--- Log opened Sat Jan 28 00:00:38 2006
00:46 < mnemoc> rehi
00:58 -!- keinek [n=keinek@200-70-144-40.mrse.com.ar] has quit ["Abandonando"]
00:59 < morfoh> wb mnemoc
01:01 < mnemoc> thx morfoh
01:01 < mnemoc> morfoh: http://www.ieiworld.com/en/product_IPC.asp?model=WAFER-9371A&e=banner180x90L :)
01:03 < morfoh> mnemoc: :)
01:04 < morfoh> mnemoc: how much does it cost ?
01:04 < mnemoc> no idea :(
01:04 < mnemoc> that's always the problem on this sh*ts
01:05 < mnemoc> south american distributor of IEI has never answered me a mail
01:06 < morfoh> :\
01:06 < mnemoc> they have amazing boards
01:09 < morfoh> mnemoc: but with all that components on board it's definitely not cheap
01:10 < mnemoc> you will never know :\
01:16 < morfoh> mnemoc: yes ... perhaps never ever
01:17 < mnemoc> morfoh: how is your reference at 2.1-test going?
01:17 < morfoh> mnemoc: quite good
01:18 < mnemoc> no regressions due to openssl and gcc update?
01:18 < morfoh> we had mysqlcc which seem to have problems with new openssl
01:19 < morfoh> 12 packages failed for now at the nearby end of stage 5
01:21 < mnemoc> i got that error before too
01:21 < mnemoc> but after rebuilding at the end it wnt fine
01:21 < mnemoc> mysqlcc and another package
01:22 < mnemoc> i blamed you and you couldn't reproduce it
01:35 < CIA-7> amery * r15085 /trunk/package/ (5 files in 5 dirs): * added missing .cache files for bcron, runit-logacct, runit-shutdown, flashplayer and djbdns from an azazel build at 2.1-test
01:39 < CIA-7> amery * r15086 /branches/2.1/package/ (6 files in 6 dirs): * added missing .cache files for bcron, runit-logacct, runit-shutdown, eclipse, flashplayer and djbdns from an azazel build at 2.1-test
02:05 < mnemoc> DicOOo.sxw is amazing.... works perfectly :)
02:06 < morfoh> mnemoc: really ? :) ... last time I tried the dictionary support within ooo I it wasn't that perfect
02:07 < mnemoc> i tried on a fresh azazel+openoffice from 2.1-test.... worked perfect
02:07 < morfoh> :)
02:08 < mnemoc> i added es_CL es_ES en_US en_UK it_IT and de_DE
02:08 < mnemoc> flawless
02:08 < morfoh> good to know :)
02:10 < CIA-7> amery * r15087 /trunk/package/office/ooo/ooo.desc: * enhanced ooo.desc by referencing DicOOo.sxw
02:10 < mnemoc> btw, i did is as luser
02:11 < morfoh> hehe :)
02:13 < mnemoc> it was also nice to open the .sxw inside firefox out-of-the-box
02:14 * morfoh hates plone's default look and feel
02:15 * mnemoc too
02:17 < morfoh> mnemoc: what site nearly comparable with T2 do you like ?
02:17 < mnemoc> ehm
02:18 < mnemoc> no idea
02:18 < CIA-7> amery * r15088 /branches/2.1-test/package/office/ooo/ (ooo.conf ooo.desc):
02:18 < CIA-7> * merged 13349,13413,14954 from trunk: improved ooo to build UI language using a $ooo_lang var
02:18 < CIA-7> * merged 15087 from trunk: referenced DicOOo.sxw on ooo.desc
02:20 < morfoh> mnemoc: Error logs from ref-2.1-test-x86-athlon-xp-generic:
02:20 < morfoh> [5] network/gsmlib [5] multimedia/avifile
02:20 < morfoh> [5] graphic/pixieplus [5] kde/superkaramba
02:20 < morfoh> [5] database/mysqlcc [5] x11/gscmxx
02:20 < morfoh> [5] multimedia/deva [5] games/xarchon
02:20 < morfoh> [5] develop/ocs [5] network/socat
02:20 < morfoh> [5] perl/mod_perl [5] java/jboss-as
02:20 < morfoh> 3298 builds total, 1685 completed fine, 12 with errors.
02:20 < morfoh> ref build is at 9-glibc
02:21 < mnemoc> socat faling due to openssl symbols if memory doesn't betrade me
02:21 < mnemoc> betray
02:24 < morfoh> ack ... undefined reference to `OPENSSL_add_all_algorithms_noconf'
02:26 < mnemoc> i'll port LIBTOOL-QUIRK and AUTOMAKE-QUIRK to 2.1
02:26 < CIA-7> jsaw * r15089 /trunk/package/gnome2/pybliographer/ (3 files):
02:26 < CIA-7> * update pybliographer (1.2.7 -> 1.2.8)
02:26 < CIA-7> * use xvfb-run for config script
02:26 < CIA-7> * copyright updates
02:27 < mnemoc> jsaw: isn't xvfb-run nice? :)
02:34 < mnemoc> lot of stupid python scripts try to import pygtk to see if it's installed :(
03:18 < CIA-7> amery * r15090 /trunk/package/office/ooo/t2-distro.patch: * fixed ooo to build fine with ooo_lang=ALL
03:24 -!- morfoh_ [n=jeru@p54BED8B2.dip.t-dialin.net] has joined #t2
03:24 < CIA-7> amery * r15091 /trunk/package/java/eclipse/eclipse.conf: * added a $PATH friendly `eclipse`
03:25 -!- morfoh [n=jeru@p54BEF32F.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)]
03:25 -!- morfoh_ is now known as morfoh
04:00 -!- mtr [n=Michael@p54AF973D.dip0.t-ipconnect.de] has joined #t2
04:16 -!- mtr_ [n=Michael@p54AFCBD1.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)]
04:17 -!- morfoh [n=jeru@p54BED8B2.dip.t-dialin.net] has quit ["using sirc version 2.211+KSIRC/1.3.11"]
07:10 < rxr> re
07:10 [Users #t2]
07:10 [@ChanServ] [ CIA-7] [ mnemoc] [ rxr ] [ _Ragnar_]
07:10 [ axion_ ] [ jsaw ] [ mtr ] [ sepp ]
07:10 [ Baldzius] [ LMJ ] [ owl ] [ valentin]
07:10 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
07:11 < rxr> re
07:18 -!- rxr_ [n=rene@e178154243.adsl.alicedsl.de] has joined #t2
07:18 -!- Topic for #t2: T2 | 2.1.1.1 and 2.2.0-epsilon RELEASED! | The System Development Environment (SDE) | http://www.t2-project.org/ | Say hello and do not hesitate to ask us any questions that you may have. | http://www.rafb.net/paste/
07:18 -!- Topic set by mnemoc [] [Sat Jan 14 21:42:03 2006]
07:18 [Users #t2]
07:18 [@ChanServ] [ CIA-7] [ mnemoc] [ rxr ] [ valentin]
07:18 [ axion_ ] [ jsaw ] [ mtr ] [ rxr_] [ _Ragnar_]
07:18 [ Baldzius] [ LMJ ] [ owl ] [ sepp]
07:18 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
07:18 -!- Channel #t2 created Sun Aug 8 19:15:33 2004
07:18 -!- [freenode-info] if you need to send private messages, please register: http://freenode.net/faq.shtml#privmsg
07:19 -!- Irssi: Join to #t2 was synced in 41 secs
07:21 -!- rxr [n=rene@e178179167.adsl.alicedsl.de] has quit [Read error: 145 (Connection timed out)]
08:02 < rxr_> valentin: awake?
