--- Log opened Thu Dec 16 00:00:05 2004 --- Day changed Thu Dec 16 2004 00:00 < rxr> "I do not work together with the Linux kernel developmers - so cdrecord" 00:00 < daja77> yeah sucks 00:00 < rxr> "does not work correcty. Please boot another operating system to burn the data" 00:00 < daja77> patched the 2.6.9 headers to get it compiled ... 00:00 < mnemoc> is that message a joke? 00:01 < rxr> "- Thanks. And do not write mails to me - I'l not answer." 00:01 < rxr> mine - yes. daja77 is fact 00:01 < daja77> yep cdrecord just writes that to me, because it can't detect scsi device 00:02 < daja77> ok it is an ide burner 00:02 < rxr> it writes a lot of noise ... 00:02 < mnemoc> rxR: i'm trying svn-HEAD to see if the leak is gone, but it fails to apply package/develop/subversion/svn_sm.patch, should i migrate it? :) 00:02 < mnemoc> port it* 00:02 < daja77> and then he is whining about patched versions shipped with distributions ... 00:03 < rxr> mnemoc: for me it is ok to get rid of this -sm.patch burden ... 00:03 < rxr> no idea what the others of you think ... 00:03 < rxr> I would patch his name in the output as well ... 00:03 < rxr> I have no problem with people on t2@exactcode.de ask cdrecord questions due to mmy 00:03 < rxr> patches 00:03 < mnemoc> o_O 00:03 < rxr> - "Do not send mail to jschilly@gmd.s.th.de" 00:04 < rxr> + "For questions and support write to t2@exactcode.de" 00:04 < rxr> what do you think ? ;-) 00:04 < daja77> wah! have to reboot to get this shit working 00:05 * daja77 cursing ... 00:05 < rxr> ? 00:05 < rxr> (reboot into solars ?) 00:05 < daja77> lol 00:05 * mnemoc moving home 00:06 < daja77> no you need to put hdc=scsi crap into kernel boot line to get these devices detected by cdrecod *grml* 00:06 < rxr> ehm 00:06 < rxr> are you in 2.6 ? 00:06 < daja77> yes 00:06 < rxr> idescsi is broken there 00:06 < rxr> at least the last time I checked ... 00:07 < daja77> argh 00:07 < rxr> you need and want to access via ATAPI 00:07 < daja77> so no chance 00:07 < rxr> ATAPI works perfectly (at least here) - better than ever 00:07 < daja77> ok i'll copy the image to the laptop which runs on 2.4 00:07 < daja77> how do iget cdrecord to recognize that 00:07 < daja77> cdrtools recompile would be ok 00:08 < daja77> Cdrecord-Clone 2.01 <- ... 00:10 < rxr> btw this -Clone confused me often - it makes me feeling I use a forked, patched version ... 00:13 < daja77> ah cdrecord dev=ATAPI -scanbus shows some results ;) 00:17 < daja77> nice burned without reboot 00:45 < rxr> There where 684 changesets with 1304 lines of commit messages. 00:45 < rxr> Approximately 373 packages got updates, 195 fixed, 373 added and 51 removed. 00:45 < rxr> Arround 16 improvements have been committed. 00:45 < rxr> Commits by contributor: 00:45 < rxr> 422 rene 00:45 < rxr> 127 jsaw 00:45 < rxr> 100 mnemoc 00:45 < rxr> 25 jeru 00:45 < rxr> 8 valentin 00:45 < rxr> 1 susan 00:45 < rxr> 1 martin 00:45 < rxr> so - redone my branch commit parser ... 00:46 < rxr> valentin: you need to commit a bit more ;-) 00:46 < rxr> just joking of course ... 00:47 * daja77 uploaded rock-theme stuff 00:55 < rxr> revisited ANNOUNCEMENT to be send in a second 00:58 < rxr> out 01:08 < mnemoc> :) 01:09 < rxr> read it ? 01:12 < mnemoc> i'm on it 01:12 < rxr> so I hope you now have a good tempalte ... 01:13 < mnemoc> !> buildcheck: checking installation... 01:13 < mnemoc> !> buildcheck: autoconf version 2.58 found. 01:13 < mnemoc> !> This version of autoconf is broken. Please .. 01:13 < mnemoc> !> least autoconf 2.59 or downgrade to version .. 01:13 < mnemoc> !> is known to work. 01:13 * mnemoc kicks svn-HEAD 01:40 < rxr> mnemoc: I guess we want a news page as frontpage ? 01:42 < mnemoc> oh, i forgot that 01:43 < rxr> I just asked if we want ... 01:43 < mnemoc> how are you managing www.exeactcode.de? CMS? raw? raw on svn? 01:43 < rxr> svn 01:44 < rxr> "svn the tool for everything but delivering fresh beer ..." 01:45 < rxr> oh - I have a trailing i on your familly name :-( 01:45 < rxr> damn 01:46 < mnemoc> qwhere? 01:46 < rxr> in both announce mails? 01:47 < mnemoc> Alejandro Mery Pellegrini <--- this is ok 01:48 < rxr> really? I remebered Pellegrin ... 01:48 < rxr> puh 01:48 < rxr> me stoning this into the head ... 01:48 < mnemoc> :D 01:49 < mnemoc> both exist :) 01:55 < rxr> http://www.exactcode.de/t2/news.html 01:55 < rxr> comments ? 