yes.
03:48 < mnemoc> can be built to use gcj for plugins?
03:58 < jsaw> hmmm. good q. I usually disabled java.
03:59 < mnemoc> as sun-java is very restricted i can't understand why so many open-developers love it :\
04:03 < jsaw> I never liked java at all...
04:04 -!- blindcod1r [~blindcode@pD958F04C.dip.t-dialin.net] has joined #t2
04:15 < jsaw> me off now, cu
04:16 -!- blindcoder [~blindcode@pD958FA2C.dip.t-dialin.net] has quit [Read error: 238 (Connection timed out)]
04:17 -!- mnemoc_ [~amery@200.75.27.56] has joined #t2
04:24 -!- mnemoc [~amery@200.75.27.40] has quit [Read error: 110 (Connection timed out)]
04:31 -!- mnemoc_ is now known as mnemoc
05:14 < mnemoc> jsaw: here?
08:18 -!- martin_ is now known as martin
08:27 -!- martin [~martin@brln-d9ba08fc.pool.mediaWays.net] has quit ["BitchX: the ONLY three day cure!"]
08:35 -!- martin [~martin@brln-d9ba08fc.pool.mediaWays.net] has joined #t2
09:23 -!- blindcod1r is now known as blindcoder
09:23 < blindcoder> moin
09:43 -!- Ragnar0k_ [loki@66-146-166-62.skyriver.net] has joined #t2
09:43 -!- Ragnar0k [loki@66-146-166-62.skyriver.net] has quit [Read error: 104 (Connection reset by peer)]
10:56 -!- rxr_ [~rene@p213.54.219.162.tisdip.tiscali.de] has joined #t2
10:56 -!- Topic for #t2: T2, a preleminary name. Submit cool naming suggestions (without linux) here ;-)
10:56 -!- Topic set by rxr [] [Wed Oct 6 17:56:16 2004]
10:56 [Users #t2]
10:56 [ blindcoder] [ daja77] [ martin] [ Ragnar0k_] [ rxr_ ]
10:56 [ CIA-9 ] [ jsaw ] [ mnemoc] [ rxr ] [ valentin]
10:56 -!- Irssi: #t2: Total of 10 nicks [0 ops, 0 halfops, 0 voices, 10 normal]
10:56 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
10:56 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup
10:56 -!- Irssi: Join to #t2 was synced in 16 secs
11:00 -!- rxr [~rene@p213.54.208.246.tisdip.tiscali.de] has quit [Nick collision from services.]
11:01 -!- You're now known as rxr
11:02 < rxr> re
11:03 < rxr> mnemoc: /me also hates java like a plague
11:19 < valentin> mnenoc: still here ?
12:48 < CIA-9> rene * r4226 /trunk/package/ (mnemoc/twoftpd/ network/twoftpd/): * moved mnemoc/twoftpd -> network/twoftpd
12:52 < CIA-9> rene * r4227 /trunk/package/network/tftp-hpa/tftp-hpa.desc: * updated tftp-hpa (0.36 -> 0.40)
14:14 < rxr> hm - damn s.th. happend to the gsmp.tfh-berlin.de box :-(
14:22 < CIA-9> rene * r4228 /trunk/package/base/coreutils/ (coreutils.conf disable-duplicates.patch):
14:22 < CIA-9> * revisited the coreutils/disable-duplicates.patch to have a higher
14:22 < CIA-9> chance to work for the next upstream release
14:22 < CIA-9> * moved head and tail from /usr/bin to /bin (I saw notes in the hotplug
14:22 < CIA-9> source, complaining distributions should have it during boot ;-)
14:36 -!- daja77 [[zl8YXDCa0@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
14:39 < mnemoc> rehi
14:41 -!- daja77 [[A+Yh4dZD0@odoaker.hrz.tu-chemnitz.de] has joined #t2
14:43 < rxr> hi mnemoc and daja77
14:57 < mnemoc> hi rxr and daja-rejoin
15:02 < CIA-9> rene * r4229 /trunk/package/x11/xorg/xorg.cache:
15:02 < CIA-9> * fixed the xorg dependencies to not include xaw3d (sorry - I added
15:02 < CIA-9> that noise :-()
15:06 < rxr> ok - I should do some math exercise soon
15:07 < mnemoc> onebase.exactcode.de?