08:02 -!- You're now known as rxr
08:11 -!- raamcm [n=root@59.145.141.50] has joined #t2
08:11 -!- raamcm [n=root@59.145.141.50] has quit [Client Quit]
08:12 -!- raamcm [n=raamcm@59.145.141.50] has joined #t2
08:13 < raamcm> Hi guys
08:19 < rxr> hi raamcm
08:22 < raamcm> hi rxr
08:23 < raamcm> thx god
08:23 < raamcm> i got help
08:23 < raamcm> last two days i has some other problem
08:23 < raamcm> then now i retun back
08:25 < raamcm> http://pastebin.com/526978
08:25 < raamcm> plz see my error
08:25 < rxr> you build the router? hm
08:25 < rxr> it might not be really up-to-date
08:26 < rxr> and seldom tested
08:26 < rxr> it was just a test at ROCK Linux times and rotting there also
08:26 < rxr> we use the embedded target to base sophisticated stuff on
08:26 < rxr> what kind of system do you like to get?
08:27 < raamcm> i need only minimal packages
08:27 < rxr> build generic with minimal selection for a start
08:27 < raamcm> i did
08:27 < rxr> or if you want to cross compile use a embedded target
08:27 < raamcm> no need
08:27 < rxr> your build error lists router target output
08:28 < raamcm> ok wht i do
08:28 < raamcm> i am got one more error also
08:28 < rxr> rerun the scripts/Config and select generic instead of router
08:28 < rxr> and then the minimal package selection preselection template
08:28 < rxr> to get a quite minimal package set
08:28 < raamcm> http://pastebin.com/526984
08:29 < raamcm> plz see that link also
08:30 < rxr> second
08:30 * rxr deep in kernel debugging
08:30 < raamcm> yeah
08:31 < raamcm> now i select generic system and select minimal package selection
08:32 < rxr> raamcm: this error is also due rotting router
08:32 < rxr> I think I just rm -rf the router target is is just old, rotting and stupid ..
08:32 < rxr> our new embedded target technology is years ahead ...
08:32 < rxr> raamcm: why did you actually tried building the router?
08:33 < raamcm> hmm
08:33 < raamcm> no basically i don't know
08:33 < rxr> raamcm: if you want a "normal" linux distribution build generic
08:33 < raamcm> ok fine
08:33 < rxr> if you want something very tiny, for cross build and embedded use, use the embedded targets
08:34 < rxr> the router was the first tiny embedded try we inherited from rock linxu and I think since it is so totally stupidly done it is time to rm it, especially with our in-producion use, commercially accepted embedded targets around
08:34 < raamcm> no i don't need in embedded linux
08:35 < raamcm> but i need only small linux
08:35 < rxr> raamcm: then router would not have been the right for you even if it would have been built
08:35 < raamcm> yeah i know
08:35 < rxr> raamcm: generic w/ minimal or minimal+xorg selection is for you then
08:35 < raamcm> that stupid option ate my one week
08:35 < rxr> raamcm: sorry
08:36 < raamcm> plz guide me
08:36 < rxr> mnemoc: when it turns light on your side of the world, let's discuss about the router target's future - e.g. it /dev/null
08:36 < rxr> mnemoc: * in /dev/null
08:36 < raamcm> how can i get basic linux build
08:36 < rxr> raamcm: select generic plus a minimal package preselection
08:36 < rxr> run Dowlonad -required
08:37 < rxr> and Build-Target
08:37 < rxr> and you are done
08:37 < rxr> raamcm: do you use 2.1 or trunk?
08:37 < raamcm> 2.1.1.1
08:38 < rxr> you want a bootable cd, right?
08:38 < raamcm> yeah exacatly
08:38 < rxr> ok, then there is one more step on 2.1 that is already optimized away on trunk
08:39 < rxr> but first build the minimal generic selection as above and then we do the bootable cd step
08:39 < raamcm> tell me
08:39 < raamcm> ok
08:39 < raamcm> can i remove the folder
08:40 < rxr> which folder?
08:41 < raamcm> t2 folder
08:41 < rxr> nah!
08:41 < rxr> why do you want to remove it
08:41 < rxr> you could remove your previous router build, but the whole folder?
08:41 < raamcm> ok
08:41 < rxr> ./script/Cleanup -build build/default-router-* or so
08:41 < raamcm> ok
08:41 < rxr> do not use rm -rf on build/... it could remove more then you think it would due to --bind mounts
08:42 < rxr> always use scripts/Cleanup ... to remove t2 builds
08:42 < raamcm> yeah i gave
08:43 < raamcm> that's everything cleaned
08:43 < raamcm> wht i should do now
08:43 < rxr> I already wrote
08:43 < rxr> ./scripts/Config
08:43 < rxr> ./scripts/Download -require
08:43 < raamcm> ok
08:43 < rxr> d
08:43 < rxr> ./scripts/Build-Target
08:43 < raamcm> yeah i have
08:44 < raamcm> hi need without X stuff
08:44 < rxr> you need no X ?
08:45 < raamcm> yeah
08:45 < rxr> then select the minimal package selection template, it is withotu X
08:45 < raamcm> yeah i did
08:45 < raamcm> ok that's going on
08:46 < raamcm> instead some curl error is coming
08:46 < raamcm> "requeded URL returend error :404"
08:48 < raamcm> that's lunch time
08:48 < raamcm> shall i come after 15 min
08:48 < raamcm> bye now
08:48 < raamcm> :-)
08:53 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
08:54 < misl> morning all
08:55 < rxr> moin misl
08:56 < misl> rxr: I will try to merge your pdksh change into 2.1
08:56 < misl> was your change related to building for mips?
08:57 < rxr> hm - I do not remeber
08:57 < rxr> the fix also might be in scripts/
08:57 < rxr> fixing extreme rude code from the early rock age
09:00 < misl> So it was not in the package itself? :(
09:01 < rxr> IIRC yes
09:01 < rxr> misl: when this damned Adaptec USB2Xchange works I take a look
09:02 < misl> rxr: I don't want to hurry you, I wanted to try and fix it myself
09:02 < rxr> misl: can you post the error you get here?
09:03 < misl> rxr: the configure script says it can't detect int and long size when crossbuilding
09:03 < misl> this results in SIZEOF_INT = 0 and SIZEOF_LONG = 0
09:04 < rxr> in which stage?
09:04 < misl> I am not sure: either 0 or 1
09:04 < rxr> that is the important bit ,-)
09:04 < misl> I will look it up
09:05 < misl> stage 1, pdksh is not build in stage 0
09:16 < rxr> i think the changes where all in scripts/
09:16 < rxr> I did change the .cache generation quite some times to make it sane
09:16 < rxr> the config.cache prefilling that is
09:17 < misl> :( that makes it harder to find and merge the change
09:20 < rxr> shoudl be in function.in
09:21 < rxr> that is why I said I would hunt it ...
09:21 < misl> let me try to hunt it ;-)
09:21 < rxr> I already use the compile time breaks to hunt it
09:21 < valentin> moin
09:21 < rxr> the bits I reffer to are in function.in
09:22 < misl> hi valentin
09:22 < rxr> hunt all peaces related to config.cache creation that are not yet in 2.1
09:22 < rxr> moin valentin
09:22 < rxr> valentin: time to phone?
09:22 < valentin> hi misl, rxr
09:22 < valentin> rxr: mom
09:29 -!- raamcm [n=raamcm@59.145.141.50] has quit [Read error: 104 (Connection reset by peer)]
09:29 -!- raamcm [n=raamcm@59.145.141.50] has joined #t2
09:30 < raamcm> :-)
09:30 < raamcm> i back
09:30 < raamcm> rxr i have one doubt
09:30 < misl> welcome back raamcm
09:31 < owl> moin
09:31 < misl> hi owl
09:31 < owl> hi misl
09:31 * misl off to do some shopping
09:31 < misl> bbl
09:31 < raamcm> thx
09:31 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit ["Chatzilla 0.9.69.3 [Firefox 1.5/2005111116]"]
09:41 < sepp> mc
09:41 < sepp> uhm :)
09:41 < sepp> moin
09:45 < LMJ> morning
09:56 < rxr> moin sepp and LMJ
09:57 < LMJ> moin rxr
09:57 < LMJ> snow here ;)
09:57 < LMJ> brb
09:57 < rxr> LMJ: ,-)
09:57 < rxr> just -10 degree here no snow since days ... :-(((
10:01 < rxr> hm
10:01 < rxr> I have the feeling the 2nd reset sequence is not comming thru to to the USB2Xchange ...