01:56 < mnemoc> let' see 01:56 < mnemoc> T2 2.1.0-alphaa 01:57 < rxr> fixed in my copy 02:02 < rxr> no more comment ? 02:03 < rxr> ok - site is gone 02:03 < rxr> exchanged them 02:03 < rxr> index -> about 02:03 < rxr> news -> index 02:05 -!- mnemoc_ [~amery@200.75.27.42] has joined #t2 02:05 < rxr> oh rehi mnemoc 02:05 < mnemoc_> :( 02:06 < mnemoc_> it looks fine btw 02:06 < rxr> 02:02 < rxr> no more comment ? 02:06 < rxr> 02:03 < rxr> ok - site is gone 02:06 < rxr> 02:03 < rxr> exchanged them 02:06 < rxr> 02:03 < rxr> index -> about 02:06 < rxr> 02:03 < rxr> news -> index 02:06 < mnemoc_> good 02:07 < mnemoc_> some statistics can be cool too 02:08 < rxr> what kind of statistics? 02:13 < mnemoc_> packages 02:13 < mnemoc_> rate of success building on main architectures 02:13 < mnemoc_> build time of reference :) 02:14 < rxr> ah - yes - will do "soon" 02:14 < rxr> as well as implanting possible version updates into the package matrix ... 02:15 < mnemoc_> and a "search on flist" :) 02:15 -!- mnemoc [~amery@200.75.27.55] has quit [Read error: 110 (Connection timed out)] 02:15 -!- mnemoc_ is now known as mnemoc 02:15 < mnemoc> bye mnemoc 02:15 -!- sparc-kly [~sparc@64.237.130.173] has joined #t2 02:16 < rxr> hi sparc-kly 02:16 < mnemoc> rxr: can you give me some advice about: http://www.geeks.cl/~amery/svn-11581-errorlog 02:17 < sparc-kly> im in VNC :D 02:17 < mnemoc> :) 02:17 < mnemoc> ssh is lighter :) 02:18 < sparc-kly> hehehe 02:18 < rxr> libtool: compile: unable to infer tagged configuration 02:18 < rxr> libtool: compile: specify a tag with `--tag' 02:18 < rxr> hm - no off hand idea 02:18 < sparc-kly> vnc run fine :D 02:18 < rxr> I saw stuff like that - but I do not remeber what it was ... 02:19 < mnemoc> :\ 02:19 < sparc-kly> i need put vnc server in Sun Ultra 2 . solaris 7 and CDE heheheh 02:21 < rxr> ok /me in bed ... 02:25 < rxr> "Opinion: Thanks, Microsoft" 02:25 < rxr> ^- says: http://www.osnews.com/story.php?news_id=9172 02:26 < rxr> ok - cu ppl 02:26 [Users #t2] 02:26 [ _Ragnar_] [ daja77] [ mnemoc ] [ nzg] [ sparc-kly] 02:26 [ CIA-9 ] [ jsaw ] [ nullslack] [ rxr] [ valentin ] 02:26 -!- Irssi: #t2: Total of 10 nicks [0 ops, 0 halfops, 0 voices, 10 normal] 02:31 -!- rxr_ [~rene@p213.54.196.182.tisdip.tiscali.de] has joined #t2 02:31 -!- Topic for #t2: T2 | The system development environment | http://www.exactcode.de/t2 02:31 -!- Topic set by rxr [] [Wed Dec 15 17:22:31 2004] 02:31 [Users #t2] 02:31 [ _Ragnar_] [ daja77] [ mnemoc ] [ nzg] [ rxr_ ] [ valentin] 02:31 [ CIA-9 ] [ jsaw ] [ nullslack] [ rxr] [ sparc-kly] 02:31 -!- Irssi: #t2: Total of 11 nicks [0 ops, 0 halfops, 0 voices, 11 normal] 02:31 -!- Channel #t2 created Sun Aug 8 21:15:33 2004 02:31 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup 02:31 < rxr_> rehi 02:31 -!- Irssi: Join to #t2 was synced in 17 secs 02:32 -!- rxr [~rene@p213.54.215.21.tisdip.tiscali.de] has quit [Nick collision from services.] 02:32 -!- You're now known as rxr 02:32 < rxr> so me back 02:32 < rxr> now I'll not get reconnects in the middle fo the day anymore 02:32 < rxr> till the morning ;-) 02:33 < mnemoc> he :) 06:18 -!- Netsplit saberhagen.freenode.net <-> irc.freenode.net quits: CIA-9 06:19 -!- Netsplit over, joins: CIA-9 09:15 < rxr> moin 10:02 < rxr> re 10:02 * rxr updating Mac OS X to 10.3.7 .... 10:02 < rxr> and showering (update slow as usual) 10:05 < rxr> hehe "linux handbook" on google gives the old rocklinux-consulting site as top hit 10:53 < rxr> rebooting to please osx ... 11:00 < rxr> re 11:00 < daja77> 8min for booting ... 11:01 < rxr> hehe no 11:01 < rxr> 8min for loading the (textil) washing machine 11:01 < daja77> :) 11:01 < rxr> and then logging into linux and X.org ... 11:24 < rxr> why is this a warning: 11:24 < rxr> ./avision.h:74: warning: comma at end of enumerator list 11:25 < rxr> I consider this an error ... :-( 13:01 < rxr> damn - my fault - this was due to I set the USB transfer timeout much lower ... 13:01 < rxr> for testing weeks ago ... 13:06 -!- gfdsa [~kvirc@213.156.44.71] has joined #t2 13:06 < gfdsa> hi all! 13:06 < gfdsa> a question if someone is here? 13:06 < rxr> hi gfdsa 13:07 < gfdsa> oh, cool, hi! 13:07 < rxr> welcome ;-) 13:07 < gfdsa> thanks! 13:07 < sparc-kly> moin all 13:07 < gfdsa> hi sparc-kly 13:07 < rxr> hey, hi sparc-kly 13:07 < gfdsa> so, may i ask? 13:07 < rxr> sure 13:07 < gfdsa> its abot building the toolchain 13:08 < gfdsa> is there an option to build with distcc? 