15:07 < rxr> when it is already a registered trademark and such we should not use it at all ...
15:08 < mnemoc> :(
15:08 < rxr> and for the community and PR reason we would spend the exta money and register a domain anyway - no need to promote our extracode.de with "t2"
15:09 < mnemoc> ack
15:09 < rxr> I mean we want to establish a community "and not redo the rock cathedral"
15:09 < mnemoc> hehe
15:12 < mnemoc> rxr: what happened to svgalib and linux-src?
15:12 < CIA-9> mnemoc * r4230 /trunk/target/mnemosyne/pkgsel/ (10 files in 4 dirs): * Mnemosyne target, fix plugins for #t2
15:12 < rxr> do you really need svgalib? *g*
15:13 < rxr> I explained the linux-src removal last night ;-)
15:13 < mnemoc> :(
15:13 < rxr> readd svgalib if you really really need it *g*
15:14 < mnemoc> svgalib is the way to have graphics browsers (links -g) without X or framebuffer
15:15 < rxr> but framebuffer is much more saner and leightweight, stable, .. then the svglib hack ...
15:15 < rxr> but readd it if you need it :-(
15:16 < rxr> I search for the r of the remocal
15:16 < mnemoc> ok, i'll move my consoles to framebuffer :|
15:17 < mnemoc> brb
15:18 < rxr> hey - when you need it readd it!
15:20 < rxr> hm - I just can not find the removal revision
15:21 < rxr> regarding the linux*-src packages I want to build kernel modules inside the kernel pacakge itself - so we waste not that much disk space with the soure package ...
15:25 < rxr> oh - coll the svn people fixed error messages
15:25 < rxr> svn 0.30 gave:
15:25 < rxr> svn: svn_io_make_dir_recursively: error making 't2-trunk'
15:25 < rxr> the newer ones yield:
15:25 < rxr> svn: Can't create directory 't2-trunk': Permission denied
15:25 < rxr> ;-)
15:33 < rxr> do we need so patch subversio for sm?
15:34 < rxr> s/so/to/ ...
15:35 -!- daja77 [[A+Yh4dZD0@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
15:35 -!- daja77 [[O8R92zJQK@odoaker.hrz.tu-chemnitz.de] has joined #t2
15:41 < rxr> oops - sed just dumped core ..
15:42 < rxr> some old version on the rs6k ...
15:48 < CIA-9> rene * r4231 /trunk/package/develop/subversion/ (subversion.desc svn_sm.patch): * updated subversion (1.0.8 -> 1.1.0)
15:57 -!- daja77 [[O8R92zJQK@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
15:57 < rxr> warning - monster commit running (sed'ing my mail address ...)
15:58 < mnemoc> re
16:00 < mnemoc> rxr: is it really desired to add all third party kernel modules into linux package? what about local rebuilds?
16:02 -!- daja77 [[lp0hr3KnN@odoaker.hrz.tu-chemnitz.de] has joined #t2
16:02 < rxr> building the modules within the linux modules as several advantages
16:02 < rxr> - you get the modules for all kernel you have
16:03 < mnemoc> but sevaral disadvantages too
16:03 < rxr> - you get the modules when you update a kernel (no manual rebuilds of all the other kernel dependend packages
16:03 < CIA-9> rene * r4232 /trunk/package/ (384 files in 346 dirs):
16:03 < CIA-9> * substituted (hopefully) all my mail addresses to the new one at
16:03 < CIA-9> exactcode
16:03 < rxr> - multiple hundred MB space saving for every kernel you touch ...
16:04 < rxr> - no need for those mol linux-mol split packages ...