10:11 < sepp> hi rxr, LMJ
10:27 < CIA-7> sebastian * r15092 /trunk/package/graphic/autopano-sift/autopano-sift.conf: * not install the win32 version
12:32 -!- raamcm [n=raamcm@59.145.141.50] has quit ["Leaving"]
12:44 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has joined #t2
12:44 < misl> :)
12:46 -!- morfoh [n=jeru@p54BED244.dip.t-dialin.net] has joined #t2
12:47 < misl> hi morfoh :)
12:47 < morfoh> hi rocket :)
12:47 < misl> :) your timing has improved
12:47 < morfoh> in what kind ?
12:48 < misl> half a year ago you arrived when I intended to leave
12:48 < misl> now I just arrived and you arrive too :)
12:49 < morfoh> hehe :)
13:17 < rxr> mon morfoh
13:17 < rxr> moin even
13:17 < rxr> morfoh: how is your AP going?
13:19 < morfoh> moin rxr ... the AP is doing well but I wasn't able to build a mipsel kernel yet, but anyway I'm already working on T2's upcomming website
13:20 < rxr> ah, the later i also good
13:20 < rxr> morfoh: you read into the details needed for a kernel for that CPU?
13:20 < rxr> MIPS kernels are extremely target cpu dependant and AFAIK there is no working open source code for the w-lan chip yet anyway
13:21 < rxr> e.g. the bcm4xxx project code is not AP ready AFAIK
13:22 < morfoh> rxr: hmmmm ... yeah I guess I have to do more information gathering. for the first I just tried to build a "generic" mips kernel
13:22 < rxr> I think just a random MIPS kernel will not boot at all
13:22 < morfoh> rxr: yeah ... of course not
13:23 < rxr> e.g. even for the sgi workstatoin boxes you need a special tweaked kernel for _each_ of those sgi machines ... :-0
13:59 < CIA-7> msluis * r15093 /branches/2.1-test/scripts/functions.in:
13:59 < CIA-7> * merged 12970 from trunk: improved configure option sanity filter
13:59 < CIA-7> to handle --with-* and --without-*
13:59 < CIA-7> * merged 12448 from trunk: removed irritating dots '.' in build
13:59 < CIA-7> log message
14:11 * misl off for a walk, bbl
14:16 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has joined #t2
14:16 < Bart-> hi
14:17 < Bart-> is there any way to verify all installed packages (except mine -y)?
14:18 < rxr> using the /var/adm checksum / md5sum files by hand ?
14:19 < Bart-> but this is the same as mine -y and I need to find all missing files, not only these modified
14:19 < rxr> donsn't mine -y also print the missing ones?
14:19 < Bart-> nope, I even removed half of coreutils to test it
14:20 * rxr would call that a bug ...
14:20 < Bart-> but -y says that most of message catalogs were modified, this is crap
14:21 < Bart-> after playing with Emerge-Pkg yesterday, my T2 is trashed again - won't boot anymore
14:21 < rxr> hm
14:23 < Bart-> I read this "only 5% oudated" handbook before breaking t2 yesterday but it says nothing about preselecting packages using Config
14:24 < rxr> package preselection exists since some time now, the handbook has a chance to listing that
14:24 < rxr> but I do not know off hand
14:24 * rxr deep into kernel code
14:26 < mnemoc> rxr: ping
14:26 < rxr> pong
14:27 < mnemoc> hi :)
14:27 < rxr> moin mnemoc
14:27 < morfoh> moin mnemoc
14:28 < mnemoc> moin rxr,morfoh
14:28 * rxr enabled scsi and usb debugging and loggin in the kernel, + custom output, scrolling info like crazzy, but still not the important bits I miss ...
14:30 < rxr> mnemoc: you pinged me because you wanted s.th. or just as moin ?
14:30 < mnemoc> rxr: guilhem said he will get a x86_64 this week, so he will start fixing kaffe
14:30 < mnemoc> 05:38:36#t2: mnemoc: when it turns light on your side of the world, let's discuss about the router target's future - e.g. it /dev/null
14:30 < mnemoc> 05:38:53#t2: mnemoc: * in /dev/null
14:31 < rxr> yep
14:31 < rxr> /dev/null, ACK ?
14:31 < mnemoc> i guess madtux wont fix it, is morfoh interested?
14:31 < rxr> mnemoc: thing is router target is 99% pure crap
14:31 < mnemoc> i have never used it
14:31 < rxr> inheriting embedded and adding some real code would be less work and a more satisfying result
14:31 < mnemoc> but yes, it's also broken on 2.1
14:31 < morfoh> mnemoc: nope ... I'm not interested in the router target
14:32 < morfoh> mnemoc: I think sedu is the better way
14:32 < mnemoc> ok, svn rm from trunk and 2.1
14:32 < mnemoc> morfoh: of course :)
14:32 < morfoh> mnemoc: hehe :)
14:32 < rxr> mnemoc: or wait until mr. miguel is showing along ... ?
14:32 < mnemoc> better
14:32 < rxr> but, well, a embedded based router would make more sense and be cleaner anyway ...
14:33 < mnemoc> yes, but we have to ask the current maintainer of that target what will he do
14:33 < rxr> guess miguel also never used it, only tried to keep this rock^Wcrap code building ...
14:33 < rxr> .o(what was our selinux support status ?)
14:33 < mnemoc> no idea
14:34 < rxr> madtux: ?
14:35 < mnemoc> yep
14:35 < mnemoc> he does some updates from time to time
14:35 < mnemoc> but i don't see any integration
14:35 * rxr neither :-(
14:35 < rxr> well, I don't need it ...
14:36 < rxr> and I rather would investigate non NSA stuff before I just US .gov code ...
14:36 < rxr> rsbac ...
14:36 < mnemoc> rsbac design is nice
14:53 * mnemoc wonders why http://mdbtools.sourceforge.net/ talks about 0.5 and http://ooo.ximian.com/packages/SRC680/ has 0.6
14:55 < rxr> hm
14:55 < rxr> their favourite custom copy?
14:59 < mnemoc> i'll try to enable mdbtools support on openoffice before continuing with java support
14:59 < mnemoc> DictOOO.swx works flawless... so now i don't care about the lack of wizards
15:00 < rxr> DictOOO is what ?
15:00 < rxr> what is mdbtools ?
15:00 < mnemoc> DictOOO a document with a nice macro to download and install dictorionaries
15:01 < mnemoc> mdbtools is stuff to read (and write?) M$ Access files
15:01 < rxr> aha
15:01 < rxr> oho
15:01 < mnemoc> ooo has a nice program called 'Base' which replicate M$ Access
15:01 < mnemoc> reading the same .mdb files, forms, reports, etc...
15:02 < mnemoc> http://ftp.services.openoffice.org/pub/OpenOffice.org/contrib/dictionaries/dicooo/DicOOo.sxw
15:02 < mnemoc> works nice :)
15:02 < mnemoc> even under luser
15:03 < rxr> can't we just package those dicts?
15:03 < mnemoc> yes, but what to install? all?
15:04 < mnemoc> ooo becomes slower exponentially
15:04 < mnemoc> and to enable them it's not just drop them in, just have to write one line per active dictionary on a file
15:05 < mnemoc> you have to... *
15:07 < rxr> valentin: on the athlon you left here IDE is a bit flanky ... :-(
15:08 < mnemoc> rxr: your ooo is 2.0.1(.1) or 2.0.0(.1) ?
15:08 < rxr> oehm
15:08 < rxr> 1.1 I think
15:08 < rxr> why?