13:09 < rxr> well - not yet 13:09 < rxr> but ccache is already there 13:09 < rxr> historically we had some custom shell, scripted cluster build mode - which due to changes does not scale well anymore (and has some other problems) 13:09 < rxr> we plan to add distcc support in 2.2 - maybe even experimetally for 2.1 13:10 < gfdsa> i see 13:10 < rxr> recently jsaw showed some interest in it - no idea if he has already started the implementation 13:10 < gfdsa> what if ill try to substitute cc with distcc? 13:10 < gfdsa> it will defenetly cause build problems, but im ready to dig in 13:11 < rxr> well, this will only work until the build chroot's into the sandbox 13:11 < gfdsa> just where to substitute :-) 13:11 < gfdsa> thats for sure 13:11 < gfdsa> i mean the toolchain only 13:11 < gfdsa> building glibc takes too long 13:11 < rxr> so you'll loose compiling distributed for the big part of the build 13:12 < gfdsa> i see it as not so big problem , to inject distcc into the target chroot 13:12 < rxr> hm - compared to the rest (when you build a desktop) glibc is not that long ... - there ix X and all the kde and gnome stuff ... - that takes much longer 13:12 < gfdsa> its a simple application 13:12 < gfdsa> no, i dont compile desktop, i do appliances 13:12 < rxr> well - there are some obstacles 13:12 < gfdsa> which? 13:13 < rxr> well - ok no big ones ... 13:14 < gfdsa> :-) 13:14 < gfdsa> i see 13:14 < rxr> have you already looked into building t2 normally? 13:14 < gfdsa> sure, today for the firswt time, after ive seen release on the fm 13:14 < rxr> adding distcc just as proof of concept should not be that hard ... 13:14 < gfdsa> yep, exactly 13:15 < rxr> we have gcc wrapper 13:15 < gfdsa> i have my own kinda-framework with distcc, so i know the curstimances (?always have trouble spelling this one?) 13:15 < gfdsa> oh, great 13:15 < rxr> that e.g. transform the optimization parameters generically 13:16 < gfdsa> i see, can you point me to where its defined? 13:16 < gfdsa> i mean the wrapper? 13:16 < rxr> AFAICS you could inject distcc simillar as we do with ccache - using the GCC_WRAPPER_OTHER (or what it was called= 13:16 < gfdsa> does it look in ENV? 13:17 < rxr> the wrapper does look into ENV variables to perform actions 13:17 < gfdsa> but GCC_WRAPPER_OTHER, will it be fetched by scripts from ENV? 13:17 < rxr> the gcc wrapper is an compiled binary executable 13:17 < rxr> it will look into env vars and perform the actions as defined by those 13:17 < gfdsa> mm, intresting.. 13:18 < rxr> I point you to the place where it is done for ccache 13:18 < gfdsa> any docs on what it accepts from ENV? 13:18 < gfdsa> oh, great 13:18 < rxr> partially in the handbook 13:18 < gfdsa> where? 13:18 < rxr> mom 13:18 < rxr> you ask to fast ... 13:18 < gfdsa> :-) 13:18 < gfdsa> type too fast, thats right 13:19 < rxr> package/develop/ccache/parse-config 13:19 < rxr> in the t2 source 13:19 < rxr> this is how ccache does it - and as I notice right now it is a bit broken 13:19 < rxr> no wonder that it did not speed building mozilla here 13:20 < rxr> (one should review every line when one forks from rock linux ...) 13:20 < rxr> ok - so here is the wrapper source with the comments which vars it evaluates 13:20 < gfdsa> shorter last one to "one should review every line when one forks" 13:21 < rxr> misc/tools-source/cmd_wrapper.c 13:21 < gfdsa> ok, ill look what is there, may be ill scream for help soon again, thanks a lot, you saved me a lot of time 13:21 < rxr> mom 13:21 < rxr> one more link 13:21 < rxr> http://dl.exactcode.de/t2-handbook/html/t2-handbookch25.html 13:22 < rxr> this is a short overview over the wrappers 13:23 < gfdsa> one more thanks! 13:23 < rxr> have fun ;-) 13:24 < gfdsa> i will, thats for sure! 13:25 < rxr> mnemoc: I try to improve the ccache handling - we could do an interim release in the beta series if it works out ... 13:26 < rxr> (as I just noticed - thanks to gfdsa asking - the g++ compiler is not wrapper) 13:27 < gfdsa> glad to be helpful, btw, what is the meaning of compilers names in the ncurses screens? 13:27 < gfdsa> transparently filenames? 13:27 < rxr> gfdsa: "compiler names in ncurses screens" ? What ? 