16:04 < rxr> mnemoc: what are the disadvantages you see?
16:04 * mnemoc thinking...
16:05 < mnemoc> - the absense of linux-src make you loose t2's patches when you want to rebuild your own customed kernel
16:05 < rxr> the major one is that potentially crappy 3rd party modules might abort a kernel build after 1hour ++ ...
16:05 < mnemoc> yes
16:06 < rxr> we could install the patchfiles with the binary linuxXY package ...
16:06 < mnemoc> what about ant (or whatever the name of the win32 driver emulator is called)
16:06 < rxr> another disadvantage is that the implementation with the current monolythic scripts is a bit tricky ...
16:06 < mnemoc> or winmodems
16:07 < rxr> ant? I know ndiswrapper ...
16:07 < mnemoc> whatever.... win32 dirver emulator :p
16:07 < rxr> what points do you have in mind regarding winmodems? - which packages is that?
16:07 < mnemoc> i have one :p
16:07 < rxr> just build it within the associated linuxXY packages as other 3rd party moduels
16:07 < rxr> the pseudo code is like this:
16:08 < rxr> (at the end of linux24.conf):
16:08 < mnemoc> also nvidia or stuff like ktcpvs
16:08 < rxr> for earch in package/*/*/postlinux.conf
16:08 < rxr> if pkgenabled $each
16:08 < rxr> build_pkg $each
16:08 < rxr> fi
16:08 < rxr> done
16:09 < rxr> I'm already working on it - should be ready within some days ...
16:09 < mnemoc> problem: $each will be built twice
16:09 < rxr> well - yes/no
16:09 < rxr> for some packages this is cool
16:09 < rxr> e.g. mol and cloop
16:09 < mnemoc> t2 needs 'extensions'
16:09 < mnemoc> but not every third party kernel module is an extension of the kernel
16:10 < blindcoder> rehi
16:10 < mnemoc> blindy!
16:10 < blindcoder> mnemoc!
16:10 < rxr> both packages have normal stuff (in mol a huge user space application) and with cloop user-space creation programs - there the normal package ships the user-space apps - and the postlinux.conf will build the kernel modules
16:10 < rxr> hi blindcoder
16:10 < rxr> for the other packages (like alsa-driver) the package when build normally is just empty (later we might like to optimize this and flag them as virtual - but that is just cosmetic)
16:10 < mnemoc> rxr: same extension don't have anything to install
16:11 < rxr> what "extensions" do you have in mind?
16:11 < mnemoc> vserver is an extension of kernel
16:11 < mnemoc> but i had to pack vserver-utils inside vserver to had something to install
16:12 < mnemoc> an extension can't live without the extendee
16:14 < rxr> ok - in what does vserver exactly need to do?
16:14 < rxr> just patch the kernel - or anything other special?
16:15 < mnemoc> patch the kernel and enable the 'feature'
16:15 < mnemoc> (this is not related to linux-src)
16:15 < mnemoc> third party kernel modules, which are not extensions are related to linux-src
16:15 < mnemoc> and local rebuilds are related to linux-src too
16:16 < blindcoder> damn, I'm running low on USB Ports
16:17 < rxr> blindcoder: plug in a hub;-)
16:17 < mnemoc> in my design for evilbk (my buildkit, not a fork, built from scratch) i allow multiple instances of the same package to be able to have linux and linux-vserver as two instances of linux one with the extension and one without
16:17 < blindcoder> rxr: well, better unplug something... 4 USB devices (none a HID) are maybe a bit much
16:18 < mnemoc> USB1.1=11MB?