15:09 < mnemoc> because my 2.0.0.1 doesn't save as open document formats or star office formats, just M$ :p
15:11 < rxr> I think I have 2.0.1.1 and it was able to save and read .odt
15:11 < rxr> even those saved on windows boxes ..
15:11 < mnemoc> good
15:11 < rxr> oh - and it was able to load windows saved impress files
15:11 < mnemoc> :)
15:11 < rxr> but it crashed every now and then and is not able to draw straight lines in the heading of documents
15:12 < rxr> those lines always get munchy misrendered, even in pdf export ...
15:12 < mnemoc> uhm
15:12 < rxr> and workign with lines in the heading has more bugs, especially when those lines are edited on a page not the first one
15:12 < rxr> the it totally fucks up ...
15:12 < rxr> oh sorry - f*cks up .. ,-)
15:13 < mnemoc> don't say blasfemies here please
15:13 < mnemoc> :)
15:14 < mnemoc> rxr: and answer the query!
15:19 < mnemoc> misl: ping
15:36 -!- mnemoc [n=amery@200.75.27.19] has left #t2 []
15:36 -!- mnemoc [n=amery@200.75.27.19] has joined #t2
15:36 < mnemoc> yuck
16:06 < valentin> rxr: why ?
16:06 < valentin> 16:08 < rxr> valentin: on the athlon you left here IDE is a bit flanky ... :-(
16:06 < valentin> what is flanky ?
16:06 < valentin> only the chanels are exchanged within linux
16:06 < valentin> but i told you irc
16:16 < misl> mnemoc: pong
16:17 < misl> Does anyone know why crossbuilding in 2.1 only has 71 packages
16:17 < misl> root@noisy:/mnt/space/t2/2.1-test# wc -l config/ref64/packages
16:17 < misl> 66 config/ref64/packages
16:18 < mnemoc> misl: crossbuilding only include packages marked to build at stage 0 and 1
16:18 < mnemoc> misl: on your commit to t2-2.1-test, the second merge has the wrong revision number
16:19 < misl> mnemoc: what revision number was there?
16:19 < misl> oeps it should have been 13448
16:20 < mnemoc> after pinging you i 'discovered' that
16:20 < misl> 12970 and 13448 were merged from trunk
16:20 < mnemoc> i have to remember when i merge to 2.1 :\
16:20 < mnemoc> why did you needed that merge?
16:21 < mnemoc> the first i mean
16:21 < mnemoc> 12970}
16:22 < misl> crossbuilding pdksh failed at stage 1, rxr mentioned he fixed that on trunk in functions.in somewhere
16:22 < rxr> valentin: lost interrupts, stalling, oopsing bruning hanging ...
16:22 < misl> so I looked up what was already merged and what not.
16:22 < misl> and continued merging
16:22 < misl> mnemoc: why? did I do something wrong?
16:23 < valentin> rxr ???
16:23 < valentin> the machine was just fine, but remember the cmos battery had been empty
16:23 < valentin> maybe there is some trash config in the bios ?
16:23 < mnemoc> misl: just curious
16:24 < mnemoc> misl: i would like to merge only needed features
16:24 < mnemoc> misl: if 12970 fix something present on 2.1, welcome
16:24 < misl> mnemoc: I am not sure which of the two merges actually made pdksh build.
16:25 < mnemoc> if it builds now, and i didn't before, and 12970 doesn't bring any regression... i'm happy with it
16:26 < misl> mnemoc: I will do a full rebuild :)
16:27 < misl> (cross)building ref64 restarted.
16:27 < mnemoc> misl: that's why i made 2.1-test :) enjoy it
16:27 < misl> mnemoc: how can I do a full cross reference build of 1600 packages instead of just 71?
16:28 < mnemoc> misl: you can't
16:28 < misl> :(
16:28 < mnemoc> misl: since stage 2 it's all done inside chroot, which doesn't work when crosscompiling
16:29 < mnemoc> misl: at T2 2.2 rene has moved more and more packages to stage 1
16:29 < mnemoc> misl: so the crossbuilding 'base' is wider
16:32 < CIA-7> amery * r15094 /trunk/package/gnome2/gdm/minimal_uid_1000.patch: * fixed GDM to only show UID>=1000 when browsing accounts
16:34 < mnemoc> misl: i have in my TODO to port LIBTOOL-QUIRK and AUTOMAKE-QUIRK to 2.1 if you want to do it :)
16:35 < Bart-> what about those who use UIDs >=500 to unify with other systems?
16:35 < misl> my todo is to actually fully build a target on noisy
16:36 < mnemoc> Bart-: uhm.... good claim
16:37 < mnemoc> misl: those quirks are need to let you build large 2.1 targets on noisy
16:37 < mnemoc> Bart-: which systems use UID>=500?
16:38 < misl> mnemoc: nice way to convince me to do it :)
16:38 < misl> I was already looking at what is needed for libtool quirk
16:38 < Bart-> mnemoc, lots of, eg RedHat based
16:38 < mnemoc> misl: please be conservative on the merges
16:39 < mnemoc> Bart-: the higher system account i now is UID=400, rsbac....
16:39 < mnemoc> know*
16:39 < misl> mnemoc: ack
16:41 < mnemoc> Bart-: well.... that patch only change the default MinimalUID at gdm.conf to the minimal uid used by T2 for users
16:41 < Bart-> mnemoc, so patched GDM will not show 500,501...? Some people may not be able to log in ;]
16:42 < Bart-> mnemoc, I don't care, I even do not have t2
16:42 < mnemoc> Bart-: but it's a good objection anyway
16:42 < mnemoc> Bart-: be t2-consistent or be RH-friendly
16:47 < mnemoc> Bart-: you win, i'll set it to 500
16:47 -!- sepp [n=sepp@p83.129.173.17.tisdip.tiscali.de] has quit [Read error: 104 (Connection reset by peer)]
16:48 < mnemoc> Bart-: 100 is let no-users leak, but 500 seems to be safe
16:48 < mnemoc> s/ is / /
16:48 < rxr> valentin: but the despite the stability the busses do not look exchanged
16:48 < rxr> maybe you mixed the cabling ?
16:49 < rxr> strange
16:49 < rxr> with the first port disabled int he BIOS it works stable ...
16:51 < CIA-7> amery * r15095 /trunk/package/gnome2/gdm/ (minimal_uid.patch minimal_uid_1000.patch): * changed GDM to show UIDs since 500 instead of 1000, to be friendly with other systems
16:52 -!- sepp [n=sepp@p213.54.151.101.tisdip.tiscali.de] has joined #t2
16:52 < rxr> other ?
16:53 < rxr> mnemoc: my uid here in the VPN is 500 since 1997 or so ,-)
16:53 < rxr> mnemoc: how is your first ahtlon in your life going ?
16:54 < mnemoc> rxr: fighting the webcam but going nice :)
16:54 < mnemoc> rxr: pmmx for now
16:54 < mnemoc> rxr: i'll burn a CD with the sources to build an azazel natively there
16:54 < rxr> you did not power up the board yet?
16:55 < mnemoc> it's running
16:55 < misl> mnemoc: should every merge be committed separately?
16:55 < mnemoc> but my shared 64kbps wifi is not nice for Download -required
16:55 < mnemoc> misl: depends on the commit message
16:56 < rxr> mnemoc: ah, you mean pmmx binaries ...
16:56 < rxr> ic
16:56 < mnemoc> misl: lot of FPIC-QUIRK merges can be a single commit
16:56 < rxr> mnemoc: the fans are a bit used and loud, that is why you got two sets of them
16:56 < misl> FPIC_QUIRCK needs to be merged too?
16:56 < misl> -C
16:56 < mnemoc> rxr: do i need to add something between the fans and the processor?