13:28 < gfdsa> a sec, ill look how you named them 13:29 < gfdsa> its "Selecting compiler" section 13:29 < gfdsa> with choices of CC CPP KCC 13:29 < gfdsa> frotran for some reason.. 13:30 < rxr> yes 13:30 < rxr> what is your question ? 13:33 < gfdsa> if i put "my-gcc" as CC there, will it be something like $a_path/my-gcc ? 13:33 < rxr> well - you can not put my-gcc there - it will only allow the selection of the compilers as shipped with t2 13:34 < rxr> compiler packages can hook in there, those can be selected 13:34 < rxr> e.g. icc ... 13:34 < gfdsa> i see 13:34 < rxr> this is because t2 rebuilds itself complete, so we so far only used compilers that are provided as package descriptions - e.g. nothing you have already in your running system 13:52 < rxr> ANNOUNCEMENT: Our ccache implementation is horrible broken (as inherited from ROCK Linux ...) 13:57 < rxr> oh my is this rediculous 13:57 < rxr> gfdsa: big thanks for pointing me to the ccache code ... 13:57 < rxr> although I should code s.th. else right now .. 14:01 < gfdsa> do you mean as a package? 14:01 < rxr> no the usage 14:02 < gfdsa> shouldnt it be more/less transparent? 14:04 < rxr> yes - but the wrapper is not utilized correctly and the wrapper is not providing the functionality at the right place 14:04 < rxr> I just fix this up .. 14:06 < gfdsa> well, im still trying to figure out how the files you pointed me to are related to the things i need 14:06 < rxr> well aside that it is slightly broken 14:06 < rxr> we have a wrapper that is called instead of gcc - which can manipluation the options and call other programs instead 14:07 < rxr> via this OTHER hook the ccache is injected - and distcc could be too 14:07 < rxr> I fix it up and then add distcc example code 14:08 < mnemoc> sure 14:08 < mnemoc> rxr: as we plan to release 2.1 before 2.0 anniversary we can release weekly betas :) 14:09 < rxr> hehe 14:09 < rxr> seems we can have -beta2 tonight .. 14:10 < mnemoc> (moin!) 14:10 < mnemoc> =) 14:10 < rxr> .oO(why does the wrapper code need to look that ugly ...) 14:10 < rxr> yeah - hi mnemoc 14:11 < mnemoc> i think one day between betas is a bit 'non serius' 14:11 < gfdsa> ok guys, ive got the idea of wrapper/gcc + cmd_wrapper, now, where is the list of commands? :-) 14:11 < rxr> well - since my tested showed nearly no speedup when buliding Mozilla on my iBook I have not used it since then ... 14:12 < rxr> I should have better taken a look at that time why mozilla was not speed up ... 14:13 < rxr> well - all you need should be in the links I posted above 14:14 < gfdsa> wow, is it binded to packages/base/gcc ? 14:15 -!- gfdsa [~kvirc@213.156.44.71] has quit ["cya later"] 14:16 -!- gfdsa [~kvirc@213.156.44.71] has joined #t2 14:32 < gfdsa> ok, forgive my dumbness, but how do you use (intend to use ) ccache in toolchain build? 14:32 < gfdsa> i mean where in these files ccache should be mentioned? 14:33 < rxr> ccache is mentioned in the base/ 14:33 < rxr> package/develop/ccache/parse-config 14:33 * rxr lunch - cu 14:33 < mnemoc> cu rxr 14:50 < mnemoc> build build build 14:50 < rxr> re 14:51 < rxr> mnemoc: build what ? 14:51 < mnemoc> t2, what else? :) 14:52 < rxr> I you (he she )? 14:52 < mnemoc> is a mantra :) 14:52 < mnemoc> build build build 14:52 < mnemoc> test test test 14:52 < mnemoc> fix fix fix 14:52 < rxr> ok ;-) 14:53 < mnemoc> like: Hare Krishna Hare Krishna Krishna Krishna Hare Hare 14:53 < mnemoc> Hare Rama Hare Rama Rama Rama Hare Hare 14:54 < mnemoc> s/Hare/Build/g s/Krishna/Test/g s/Rama/Fix/g :) 15:16 < rxr> hm - the wrapper stuff does indeed does "something" - even for C++ 15:16 < rxr> but it is highly unintuitive and hard to follow the path of execution ... 15:16 < gfdsa> its rather terrible... -) 15:16 < gfdsa> harihari 15:17 < rxr> and since first ccache is executed and then the wrapper run again the output is associated with the wrong gcc options - resulting in the possibility that the wrong data is pulled out of the cache ... 15:17 < rxr> e.g. (AFAICS) doing this: 15:18 < rxr> GCC_WRAPPER_INSERT="-mcpu=i686" 15:18 < rxr> gcc ... 15:18 < rxr> GCC_WRAPPER_INSERT="-mcpu=powerpc" 15:18 < rxr> gcc ... 15:18 < rxr> will give you the cached result from the first with the current code ... 15:18 < gfdsa> rxr: thanks for helping me, but i reconsiderred the usefulness of this project for me, its slightly not my direction, its too "distributive" alike... 