16:18 < blindcoder> no idea... I'm just a USB user so I only care about it working :{
16:18 < blindcoder> :)
16:19 < rxr> mnemoc: we can think of s.th. better for the future - for 2.1 I want it cleaner then 2.0 and 2.1 ROCK was and until October the 30th .. ;-)
16:19 < mnemoc> sure
16:20 < rxr> for 3.0 we can thing of revolutional designs that surpase the current improvement
16:20 < mnemoc> i never asked to include 'intances' on t2-2.1, i just think removing linux-src is wrong :p
16:20 < mnemoc> but it's mostly a feeling
16:24 < mnemoc> cool, acm people added online books :D
16:24 -!- CIA-9 [~CIA@to.je.spocco.com] has quit [Remote closed the connection]
16:25 -!- daja77 [[lp0hr3KnN@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
16:27 < rxr> ouhm - what happend with CIA-9 ?
16:28 < rxr> acm what?
16:28 < mnemoc> www.acm.org
16:28 < mnemoc> http://acm.books24x7.com/
16:29 < rxr> when linuxXY-src would be right - we would have a -src of every package ...
16:29 < rxr> and you see that this is ugly by the xfree86-compiled-source and ffmpeg-source s.th. packges I removed in T2 ...
16:31 -!- CIA-9 [~CIA@to.je.spocco.com] has joined #t2
16:31 < rxr> ah ;-)
16:32 < mnemoc> rxr: linux is not such as every other package, is the kernel
16:32 < mnemoc> the most rebuilt package
16:32 -!- daja77 [[tJhFl75hP@odoaker.hrz.tu-chemnitz.de] has joined #t2
16:32 < rxr> it is the most rebuild package? why this?
16:33 < rxr> and if you are right and it is - it is even more important to get up-to-date 3rd party modules selected with every rebuid ...
16:34 < mnemoc> what's the first thing your rebuild after the machine is installed (if you rebuild any)
16:35 < rxr> not the kernel ;-)
16:35 < rxr> really - nowadays I seldom rebuild a custom kernel ...
16:35 < rxr> but anyway - if you do you are happy when you get all the modules alongside ...
16:36 < rxr> and if you want to build it manually you can untar it as you do for every other package, too ...
16:36 < mnemoc> if third party modules are installed on modules/misc there is no problem if you rebuild the same version of the kernel
16:36 < rxr> auch - this is even more evil - the modules might crash and burn on you, e.g. changes in smp, vs. preemtion, vs. power-management ...
16:37 < rxr> you can be lucky if the modules load - but most probably they will crash sooner or later ...
16:37 < rxr> (where I'm not so sure if I'm lucky if the module laod and later crash on me ... ;-)
16:38 < mnemoc> =)
16:38 < mnemoc> in worse case i can emerge the third-party stuff after customizing the kernel
16:41 < rxr> but the idea is powerful - especially when you combine it with jsaw's idea to make it possible to dive thru make menuconfig when you run Emerge-Pkg ;-)
16:41 < rxr> then all the 3rd party stuff is build with that selected config all automatically ...
16:43 < mnemoc> :)
16:55 * mnemoc reading mail
16:55 < rxr> there might not be much more new stuff for you in it - just a summary to let other's now ...
16:55 < rxr> know even
16:56 < mnemoc> :)
17:14 < mnemoc> rxr: on Cleanup i get _lot_ of WARNING: Non-Standard Perm at $everydir .... do you have this too?
17:15 < rxr> jep
17:15 < rxr> I think you get those when you checkout as user ...
17:16 < mnemoc> i did it as root
17:23 < blindcoder> if the dirs are not 0755 or 0775 you get those
17:23 < blindcoder> ie, remove those s bits
17:25 < mnemoc> # stat target/ -c %a
17:25 < mnemoc> 1755
17:25 < mnemoc> uhm
17:26 < mnemoc> why the $badword are they +t??
17:27 < mnemoc> # svn co http://svn.exactcode.de/t2/trunk t2-temp
17:32 < rxr> no idea
17:33 < rxr> hm - right now I do not get them anymore :-(
17:33 < mnemoc> http co is slooooooow
17:33 < rxr> yes - a bit slower than svn:// ;-)
17:34 < rxr> you can co svn:// and switch --relocate ;-)
17:34 < mnemoc> i want to see this +t on a clean checkout, and fix it if it still there
17:34 < mnemoc> Checked out revision 4232.