16:56 < mnemoc> misl: many are missing
16:57 < mnemoc> misl: i'm just one person :(
16:57 < rxr> mnemoc: nope
16:57 < mnemoc> rxr: good :) i'm not used to AMD and it's thermal thingies
16:57 < rxr> mnemoc: those are from the slot aera, where those modular have been rock solid and easy to plug togheter
16:58 < mnemoc> ic
16:58 < rxr> mnemoc: it is like pII or pIII or what it was
16:58 < rxr> or alpha
16:58 < misl> mnemoc: no problem
16:58 < misl> Can I start merging AUTOMAKE_QUIRK independantly?
16:58 < rxr> AMD brought the i/o protocol and package from DEC
16:58 < mnemoc> misl: sure, i'll join them when merging to 2.1-stable
17:04 < valentin> rxr: it was exchanged relative to BIOS not to cabel..
17:04 < valentin> strange
17:04 * valentin away for food
17:13 < rxr> valentin: two words: pie sea ...
17:13 < rxr> on a sgi this would not hapend ...
17:13 < rxr> there even the mac would be permanent even after a system board exchange ...
17:18 < morfoh> rxr: I'd like to move the "Platform" section from the main menu to "About -> Supported Platforms" on the website
17:18 < morfoh> to not bloat the main menu
17:18 < rxr> why?
17:19 < rxr> I find supported platforms major ...
17:19 < rxr> valentin: this athlon gets quite hot
17:19 * rxr wonders if that is ok and normal ..
17:20 < misl> mnemoc: is it safe to raise the libtool version too?
17:20 < mnemoc> misl: please don't
17:20 * rxr installing 2.4.25 ...
17:20 < mnemoc> misl: only if it's highly needed, and harmless
17:23 < morfoh> rxr: I agree that "Platform" is "major" but the current structure doesn't look that nice organized
17:23 < misl> mnemoc: no problem but merging r14222 (LIBTOOL_QUIRK) included raising the version to 1.5.20
17:23 < misl> I will stay at 1.5.14
17:23 < mnemoc> misl: merge it partially :)
17:23 < misl> mnemoc: I already did :)
17:24 < mnemoc> thanks :)
17:24 < morfoh> any other voices regarding the site navigation ?
17:27 < mnemoc> ?
17:28 < rxr> morfoh: architectures and targets shold be major entries in the top=level navigatoin
17:28 < mnemoc> architectures, targets, packages
17:28 < mnemoc> as on t2-root
17:29 < morfoh> ok
17:29 * mnemoc hate the 'just' on t2 frontpage
17:30 < mnemoc> morfoh: what happened to the backup you took from www3?
17:31 < morfoh> mnemoc: it wasn't working so I have to do it manually :\
17:31 < mnemoc> o_O
17:32 < rxr> mnemoc: haven't I removed a just somewhere alrady
17:32 < rxr> maybe it was at some more places ...
17:33 < rxr> so 2.4.25 compiling ..
17:33 < mnemoc> rxr: from the fliers :)
17:33 < rxr> damn - what old stuff I test out today is quite rediculous ...
17:33 < rxr> so I can not say I do not run 2.4 since years ... - well in production at least that still is ... ,-)
17:37 < mnemoc> why are you trying linux24?
17:38 < rxr> to review if a ancient bit of patch works with this usb2xchange
17:38 < rxr> I rewrote it 99% to be clean and 2.6, it only works with the usb1 flavour not the usb2 one
17:39 < rxr> and fro the i/o I do not see a difference to the windows snoop I made ...
17:39 < rxr> thus I want to see if this orig patch someone claimed to work with 2.4.25 works in real
17:39 < rxr> with 2.4.31 it did not work with the same behaviour as my 2.6 patch
17:39 < rxr> gcc 3.2.3 onthe box - heh
17:39 < mnemoc> uhm
17:40 < mnemoc> :)
17:40 < rxr> .oO (ROCK 2.0.1)
17:40 < rxr> at least a release made by /me ,-)
17:40 < mnemoc> i was expecting 1.5 :p
17:40 < rxr> at least old enough to run this old kernel ,-)
17:41 < rxr> IIRC gsmp is a hacked up 1.7 ...
17:41 < mnemoc> did th abandon rock?
17:43 < rxr> no idea
17:43 < rxr> rock as usual
17:43 < rxr> stable releas? what is that?
17:43 < mnemoc> but neither merges
17:44 < rxr> I do not monitor 2.1 ..
17:44 < rxr> only submaster and lurker when I have too much free time or need to |relex|
17:44 < mnemoc> and the question about udev on 2.0.4 at $topic has been there for half a year
17:44 < rxr> but lurker recently broke IIRC
17:44 < mnemoc> doh
17:44 < rxr> mnemoc: yeah
17:44 < rxr> mnemoc: without even real udev support in trunk at all ..
17:44 < mnemoc> :)
17:45 < rxr> and netrunner needing help from the T2 inventor rather than the udev README ...
17:45 < misl> grep -c FPIC T2-Changelog --> 27
17:45 < mnemoc> README wasn't complete enough
17:46 < mnemoc> rxr: did you see the free flaming at rock's new page?
17:46 < mnemoc> a FAQ: why not C++, python or java instead of bash
17:46 < rxr> no
17:46 < rxr> have not seen that
17:46 < rxr> free flaming?
17:46 < rxr> what do you mean with free flaming?
17:47 < mnemoc> atack without reasons
17:47 < rxr> heh
17:47 < mnemoc> "Why not rewriting it in Ruby, Python, C++ or Java?"
17:47 < CIA-7> jsaw * r15096 /trunk/package/develop/gwenhywfar/ (gwenhywfar.conf gwenhywfar.desc):
17:47 < CIA-7> * update gwenhywfar (1.99.4 -> 1.99.5)
17:47 < CIA-7> * fix for ssl
17:47 < CIA-7> * copyright update
17:48 < rxr> mnemoc: it is strange to see them throwing all the stuf over the boat the got for free, first the handbook and now the homepage
17:48 < rxr> wonder if they will ever fill them with more content then they already had
17:48 < mnemoc> i doubt
17:48 < mnemoc> but they have LinuxQuestions.org :)
17:49 < rxr> *lol*
17:52 < rxr> http://it.slashdot.org/it/06/01/28/1414209.shtml
17:52 < rxr> ^- bugfixing in the commercial world
17:52 < rxr> "known as secret knowledge base article"?
17:53 < rxr> mnemoc: remeber me when we start to fall into this trap ...
17:54 < mnemoc> promised
17:56 < mnemoc> rxr: where is T2' svg?
17:58 < morfoh> rxr: I guess we'll use the current logo .. right ?
17:59 < mnemoc> that's _the_ logo
18:00 * LMJ forgot how to "refresh" the tree when LMJ just added a new package, mnemoc should know that please ? ;)
18:00 < morfoh> mnemoc: ok, if it's the logo I'll not ask about it anymore
18:01 < mnemoc> morfoh:
18:01 < mnemoc> err
18:02 < mnemoc> morfoh: when we get a new one we will change it everywhere, but until now every logo/mascot proposal has failed
18:02 < mnemoc> LMJ: svn up?
18:02 < LMJ> no mnemoc
18:02 < LMJ> damned, i forgot the command again
18:02 < LMJ> i just made another little .desc
18:03 < LMJ> ./script/Emerge-Pkg can't see it yet
18:03 < mnemoc> Config -oldconfig
18:03 < LMJ> TATATA
18:03 < LMJ> thanks
18:03 * LMJ is writing it down ;)
18:04 < LMJ> == 01/28/06 19:04:53 =[9]=> Finished building package geany.
18:04 < LMJ> i made a package for that : http://geany.uvena.de/
18:05 < LMJ> a text editor, neighboor to nedit or beaver, little bit more oriented development
18:05 < morfoh> mnemoc: yeah ... I know the "logo story" ... I was just asking ... because someone made some nice logo perhaps, that I don't know about
18:06 < rxr> me most probably away in some minutes - cu
18:07 < LMJ> bye
18:07 < mnemoc> cu rxr
18:07 < CIA-7> rene * r15097 /trunk/package/x11/rxvt-unicode/rxvt-unicode.desc: * updated rxvt-unicode (7.1 -> 7.4)
18:07 < mnemoc> LMJ: i'm happy with pida :)
18:07 < LMJ> i dont know it
18:08 < LMJ> so, should i submit the package or not in that case ?