15:19 < rxr> well - what do you exactly need? 15:19 < gfdsa> i need a framework for building embedded-alike systems 15:19 < rxr> note: not that you normally mix x86 and powerpc as I did above, just to expalain the crazzyness - just using pentium and pentium4 is the usual example ... 15:19 < rxr> gfdsa: well t2 is that ... 15:20 < gfdsa> not really 15:20 < rxr> we build lightweight embedded systems with it ... 15:20 -!- madtux [~mike@200.91.101.101] has joined #t2 15:20 < madtux> hello 15:20 < rxr> and it is even possible to build 4MB sized PDA images .... 15:20 < gfdsa> say we need glibc on the platform, not ulibc 15:20 < rxr> gfdsa: what do you miss or ... 15:20 < rxr> gfdsa: we have uclibc and dietlibc ... 15:20 < gfdsa> i said not ulibc 15:21 < gfdsa> do you have a procedure not to copy timezones, locales, mans, includes blabla from glibc to target? 15:21 < gfdsa> ars? 15:21 < gfdsa> and, its too complicated for my needs 15:22 < rxr> oh - overread the not 15:22 < gfdsa> packaging system for example ... 15:22 < rxr> but we have glibc still - what do you miss ? 15:22 < gfdsa> i dont need any 15:22 < mnemoc> gfdsa: you can use flistdel on target's pkg_glibc.conf 15:22 < rxr> you need none or libc-avr or so ? 15:22 < gfdsa> installations procedures that strip all that stuff 15:23 < gfdsa> mnemoc: you mean describing which files to copy? 15:23 < gfdsa> libc-avr? 15:23 < mnemoc> flistdel is a pattern to remove stuff from file list 15:23 < mnemoc> that filelist is used to package 15:23 < gfdsa> mnemoc: i see, its an option 15:23 < rxr> for example for freestanding Atmel Atmega embedded systems 15:25 < gfdsa> well, all i needed was to substitute gcc with distcc for building the toolchain, i spent an hour, didnt find how to, take someone not related to rock-linux and show him that $x=$y .. 15:25 < gfdsa> and taht the first thing i need 15:25 < gfdsa> IMO, its too "automated" 15:25 < gfdsa> too common 15:26 < mnemoc> gfdsa: you don't substitute gcc with distcc, you prefix GCC with distcc, like it's done with ccache 15:27 < gfdsa> mnemoc: and it runs distcc with ARGV[1] "gcc" ? 15:27 < rxr> yes 15:27 < gfdsa> distcc gcc dummy.c? 15:27 < rxr> ack 15:28 < gfdsa> and if its not gcc but /opt/toolchain/bin/some-weired-hacked-my-gcc? 15:28 < gfdsa> will wrapper handle it? 15:28 < rxr> then you use that 15:28 < rxr> well - to be honest t2 is not designed to inject prebuild stuff from the host 15:29 < gfdsa> not that i need it, just examples... 15:29 < gfdsa> why should there be a wrpaper? 15:29 < rxr> well - if you include some hacked gcc into t2 which is than build by t2 you wi 15:30 < rxr> would inject that name into the distcc argument 15:30 < gfdsa> from all the parse-config ive seen till now, it only complicates the things i need to do 15:30 < rxr> distcc your-special-cc hello.c 15:30 < rxr> otherwise distcc would not know which cc to call ... 15:30 < rxr> and would fall back to gcc or so ... 15:30 < rxr> and it would sometimes not able to differentiate between C or not C, e.g. C++ 15:31 < rxr> when you do: 15:31 < rxr> g++ hello.c <- c++ file 15:31 < rxr> and try distcc hello.c 15:31 < rxr> the executed gcc would fail due to C++ source .... 15:31 < gfdsa> i dont get the point, sorry 15:32 < rxr> mnemoc: I change the OTHERS handling to not call it in between but use it for the final execution at the end ... 15:32 < rxr> NAME 15:32 < rxr> distcc - distributed C/C++/ObjC compiler 15:32 < rxr> SYNOPSIS 15:32 < rxr> distcc [COMPILER OPTIONS] 15:32 < rxr> man distcc 15:33 < gfdsa> i dont need to man distcc, i know what it does, i just dont understand what you are trying to explain.. 15:34 < rxr> let use summarize: 15:34 < rxr> 15:25 < gfdsa> well, all i needed was to substitute gcc with distcc for 15:34 < rxr> building the toolchain, i spent an hour, didnt find how to, take 15:34 < rxr> someone not related to rock-linux and show him that $x=$y .. 15:34 < rxr> 15:25 < gfdsa> and taht the first thing i need 15:34 < rxr> 15:25 < gfdsa> IMO, its too "automated" 15:34 < rxr> 15:25 < gfdsa> too common 15:34 < rxr> 15:26 < mnemoc> gfdsa: you don't substitute gcc with distcc, you prefix GCC 15:34 < rxr> with distcc, like it's done with ccache 15:34 < rxr> 15:27 < gfdsa> mnemoc: and it runs distcc with ARGV[1] "gcc" ? 