17:35 < mnemoc> # stat t2-temp/ -c %a
17:35 < mnemoc> 1755
17:35 < mnemoc> wtf
17:35 < mnemoc> # stat t2-temp/target/ -c %a
17:35 < mnemoc> 1755
17:38 < mnemoc> rxr: any idea why is this happening?
17:38 < rxr> hm - no does not happen here ...
17:39 < rxr> what svn version?
17:39 < blindcoder> what is +t?
17:39 < rxr> sticky
17:39 < blindcoder> meaning what?
17:39 < rxr> blindcoder: ?
17:39 < mnemoc> # svn --version
17:40 < mnemoc> svn, version 1.0.5 (r9954)
17:40 < blindcoder> rxr: what does it do?
17:40 < blindcoder> rxr: what does sticky bit do?
17:40 < rxr> I see - I need to upadte the "t2 handbook" then ,-)
17:41 < rxr> STICKY FILES
17:41 < rxr> On older Unix systems, the sticky bit caused executable files to be
17:41 < rxr> hoarded in swap space. This feature is not useful on modern VM sys-
17:41 < rxr> tems, and the Linux kernel ignores the sticky bit on files. Other ker-
17:41 < rxr> nels may use the sticky bit on files for system-defined purposes. On
17:41 < rxr> some systems, only the superuser can set the sticky bit on files.
17:41 < rxr> STICKY DIRECTORIES
17:41 < rxr> When the sticky bit is set on a directory, files in that directory may
17:41 < rxr> be unlinked or renamed only by root or their owner. Without the sticky
17:41 < rxr> bit, anyone able to write to the directory can delete or rename files.
17:41 < rxr> The sticky bit is commonly found on directories, such as /tmp, that are
17:41 < rxr> world-writable.
17:41 < blindcoder> thanks :)
17:42 < mnemoc> any idea why mine are +t and your not?
17:42 < rxr> blindcoder: no problem *g*
17:42 < rxr> mnemoc: No - I really have no idea ...
17:42 < blindcoder> mnemoc: what's the stat of the dir you're trying to co in?
17:42 < rxr> can you test if running the checkoug on a tmpfs makes a difference?
17:43 < mnemoc> # stat . -c %a
17:43 < mnemoc> 755
17:43 < rxr> maybe you hit a bug in the kernel or so ...
17:43 < rxr> what FS are you on?
17:43 < rxr> maybe strace svn and look what it does ...
17:43 < mnemoc> lvm1+reiser
17:43 < mnemoc> rxr: that looks as a big _pain_
17:43 < rxr> hm - reiser here, too ...
17:44 < rxr> do you ever experienced it before?
17:44 < mnemoc> with rock too
17:44 < mnemoc> i'll test updating svn
17:44 < rxr> hm - it most probably is not svn itself ...
17:45 < rxr> I never saw this - regardless what version in use ... - I even use 0.30 or so on the rs6k .. (yes - I should update ... ;-)
17:45 < rxr> I saw some of the permission warnins - but that was a tree which I checked out as normal user ....
17:45 < rxr> and the check complains about non-root files IIRC ...
17:46 < blindcoder> rxr: how much would existing rock targets have to be altered to be usable with t2?
17:46 < rxr> test if checking out just some files (e.g. target) yields the same
17:46 < mnemoc> mom
17:46 < rxr> you then can run such a "tiny" checkout in strace and scroll thru to see if it is in user-space what goes wrong ...
17:46 < rxr> right now - nearly nothing needs to be altered
17:47 < blindcoder> pkgsel I wager?
17:47 < rxr> just see the list for the linux*-src packages beeing removed
17:47 < blindcoder> root@fuzzy:/usr/src# stat t2-temp/ -c %a
17:47 < blindcoder> 755
17:47 < blindcoder> root@fuzzy:/usr/src# svn --version
17:47 < blindcoder> svn, version 0.35.0 (r7994)
17:47 < rxr> yes - ok you need to tweak your selection if you select per repsoitory ...