18:09 < misl> Error logs from ref64-2.1-test-x86-64-cross-generic:
18:09 < misl> 71 builds total, 71 completed fine, 0 with errors.
18:10 < mnemoc> LMJ: if you will maintain the package, sure
18:10 < mnemoc> misl: why cross and not native?
18:11 < LMJ> ok diff /dev/null geany.desc > email to ??
18:11 < misl> When doing native I get lots of failures too. I think this is due to noisy not being fully 64 bits
18:11 -!- YeT2 [n=chmo@i538785FC.versanet.de] has joined #t2
18:11 < mnemoc> LMJ: ./package/*/geany/geany.{desc,cache}
18:12 < misl> but back then I thought cross building builds is it all, now I now better
18:12 < mnemoc> misl: what about reinstalling noisy?
18:12 < LMJ> what's the good ML address ?
18:12 < mnemoc> LMJ: aren't you subscribed?
18:12 < LMJ> to severak ones
18:12 < LMJ> i wanna be sure
18:12 < misl> mnemoc: Still waiting for rxr :( to do so. I probably mess things up.
18:13 < misl> last time rxr said noisy might be good for hunting the final x86-64 bugs
18:14 < LMJ> got it
18:14 < LMJ> t2@
18:14 < mnemoc> misl: when will you invite him again? don't you want to send noisy to berlin? :p
18:15 < mnemoc> LMJ: mail filters are a good-thing-to-have (tm)
18:15 < LMJ> got some ;)
18:15 < LMJ> ok
18:15 < LMJ> sent little mnemoc ;)
18:15 < misl> mnemoc: I like to keep noisy close by, I sort of grew fond of it :)
18:16 < mnemoc> misl: :D
18:16 < mnemoc> misl: what about the sound-resistant rack?
18:17 < misl> mnemoc: I am still looking for one
18:18 * morfoh will f* off and is going to hell now ...
18:18 < morfoh> cu later folks
18:21 < mnemoc> morfoh: ?
18:21 -!- nzg [n=tschmidt@p54AFE524.dip.t-dialin.net] has joined #t2
18:21 < nzg> moin
18:21 < mnemoc> moin nzg
18:21 < nzg> hi mnemoc
18:22 < misl> morfoh: is that a nice club?
18:22 < misl> heard some awful rumor about it. ;-)
18:22 < misl> cu morfoh
18:25 < morfoh> misl: nah! ... I'm just off for shopping some food and things I'll need for me and the rat to survive the weekend ;)
18:26 < morfoh> moin nzg
18:26 < nzg> moin morfoh
18:26 < morfoh> mnemoc: does it answer your "?" ;)
18:26 < misl> morfoh: If it is too cold there in Berlin and you need some food, you could always eat your rat :D
18:28 < mnemoc> morfoh: yes :)
18:28 < morfoh> misl: I think I should better go to Amsterdam to eat you :D
18:28 < valentin> re
18:28 < mnemoc> misl: the rat is not good enough for more than 4h
18:28 < mnemoc> wb valentin
18:28 < valentin> LMJ: geany usable with c++ ?
18:28 < valentin> hi mnemoc
18:28 < morfoh> misl: you're fatter than the poor rat :p
18:29 < misl> morfoh: maybe, but I am not in Berlin :p
18:29 < morfoh> misl: don't worry my friend :p
18:29 < mnemoc> morfoh is alergic to dutchs
18:30 < morfoh> mnemoc: really ... didn't knew that before :p
18:30 < mnemoc> his ass will become even fatter if he ate one
18:30 < misl> then he should move, his country is right next to a whole lot of them :)
18:30 < morfoh> mnemoc: I just know one with a fat ass here
18:31 < mnemoc> misl: he is at the other border :)
18:31 < mnemoc> morfoh: misl?
18:31 < morfoh> mnemoc: you
18:31 < misl> now mnemoc I am not showing you nude pictures
18:32 < misl> now = no
18:32 < valentin> behave yourself everybody
18:32 < morfoh> now I see what you're doing all the time on noisy ...
18:33 < mnemoc> sorry uncle valentin
18:33 < morfoh> valentin: who is misbehaving ?
18:33 < morfoh> mnemoc: :)
18:33 < misl> morfoh: that's why I want to keep noisy close by :p
18:34 < misl> and I won't send it to Berlin
18:34 < misl> sorry uncle valentin
18:34 < misl> :( whatever I did with noisy I can't do it anymore
18:35 < morfoh> misl: ic ic ... now I can imagine what multimach is meaning (to be a bit more metaphoric for uncle valentin's health ;)
18:36 < morfoh> ok ... I'm off now ... cu
18:36 -!- sparc-kly [n=sparc-kl@64.237.241.61] has joined #t2
18:36 * sparc-kly brbrbr
18:59 < rxr> re
19:05 < mnemoc> wb rxr
19:05 < mnemoc> rxr: what do you have at c:/windows/Fonts ?
19:09 < rxr> mnemoc: ?
19:09 < rxr> should I boot windows now or experimental mount the NTFS ?
19:09 < mnemoc> rxr: wine :)
19:11 < rxr> mnemoc: I have not used wine the last days ...
19:11 < rxr> I no archivista box online
19:11 < rxr> and the athlon with naturally speaking as well ...
19:12 < mnemoc> ok, don't worry then
19:15 < LMJ> did you guys get my little package ?
19:17 < mnemoc> LMJ: subscribing to the mailing list is not as hard as you may think
19:17 < LMJ> i'm suscribed to t2 and t2-svn
19:17 < LMJ> since a couple of months
19:20 < mnemoc> good :)
19:21 < mnemoc> now i answer, yes... your mail arrived :)
19:22 -!- YeT2 [n=chmo@i538785FC.versanet.de] has quit ["Leaving..."]
19:22 < rxr> hehe - mnemoc you are soooo evil ,-)
19:23 < LMJ> mnemoc is evil²
19:23 < LMJ> in fact
19:23 < mnemoc> :(
19:23 < LMJ> mnemoc is the evil of the evil
19:23 < LMJ> ;)
19:23 < mnemoc> do you know how my desktop target is called?
19:23 < mnemoc> named*
19:23 < LMJ> evilmnemoc ?
19:24 < mnemoc> azazel
19:24 < misl> LMJ: better be nice to mnemoc if you want him to add your geany package ;-)
19:24 < mnemoc> http://svn.exactcode.de/mnemosyne/azazel/trunk/
19:24 < LMJ> hehe, i'm sharing it, i already have it here ;)
19:25 < mnemoc> LMJ: having packages only at your working copy will give you troubles :)
19:25 < misl> LMJ this time your the one being evel :)
19:25 < misl> evil
19:26 < mnemoc> frenchs are evil per se
19:26 < LMJ> evilmnemoc will come into my dreams and transform them to nightmares ;)
19:26 < LMJ> look at you ;)
19:26 < LMJ> lol
19:26 < misl> does anyone know the java tool Maven?
19:26 < misl> http://maven.apache.org
19:26 < mnemoc> misl: yes, and java repository
19:27 < mnemoc> s/yes/you/
19:27 < misl> I think their dependency stuff is okay
19:28 < misl> every package in the repository knows its dependencies
19:28 < misl> even multiple versions of the same packages are supported
19:29 < misl> maybe this is something for t2 as well
19:31 < misl> http://maven.apache.org/guides/introduction/introduction-to-dependency-mechanism.html
19:32 < mnemoc> misl: valentin is designing the dependencies system, you may want to talk to him
19:32 < misl> valentin: ping
19:36 < valentin> pong
19:37 < valentin> where does maven know the dependecys from ?