15:34 < rxr> 15:27 < rxr> yes 15:34 < rxr> 15:27 < gfdsa> distcc gcc dummy.c? 15:34 < gfdsa> what i do understand, is the idea to have a description file + config file +someotherfile, for just adding a compiler 15:34 < rxr> I just wanted to explain why supplying a compiler as first arguemnt is important 15:35 < gfdsa> hm, it is important, no doubt 15:35 < gfdsa> may be now i understand you more clearly 15:35 < gfdsa> so how do i "prefix" gcc? 15:35 < mnemoc> gfdsa: one of our goals to t2-2.2 is transparent use of distcc to do cluster builds 15:36 < rxr> gfdsa: var_append GCC_WRAPPER ":" "distcc" 15:36 < gfdsa> rxr: where? 15:36 < gfdsa> in packege/gcc/? 15:36 < rxr> well - if you just want it hardcode it somewhere ... 15:37 < gfdsa> :-) 15:37 < rxr> e.g. in gcc/parse-config 15:37 < rxr> or config/parse-config 15:37 < gfdsa> predictable... 15:37 < gfdsa> oh, config/parse looks better 15:37 < mnemoc> config/parse-config is the first to be loaded 15:38 < gfdsa> but still, what builds it affects? i need to maintain 5 build simulationsly, and all have diffrent building clusters 15:38 < mnemoc> so distcc will be the 'last' to be called on nested executing 15:38 < rxr> scripts/parse-config I meant ... 15:38 < mnemoc> yes :) 15:38 < gfdsa> mnemoc: if it is the first, then what will override the vars? 15:38 < mnemoc> config/parse-config doens't exist :p 15:38 < rxr> yes - I meant scripts 15:39 < gfdsa> it does not 15:39 < mnemoc> grep -R "var_append GCC_WRAPPER" * 15:39 < gfdsa> did already 15:42 < mnemoc> rxr: rxvt-unicode-4.4 is no longer available, /me updating to 4.6 15:42 < rxr> 4.6? 15:43 < mnemoc> yep 15:43 < mnemoc> http://dist.schmorp.de/rxvt-unicode/ 15:47 < mnemoc> uhm... svk is doing something awful.... 15:50 * rxr performing the rxvt update 15:51 < mnemoc> better 15:52 < mnemoc> my branch got somehow corrupted with yestardays leakings 15:55 < mnemoc> rxr: http://freshmeat.net/projects/t2/?branch_id=38672&release_id=181818 <--- notice something :D 15:55 < mnemoc> +? 15:56 < rxr> yep - typo fixed 15:57 < mnemoc> "Jeff" thought that was his typo editing the message i sent :p 15:57 < rxr> hehe 15:57 < CIA-9> rene * r5078 /trunk/package/x11/rxvt-unicode/rxvt-unicode.desc: * updated rxvt-unicode (4.4 -> 4.6) 15:58 < daja77> the package you created for that debian guy? 15:58 < rxr> yep 15:58 < rxr> but I started to use it ... 15:58 < rxr> some 3.x or early 4.x version was great - all my chars worked just fine 15:58 < rxr> however recent 4.x version do not anymore ... 15:59 < rxr> I need to investigate or rereview the terminal emulators .. 15:59 < daja77> you still have no t-shirt *gg* 16:00 < madtux> t-shirt?! where?! i want?! 16:00 < mnemoc> rxr: 4.6 still weird? 16:00 < daja77> madtux: debian t-shirt? 16:00 < madtux> yuk 16:00 < madtux> though i got several somewhere in my closet 16:00 < daja77> *gg* 16:00 < madtux> i want t2 tshirts :) 16:00 < rxr> madtux: print them 16:00 < mnemoc> madtux: what about the t-shirts with the daemon? 16:01 < madtux> mnemoc: yes people have given me *bsd tshirts too 16:01 < madtux> rxr: gimme some logos :) 16:01 < mnemoc> T^2 logo is nice 16:01 * daja77 wants a mad tux t-shirt 16:01 < rxr> madtux: some debina guy tought he is smarter than we and annoyed me over an hour at the cebit ... finally he wanted that I package his app there ... and annoyed be further with terminal and perl unicode wchar problems ... 16:02 < madtux> rxr: and the guy still lives? 16:02 < rxr> but finally he said we have done well standing his monologe and wanted to hand a debian t-shirt to me - however I had not time to collect it during the next days at cebit ... 16:03 < mnemoc> he ate rene's tee mug :) 16:03 < madtux> :) 16:03 < rxr> nah - I do not hit people - especially not when they sit in the gcc streeing comitte ... 16:03 < madtux> daja77: what for? 16:03 < madtux> oh.. 16:03 < madtux> :) 16:03 < mnemoc> =) 16:04 < daja77> that might be a reason to hit them ... 16:04 < rxr> hehe 16:04 < rxr> especially when he does non ANSI C++ stuff in his rxvt-unicode thing .. 16:04 < rxr> but he already fixed that around 4.0 or so ... 16:04 < daja77> madtux: fun :) 16:05 < madtux> daja77: we'll work something out for march :) 16:05 < daja77> ;) 16:09 < rxr> well freshmeat shows up quite nicely in the stats ... 16:09 < rxr> now if we could get the distrowatch entry some day ... 