17:47 < blindcoder> I'd say it's svn
17:47 < blindcoder> rxr: no, I select per package
17:48 < blindcoder> rxr: I'd like to see how LVP compiles with t2 :-)
17:48 < rxr> then you should not need to tweak anything - AFAICS
17:48 < rxr> just that the linux*-src removal currently causes these tiny regressions that are on-the-way of beeing fixed
17:48 < blindcoder> I se
17:48 < blindcoder> e
17:49 < rxr> I removed stuff that I expected not to be maintained in t2 - lvp was one of those
17:49 < rxr> I wanted a basic t2 and no shouting I would steal everyones code ...
17:49 -!- daja77 [[tJhFl75hP@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
17:49 < mnemoc> # strace -o svn.strace svn co http://svn.exactcode.de/t2/trunk/Documentation t2-doc
17:50 < blindcoder> sure :)
17:50 < rxr> so /me doing a bit math - /me in the background ...
17:50 < blindcoder> well, to be honest, I'll use whatever will be easier for me.
17:50 -!- daja77 [[gZMFU9g6X@odoaker.hrz.tu-chemnitz.de] has joined #t2
17:52 < mnemoc> lstat64("t2-doc", 0xbfffedb0) = -1 ENOENT (No such file or directory)
17:52 < mnemoc> mkdir("t2-doc", 07777) = 0
17:52 < mnemoc> lstat64("t2-doc/.svn", 0xbfffed40) = -1 ENOENT (No such file or directory)
17:52 < mnemoc> mkdir("t2-doc/.svn", 07777) = 0
17:52 < mnemoc> # stat -c %a t2-doc/
17:52 < mnemoc> 1755
17:52 < blindcoder> mnemoc: what does umask say?
17:52 < mnemoc> 0022
17:53 < blindcoder> strange
17:53 < mnemoc> yeah
17:53 < mnemoc> i'll try with svn 1.0.8 now
17:56 < mnemoc> # stat -c %a t2-doc/
17:56 < mnemoc> 755
17:56 < blindcoder> HAH! I WAS RIGHT!
17:56 < blindcoder> IT IS a bug in svn :D
17:56 -!- daja77 [[gZMFU9g6X@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
17:56 < mnemoc> yeah :D
17:56 < rxr> interesting
17:57 < blindcoder> nice to see that I am still able to predict some things :)
17:57 < mnemoc> hehe
17:57 -!- daja77 [[HbrsTpDcJ@odoaker.hrz.tu-chemnitz.de] has joined #t2
18:04 -!- daja77 [[HbrsTpDcJ@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
18:08 < rxr> blindcoder: "predict"? - hehe
18:08 < blindcoder> rxr: yes. Example: The day you left ROCK I said to esden that you will do a fork :)
18:09 -!- daja77 [[6QpsUWXbZ@odoaker.hrz.tu-chemnitz.de] has joined #t2
18:09 < mnemoc> wasn't obvious?
18:09 < blindcoder> mnemoc: PSST! You're ruining everything :)
18:10 < mnemoc> :p
18:11 < blindcoder> guess I will release Cocktail Bar 1.5 today
18:12 < mnemoc> german only?
18:12 < blindcoder> german and english
18:13 < blindcoder> the "big" database is german only because I'm too lazy to translate 315 Cocktails
18:16 < rxr> blindcoder: that was an easy prediction - I spend most of my last 3 years or so working on ROCK, doing much PR, ... and need some working system ...
18:17 < rxr> and although we'll rewrite the build system - the "package parameter files" are a source I do not want to miss ...