19:38 < misl> valentin: every package has a pom.xml which specifies its dependencies
19:38 < misl> follow the link I mentioned before
19:39 < misl> I particularly like hte dependency scope, but I would add optional as well.
19:40 < valentin> our greatest problem for now is the input, i.e. what to put into that pom.xml
19:40 < mnemoc> valentin: python deps are a mess
19:41 < valentin> since our current system does to unsharp recording
19:41 < mnemoc> are we ignoring reads at configure phase?
19:41 < misl> I compare the pom.xml with t2 .desc files
19:42 < misl> it only contains a bit more information
19:42 < mnemoc> a bit?
19:42 < valentin> misl: our main problem is that t2 deps are stage9 deps
19:42 < misl> we could start of transforming the .cache dependency information. But later this needs to be fine tuned by hand
19:42 < valentin> therefore we have paradox deps
19:44 < valentin> misl: i like the idea of distinguishing between build and runtime deps
19:44 < valentin> however we have another (independend) aspect: required or optional
19:45 < mnemoc> valentin: i think we should do somekind of dependencies reporting system
19:45 < mnemoc> valentin: exporting .cache files
19:45 < mnemoc> valentin: something always present is required
19:45 < valentin> mnemoc: ?
19:45 < mnemoc> valentin: if it wasn't reported by someone, it's optional
19:45 < misl> the big difference is that we compile from source and maven has prebuilt stuff, so circular dependencies is no problem for maven.
19:46 < mnemoc> valentin: [E] deps are runtime, optional runtime... we don't care
19:46 < valentin> misl: our dependecys would not be 'circular' if we would distinguis between build scenario (e.g toolchain vs stage 1 vs stage 5 etc)
19:47 < valentin> mnemoc: most deps are runtime and build time, but some are build time only
19:47 < valentin> there are very few that are runtime only
19:47 < rxr> this stage 5 vs. stage 9 talk brings us nothing
19:47 < valentin> runtime only would be fonts, i18n files, firmware, etcetc
19:47 < valentin> rxr: i feared you would say that :)
19:48 < valentin> know we had this topic before
19:48 < valentin> but with stage9 deps we have all those A -> B vs. B -> A deps
19:49 < mnemoc> which we can only solve building A at i and i+k stages
19:49 < mnemoc> and B after A at stage i
19:50 < valentin> mnemoc: but where do you know except hardcoding ?
19:50 < misl> mnemoc: is this dividing stage 9 in 2 stages as well?
19:50 < valentin> i'd like to drop those stages, anyway
19:51 < mnemoc> i think you can't
19:51 < valentin> why not ?
19:51 < rxr> sorry
19:51 < rxr> my ipw2200 driver corrupted
19:51 < rxr> had to reboot ...
19:51 < mnemoc> :(
19:51 < misl> valentin: I don't get the runtime part you mentioned (fonts, i18n,..)
19:51 < mnemoc> bye rxr
19:51 < rxr> only had occational firmware errors before - never a total f*ckup ..
19:51 < misl> these are specified by the target
19:51 < rxr> well, Intel code ...
19:51 < valentin> misl: runtime is stuff that is not needed to build a package
19:51 < rxr> it does not matter that we have stage 5 to 9 right now
19:51 < misl> no package needs to be dependant on those.
19:52 < rxr> when we have real dependency tracking we would have all deps including optional marks
19:52 < mnemoc> why do we worry about buildtime-only deps?
19:52 < valentin> misl: well, we do not need those deps for building
19:52 < rxr> the build system should itself use those circles to build one pass first and circular depending packages finally thereafter
19:52 < mnemoc> buildtime and runtime-only are the only important imo
19:52 < misl> mnemoc: build time only deps can be easily filtered
19:53 < misl> afterwards
19:53 < valentin> mnemoc, misl : it would be nice to have such dependency info for image postprocessing
19:53 < mnemoc> sure, but 'it would be nice' can wait, doesn't it?
19:53 < rxr> 19:48 < valentin> but with stage9 deps we have all those A -> B vs. B -> A deps
19:53 < rxr> 19:49 < mnemoc> which we can only solve building A at i and i+k stages
19:53 < rxr> ack
19:53 < valentin> damn, too many ppl talking here at one time :)
19:53 < rxr> this is real world
19:54 < rxr> crap depends on crap and we have to live with that
19:54 < valentin> yes
19:54 < valentin> so why do we need stages ?
19:54 < rxr> build as much as possible in one pass, and the ones that missed optional deps in a finalizing stage
19:54 < rxr> valentin: our rework would not have those "stages"
19:54 < rxr> we would only have some staging that involves:
19:54 < valentin> rxr: that is what i think
19:54 < rxr> - bootstrap cross toolchain
19:54 * mnemoc wants to see that graph-magic
19:55 < rxr> - cross build as much as possible
19:55 < rxr> - native build first pass
19:55 < rxr> - native build, finallizing circular deps pass
19:55 < valentin> mnemoc: the graph magic is trivial, but i need the input (which our current caches and desc do not give yet)
19:55 < rxr> valentin: those .cache are much better now
19:55 < valentin> rxr: why distinguishing between stage 2 and 3 ?
19:56 < rxr> if you hand us a A0 poster with annoying deps we will make sure the next .cache regen is denoised even more
19:56 < rxr> valentin: not distringuish
19:56 < valentin> ok
19:56 < rxr> valentin: cross build as much as possible is followed by native build the reminder that is not cross aware ..
19:57 < valentin> maybe some crap package even needs to be build thrice
19:57 < rxr> valentin: at our uni I can print A0 posters on plain paper for 10EUR or whatever ...
19:57 < rxr> hand me a 1xx MB .pdf with the graph and we have an overview
19:57 < valentin> rxr: think that would be waste of paper for there are too much deps per packages and you would not be able to see anything :/
19:57 < rxr> I have a dual core *2GHz *1.5GB RAM if you need cpu cycles
19:57 < mnemoc> python is the main failure on .cache files
19:57 < valentin> however i can try to remove transitivity
19:59 < rxr> /we dinner - cu
19:59 < misl> guten apetit
20:00 < valentin> so you propose to look at each cycle and decide which dep to mark optional manualy ?
20:00 < misl> who is you?
20:00 < rxr> of course we have to mark optional ones manually when time comes
20:01 < mnemoc> that's where i propose the reporting system
20:02 < valentin> rxr: i have an idea
20:02 < valentin> we could even do it automaticaly for existing cycles
20:02 < valentin> by assuming trunk builds and looking at the P flag
20:02 < CIA-7> rene * r15098 /trunk/package/xorg/fdclock/ (. fdclock.cache fdclock.desc): * added fdclock (2006-01-26) - A new-style example application clock
20:03 < rxr> valentin: that would only catch a few
20:03 < valentin> rxr: sure ?
20:03 < valentin> it would catch enough to resolve the cycles
20:03 < misl> also comparing early stage cache files with later stage cache file will catch some
20:03 < mnemoc> i guess only experimentation can answer
20:04 < valentin> for now i will just use [P] info to break all cycles.
20:04 < misl> difference is optional
20:04 < valentin> misl: we do not have those cache files for different stages
20:04 < misl> would it be hard to have them
20:05 < misl> we only need them once :)
20:07 < mnemoc> on the next reference build we could do some temporal hackery to generate var/adm/cache.$stagelevel/$pkg
20:08 < mnemoc> but, does that help?
20:08 < valentin> another question: in the not to far future, where will we read deps from ? still from noised caches ?