16:09 < madtux> :) 16:21 < CIA-9> rene * r5079 /trunk/package/develop/ccache/parse-config: 16:21 < CIA-9> * fixed (and improved) the ccache parse-config matching of the 16:21 < CIA-9> cache hits and misses status output 16:21 < CIA-9> * fixed ROCK -> TOOLCHAIN cache name and added a comment 16:22 < mnemoc> tested? 16:22 < rxr> sure 16:22 < rxr> I'm just debugging the remaining issues ... 16:23 < madtux> gut gut 16:23 < rxr> madtux: und, wann commited du deine Modifikationen nach trunk ? 16:23 < rxr> *g* 16:23 < rxr> s/commited/commitest/ 16:25 < madtux> i will soon. 16:25 < madtux> got several updates to test i have been already playing around to move the $foo out of Documentation/ 16:26 < daja77> madtux german improves, so he can read the mails himself *ggg* 16:26 < madtux> daja77: there way too much to learn for mew to do that 16:26 < rxr> mails ? 16:29 < rxr> ouhm 16:29 < daja77> he'll get mails due his clt registration ... 16:30 < madtux> rxr: for the event at chemnitzen or however its written 16:30 < rxr> -en 16:30 < madtux> :) 16:30 < madtux> danke mein freund :) 16:31 < rxr> daja77: did you read my comment that your diet.tu-chemnitz.de is far superior to the degrading dict.leo.de ? 16:31 < daja77> oh yes, read that ;) 16:32 < daja77> i can tell the webmaster next time i meet him 16:33 < CIA-9> rene * r5080 /trunk/package/develop/ccache/parse-config: 16:33 < CIA-9> * fixed the other ccache statistic matching accordingly 16:33 < CIA-9> -This line, and those below, will be ignored-- 16:33 < CIA-9> M ccache/parse-config 16:33 < rxr> -> Compiler Cache Hits while pkg build: 97.83% (90 hits, 2 misses) 16:34 < mnemoc> LOL, you removed the first '-' of the message :p 16:34 < rxr> my last 5 dict request at leo did not yield any result - your uni one did provide a bunch for each ... 16:35 < rxr> damn 16:35 < rxr> yep 16:35 < rxr> sorry 16:35 < madtux> :P 16:35 < rxr> so - now let's track why C++ is not cached correctly .. 16:37 < rxr> -> Compiler Cache Hits while pkg build: 1.15% (1 hits, 86 misses) 16:48 < rxr> ok - so let's hope my modified cmd_wrapper works as expected - otherwise I'll have /fun/ debugging it ... 16:58 < rxr> ok - I have fun debugging 17:04 < rxr> have you seen this one: 17:04 < rxr> http://www.mozilla.org/images/nyt_ad_large_2004.png 17:07 < rxr> http://www.mozilla.org/press/nytimes-firefox-final.pdf 17:07 < rxr> ^- xpdf needs quite some time to render the pdf ... ;-) 17:09 < mnemoc> that should have been expensive 17:10 < rxr> as far as I read they collected some hundred thousand US$ 17:28 * mnemoc is now part of svk's changelog :) 17:29 < mnemoc> --verbatim option to skip adding of svk's headers 17:41 < rxr> congratulations! 17:41 [Users #t2] 17:41 [ _Ragnar_] [ daja77] [ jsaw ] [ mnemoc ] [ nzg] [ sparc-kly] 17:41 [ CIA-9 ] [ gfdsa ] [ madtux] [ nullslack] [ rxr] [ valentin ] 17:41 -!- Irssi: #t2: Total of 12 nicks [0 ops, 0 halfops, 0 voices, 12 normal] 17:41 < mnemoc> :p 17:43 < rxr> (we need to rewrite the wrapper some day) 17:44 < mnemoc> yes 17:44 < mnemoc> we may also use wrapper to solve full-buildability 17:44 < mnemoc> we may also use wrapper to solve full-crossbuildability 17:44 < rxr> as long as we get rid of this hacky piece of code 17:45 < rxr> you do not want how long I read and debug on it right now .. 17:45 < rxr> :q! 17:46 < rxr> sorry 17:46 < mnemoc> have you designed the rework of it? 17:46 < rxr> no - I just want the hack to work right now ... 17:46 < rxr> I need to do other stuff .. 17:47 < mnemoc> obviously 17:47 < gfdsa> rxr: you do need to rewrite it... (my 5 cents) 17:48 < rxr> yes - I know 17:48 < mnemoc> :D 17:48 < rxr> just that I do not want to do it right now .. 17:48 < rxr> I have some commercial thing to hand out monday - and i shoudl work on that ... 17:48 < gfdsa> if i wouldnt need the results right now, i would help you... 17:48 < rxr> I just want this damn thing to do what it should do ... 17:48 < gfdsa> the ideas inside t2 are very good 17:49 < gfdsa> someone had a bad design in the beginning... IMO 17:49 < rxr> *g* 17:50 < CIA-9> mnemoc * r5081 /trunk/: 17:50 < CIA-9> mnemoc * r5082 /trunk/: 17:50 * mnemoc wonders what svk is doing 17:50 < CIA-9> mnemoc * r5083 /trunk/ (. package/develop/ccache/parse-config): 17:50 < CIA-9> mnemoc * r5084 /trunk/: 17:50 < rxr> mnemoc: ! 17:50 * mnemoc checking diffs 17:51 < rxr> may I hit you ? 