18:17 < blindcoder> rxr: yes. I just remembered what you even said about maintaining devfs
18:18 < blindcoder> which will be gone in 2.8 anyway
18:18 < blindcoder> http://www.kerneltraffic.org/kernel-traffic/kt20040906_273.html#1
18:25 < mnemoc> imvho devfs is good for linux, udev is good for hurd
18:26 < blindcoder> I never really had a look at hurd
18:26 < blindcoder> I prefer things that work _now_ :)
18:26 < mnemoc> devfs :D
18:39 < martin> rehi
18:48 < rxr> hi m
18:48 < rxr> atin ;-)
18:49 < rxr> I decided to go udev - it works, and I waste no time on maintaining devfs ... ,-)
18:49 < rxr> the next linux magazin will feature an overview article about udev
18:49 < rxr> with some devfs comparision ...
18:49 < rxr> will be out in a month ...
18:50 < rxr> the current issue has an X.org article from me ;-)
18:50 < rxr> and for the 2005/01 I also have two in queue ...
18:52 < blindcoder> well, maybe I should have a look at udev soon, too
18:52 < rxr> so - me soon away ...
18:55 < rxr> just invest in that linux magazin issue *g*
18:55 < blindcoder> that reminds me..
18:55 * rxr needs to stay free journalist a bit longer ;-)
18:56 < blindcoder> I have to send a "moved" notice to Linux User...
19:00 < blindcoder> it's really sad that some people just can't communicate civilized :(
19:02 < mnemoc> uhm?
19:03 < blindcoder> I just asked xor if he allows me to convert his cocktail website to my cocktailbar format (the aforementioned 315 cocktails)
19:03 < blindcoder> well, he told me in unmistakingly language that I'm not because I once said something bad about him 3-4 years ago
19:13 < blindcoder> Personally, I don't know what's with calling someone an asshole if he behaves like one *shrug*
19:15 < mnemoc> :)
19:17 < blindcoder> hmm
19:18 * blindcoder just read rxrs mail wrt linux-src
19:18 < blindcoder> the solution sounds like the pkg_${pkg}_post.conf files
19:18 < blindcoder> wouldn't it be easier to use that instead of postlinux.conf ?
19:18 < blindcoder> or is that not in t2?
19:19 < mnemoc> i hope pkg_${pkg}_post.conf still there
19:19 < rxr> yes it is
19:19 < rxr> blindcoder: nope - you need the download file and version and such
19:20 < rxr> thus the continue with building this other package thing ...
19:20 < blindcoder> ah, okay
19:20 < rxr> if you use the current pkg_*_post.conf stuff you would need to hardcode the files, versions, ...
19:20 < blindcoder> which would suck because it will always be missed with package updates
19:21 < rxr> jep - thus I'm working on a cleaner, less hardcoded version ;-)
19:21 < rxr> so /me off - cu around midneight - or in the morning ...
19:21 < mnemoc> cu rxr
19:21 < blindcoder> bye
19:26 -!- daja77 [[6QpsUWXbZ@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
19:26 -!- daja77_ [[Ed8Jhlb4C@odoaker.hrz.tu-chemnitz.de] has joined #t2
20:15 -!- daja77_ [[Ed8Jhlb4C@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
20:15 -!- daja77 [[0f673MzVA@odoaker.hrz.tu-chemnitz.de] has joined #t2
20:22 -!- daja77 [[0f673MzVA@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
20:23 -!- daja77 [[PIMX+YLc6@odoaker.hrz.tu-chemnitz.de] has joined #t2
20:30 -!- daja77 [[PIMX+YLc6@odoaker.hrz.tu-chemnitz.de] has quit [Remote closed the connection]
20:36 < blindcoder> okay, guess i'm now done with typo3
20:38 < blindcoder> once you've tamed that beast it's really nice to work with it
20:50 -!- daja77 [[IEbLaMkfC@odoaker.hrz.tu-chemnitz.de] has joined #t2
20:51 < blindcoder> www.crash-override.net now running on typo3 ^^
20:54 -!- martin [~martin@brln-d9ba08fc.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
22:34 -!- martin [~martin@brln-d9ba0174.pool.mediaWays.net] has joined #t2
--- Log closed Sat Oct 09 00:00:18 2004