20:08 < valentin> mnemoc: maybe find even more optional stuff than by looking at [P{
20:09 < mnemoc> valentin: i imagined [DEP] $pkg $attribute ... .cache files are not that noisy this days
20:10 < valentin> mnemoc: but caches are autogenerated, whereas $attibute is tuned manually
20:10 < valentin> and recompiling a package would destroy the exiting cache file with all its knowledge inside
20:11 < mnemoc> that's why i insist in the reporting system, with a daemon generating $confdir/$pkg.cache using old $pkg.cache and trusted cache files sent to it
20:11 < mnemoc> and $confdir/$pkg.desc:[E] for runtime-only
20:12 < valentin> mnemoc: that is what i think about, too. only that i'd rather manage the result in another file than .cache
20:12 < mnemoc> .deps ?
20:12 < valentin> like that, yes
20:13 < mnemoc> var/adm/cache/$pkg reflects the build, $confdir/$pkg.whatever is just build time, last state and a list of deps. unrelated to any build
20:14 < mnemoc> cp var/adm/cache/$pkg $confdir/$pkg.cache is just the current method to populate them
20:14 < mnemoc> we could also hardcode attributes at [E]
20:15 < mnemoc> $pkg.desc:[E] -foo meaning foo is an optional dep of $pkg
20:15 < misl> I am off guys, cu tomorrow
20:15 < mnemoc> giving the final hints to the scheduler
20:15 < mnemoc> cu misl :)
20:15 < misl> boy did I start a discussion over here ;-)
20:15 < misl> bye bye
20:15 < mnemoc> restart :)
20:16 < misl> mnemoc:ack
20:16 -!- misl [n=chatzill@82-217-66-150.cable.quicknet.nl] has quit ["Chatzilla 0.9.69.3 [Firefox 1.5/2005111116]"]
20:16 < CIA-7> aldas * r15099 /trunk/package/develop/gambas/gambas.desc: * updated gambas (1.0.13 -> 1.0.14)
20:22 < CIA-7> amery * r15100 /branches/2.1-test/package/emulators/wine/wine.desc: * merged 14927 from trunk: updated wine (0.9.5 -> 0.9.6)
20:23 < CIA-7> aldas * r15101 /trunk/package/gnustep/gorm/gorm.desc: * updated gorm (1.0.1 -> 1.0.4)
20:27 < CIA-7> aldas * r15102 /trunk/package/editors/ne/ (ne.conf ne.desc page-up-down.patch): * updated ne (1.41 -> 1.42)
20:36 < rxr> hm,
20:36 < mnemoc> ?
20:37 < rxr> either evas or my Evas helper code has endian or so issues on x86-64
20:37 < rxr> mostly used it on PowerPC so far
20:37 < rxr> evas_software_x11_test looks fine
20:37 < rxr> so mostly my own code I ear
20:38 < rxr> ah - just my screenshot code ,-)
20:38 < rxr> ok - will see tomorrow
20:38 < rxr> n8 all
20:39 < CIA-7> aldas * r15103 /trunk/package/gnome2/gnome-volume-manager/gnome-volume-manager.desc: * updated gnome-volume-manager (1.5.10 -> 1.5.11)
20:40 < valentin> gn8 rxr
20:41 < mnemoc> gn8 rxr
20:43 < LMJ> n8 too
20:43 < LMJ> i'm off
20:47 < Bart-> me\ 2 off
20:48 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has quit ["Who's General Failure and why he's reading my disk?"]
21:11 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has joined #t2
21:11 < Bart-> re
21:12 < Bart-> mnemoc, is there any debug-mode in runit? This pig stalls after "Entering stage 2" but logs are rather clean
21:13 < mnemoc> Bart-: getty services are failing probably, hack some debug stuff inside /etc/runit/getty-$i/run
21:14 < mnemoc> Bart-: are they linked at /service right?
21:14 < Bart-> I'll check
21:15 < Bart-> actually /etc/runit/getty-1/run is a /bin/sh script
21:15 < mnemoc> of course
21:15 -!- axion_ [n=axion@cpe-24-195-94-82.nycap.res.rr.com] has quit [Read error: 110 (Connection timed out)]
21:16 < Bart-> I have agetty[532]: /dev/vc/1: No such file or directory in 'messages' but: 1) I had it before and T2 worked, 2) These devices exists
21:16 < mnemoc> uhm
21:17 < Bart-> or at least they are created when udevstart is run in chroot, I need to modify scripts to do 'ls /dev/*' during boot
21:18 < mnemoc> ls /dev/* > /dev/console
21:19 < mnemoc> stdout of stage 2 services is passed to /service/foo/log/run if present
21:19 < mnemoc> if not, dropped
21:20 < Bart-> there are no files or dirs lacced 'log'
21:20 < Bart-> *called ;]
21:22 < mnemoc> getty don't suppose to need loggers
21:22 < mnemoc> that's why i told you to explicitly redirect debug flags to /dev/console
21:32 < Bart-> strange things happen in this crap, so there is no chance to see what runit is doing?
21:32 < mnemoc> nope
21:33 < mnemoc> add a call to sh on etc/runit/2 before the services
21:33 < mnemoc> then try to exec ./run scripts by hand and watch
21:34 < mnemoc> right after 'entering stage 2' you should get your getties
21:34 < mnemoc> but for some reason that's not happening
21:35 < mnemoc> to you
21:39 < Bart-> ok, ready to test, rebooting
21:39 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has quit ["Who's General Failure and why he's reading my disk?"]
21:45 < CIA-7> aldas * r15104 /trunk/package/network/nss_ldap/nss_ldap.desc: * updated nss_ldap (247 -> 248)
21:47 < CIA-7> amery * r15105 /branches/2.1-test/package/base/gcc/parse-config: * merged 15066 from trunk: changed gcc to not use stack protector for wine
21:56 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has joined #t2
21:56 < Bart-> mnemoc, no way to run sh: "No job control in this stage" and then reboots
21:56 < mnemoc> o_O
21:57 < mnemoc> Bart-: next try, explicitly turn runit/2 into your getty
21:57 < mnemoc> exec /sbin/agetty -L -i -I '\012\015\012Maintenance Console:\012' 38400 vc/1 linux :)
21:57 < mnemoc> or console instead of vc/1
21:58 < mnemoc> ldd /sbin/agetty before that exec can be usefull
21:58 < mnemoc> -l
22:09 < Bart-> last try for today since it's late, cu
22:09 -!- Bart- [n=bart@osiris.osk.enformatic.pl] has quit ["Who's General Failure and why he's reading my disk?"]
22:23 < morfoh> re
22:24 < mnemoc> wb morfoh
22:43 < morfoh> thx mnemoc :)
23:02 -!- sparc-kl1 [n=sparc-kl@64.237.254.13] has joined #t2
23:11 < CIA-7> jsaw * r15106 /trunk/package/office/libofx/libofx.conf:
23:11 < CIA-7> * fix for libdir != *lib
23:11 < CIA-7> * copyright update
23:13 -!- menomc [n=amery@200.75.27.27] has joined #t2
23:13 -!- mnemoc [n=amery@200.75.27.19] has quit [Nick collision from services.]
23:15 < CIA-7> jsaw * r15107 /trunk/package/office/libofx/ (gcc4.patch libofx.desc no-hardcoded-libdir.patch): * missing parts of r15106
23:15 -!- menomc is now known as mnemoc
23:17 -!- sparc-kly [n=sparc-kl@64.237.241.61] has quit [Read error: 110 (Connection timed out)]
23:35 < CIA-7> jsaw * r15108 /trunk/package/security/libchipcard/ (64bit.patch libchipcard.conf libchipcard.desc):
23:35 < CIA-7> * 64bit fix
23:35 < CIA-7> * copyright update
23:37 -!- nzg [n=tschmidt@p54AFE524.dip.t-dialin.net] has quit ["Leaving"]
23:45 < CIA-7> jsaw * r15109 /trunk/package/security/libchipcard2/ (libchipcard2.conf libchipcard2.desc):
23:45 < CIA-7> * update libchipcard2 (1.9.19beta -> 1.9.20beta)
23:45 < CIA-7> * fixes for libdir != *lib
--- Log closed Sun Jan 29 00:00:44 2006