17:51 * rxr slaps mnemoc 17:51 < mnemoc> not need, i'm doing it myself 17:51 < mnemoc> svk has it's own life 17:51 < rxr> well - it needs 50+ perl packages - it most probably can speak soon 17:52 < rxr> .oO(my name is perl script) 17:52 < mnemoc> :) 17:52 < rxr> (cpan module 7 of 54) 17:53 < mnemoc> now svn diff -r5081 http://svn.exactcode.de/t2 doesn't work :\ 17:53 < mnemoc> .oO( what the heck is happening in my box!!! )o 17:53 < madtux> you know mnemoc this svk thing is not really convincing me.. 17:54 < rxr> mnemoc: I have to clean up trunk now ? 17:54 < madtux> rxr: no, he has deleted it 17:54 < mnemoc> i'm over it 17:54 < rxr> hasn't svk some dry run mode ? 17:55 < mnemoc> 5081,5082 are NOP 17:55 < mnemoc> yes, -C 17:55 < rxr> except svk properties ? 17:55 < mnemoc> but it doesn't do exactly the same with and without -C 17:55 < mnemoc> merge tickets :\ 17:55 < madtux> damn I'm SICK of problem at home. 17:55 < madtux> :( 17:56 < mnemoc> real problems are _always_ at home 17:56 < madtux> mnemoc: yes but i'm tired mnemoc .. 17:56 < madtux> i'm really tired 17:56 < mnemoc> :\ 17:56 < madtux> i want to live my life and deal with _MY_ problems 17:57 < madtux> mm.. i'm talking what i think i better shut. 17:57 < mnemoc> 5083 es NOP too (according to svn diff) 17:58 < mnemoc> what? and 5084 es simply empty? 17:58 < rxr> mnemoc: ! 17:59 < madtux> mnemoc: please don't missunderstand me but u are starting to scare me 17:59 < mnemoc> rxr: :( 17:59 < madtux> mnemoc: wtf are u doing? 18:00 < rxr> madtux: mnemoc is shiffering enough already I guess we do not need to further support this ... 18:00 < madtux> very well. 18:00 * mnemoc shamed 18:00 < rxr> ah - shit happens ... 18:01 < mnemoc> thanks :) 18:01 < rxr> s/ah/nah/ 18:11 < rxr> damn - enlightenment just core dumped .. 18:18 < mnemoc> http://svk.elixus.org/index.cgi?SVKSucks <--- the TODO list of svk 18:19 * mnemoc gone to lunch 18:22 -!- gfdsa [~kvirc@213.156.44.71] has quit ["cya later"] 18:29 -!- madtux [~mike@200.91.101.101] has quit ["leaving"] 19:47 < CIA-9> rene * r5085 /trunk/misc/tools-source/cmd_wrapper.c: 19:47 < CIA-9> * resorted some parts of the cmd_wrapper - to parse the options before 19:47 < CIA-9> the other wrapper is executed - so ccache or distcc get the correct 19:47 < CIA-9> arguments - contains some debugging still - to be removed and 19:47 < CIA-9> revisited 20:54 * mnemoc *hates* Worm.Zafi.D! 20:58 < rxr> what worm 20:58 < rxr> ? 20:58 < mnemoc> Zafi.D 20:59 < mnemoc> i has my smtp server on 95% :\ 20:59 < mnemoc> it* 21:00 < mnemoc> 4 mails per seconds with that damn worm 21:01 < mnemoc> OT: has anyone played with GOsa? https://gosa.gonicus.de/ 21:09 < rxr> no - I not ... 21:38 < mnemoc> rxr: i'll (try to) commit a tested proposal for ABORT_ON_ERROR stage based 21:38 < rxr> ouhm 21:38 < rxr> or better post the patch first .. 21:39 < rxr> but if you prefer a commit ... 21:39 < mnemoc> ok 21:39 < mnemoc> i'll post 21:39 < mnemoc> but svk will not win me :) 21:44 < rxr> not win? 21:45 < rxr> you mean iit is not good enogh? 21:45 < mnemoc> it is great 21:46 < mnemoc> but it acts 'odd' because i have to mirrors with the same UUID because rock-trunk and t2-trunk share their past 21:46 < mnemoc> rxr: mail sent, comments? 21:49 < mnemoc> rxr: on the choice i should have s/on/after/g 21:54 < mnemoc> fixed* 21:55 < mnemoc> uhm... and s/Abort/Continue/g 21:55 < mnemoc> i'm too stupid this days 22:04 < mnemoc> comment replyed 22:33 < rxr> -> Compiler Cache Hits while pkg build: 100.00% (361 hits, 0 misses) 22:33 < rxr> I had evening meal .. 22:34 < rxr> I boot into osx to continue my com. work and repluy 22:34 < rxr> cu 22:35 < CIA-9> rene * r5086 /trunk/package/x11/xscreensaver/xscreensaver.desc: * updated xscreensaver (4.18 -> 4.19) 22:57 < rxr> re 22:58 < rxr> had some more mail stuff in the queue ... 23:08 < rxr> mnemoc: I find relation easier to read if the variable is written first ... 23:08 < rxr> $stagelevel -le $ROCK_CFG_FIXED_ERROR_STAGE_CONSTANT ... 23:08 < rxr> but maybe it is just me .. 23:08 < rxr> I might also not check for emptyness explicitly 23:09 < rxr> either it is defined or the build will not be aborted ... 23:09 < rxr> I crawl into bed ... 23:09 < rxr> cu early in the morning .. 23:09 < mnemoc> i did that -n for compatibility 23:09 < mnemoc> but i can remove it 23:10 < mnemoc> i'll write the comparisons in that order you proposed --- Log closed Fri Dec 17 00:00:28 2004