T2 IRC Log: 2010-05-30

This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.

« prev | next »

--- Log opened Sun May 30 00:00:03 2010
10:34 -!- zod [~zod@smitty.knid.net] has joined #t2
10:34 < zod> hi there
10:41 < rxr> hey there
10:42 < rxr> cool thing, Lena won the Eurpean Song Contest, just the 2nd time for Germany ever: http://rene.rebe.de/2010-05-30/we-have-won-the-eurovision-song-contest-in-oslo/
10:42 < zod> i'm trying to use ./scripts/Config but after building luabash it fails with "bash: symbol lookup error: src/luabash.so: undefined symbol: find_variable"
10:43 < rxr> hm, though that part is certanyl the most esoteric in the T2 build / config system, you are the first to have a problem with it, ever, what kind of OS are you running on?
10:43 < zod> Debian testing
10:44 < rxr> hm, strangely interesting
11:55 < zod> ok, somehow my bash-package was broken. config works now but trying to build a minimal target (bootstrap) fails because gcc can't be built...
12:44 -!- mpp [~mpp@2001:470:9a41:107:c1ce:ce52:5b18:7d4d] has joined #t2
13:03 < mpp> moinsen
14:55 -!- mpp [~mpp@2001:470:9a41:107:c1ce:ce52:5b18:7d4d] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]]
17:36 < rxr> moin moin mpp
18:06 -!- mpp [~mpp@2001:470:9a41:107:c1ce:ce52:5b18:7d4d] has joined #t2
18:06 < mpp> re
18:07 < rxr> hi mpp
18:07 < mpp> hey finally
18:07 < rxr> depends :-)
18:07 < mpp> so i did some long research on my build problem.
18:08 < mpp> long story short: the x86_64 build host has some troubles with gcc prior to 4.5
18:08 < mpp> i have used the standard iso images provided by t2
18:08 < mpp> and completely fresh vms
18:09 < mpp> in the end i used a 32bit build
18:09 < rxr> to reproduce this?
18:09 < mpp> yes
18:09 < rxr> strange, that both build here
18:09 < mpp> i know
18:09 < rxr> i tag some new rc soon and try again
18:09 < mpp> BUT - you most certainly do not use bleeding edge builds as a host
18:10 < mpp> it is not really a problem
18:10 < mpp> but if you are not experienced in t2 - u spend alot of time
18:10 < mpp> so it's really not necessary
18:11 < mpp> a regular minimal build iso would greatly improve spreading t2
18:11 < mpp> imho
18:11 < mpp> that is
18:11 < mpp> anyways
18:11 < mpp> i have a 32 bit iso building today based on a gcc 4.5 host
18:11 < mpp> all vanilla svn
18:11 < rxr> yes, I will
18:11 < mpp> i would like to upload this to unofficial
18:11 < rxr> sure
18:12 < mpp> just in case - and for the sake of reproducing scenarios like these
18:12 < mpp> i also patched mysql
18:12 < mpp> to work
18:12 < mpp> again
18:12 < mpp> :-)
18:12 < rxr> what pach? mysql works here :-(
18:12 < mpp> the patch involved the configure - and not the configure.in as it should be
18:12 < mpp> yes ,yes on ur upgraded host and toolchains
18:13 < mpp> surely no prob as mentioned above.
18:13 < rxr> what configure detail?
18:13 < mpp> anyways i'll put an to this
18:13 < mpp> http://svn.exactcode.de/t2/trunk/package/database/mysql/gcc45.patch
18:14 < mpp> hold on i'll check my repo to see the diff
18:14 < mpp> it shoudl be Makefile.in and NOT Makefile.am
18:14 < mpp> in the patch
18:14 < rxr> ah :-)
18:15 < rxr> yeah, in theory .am might be the best for upstream, but indeed I often just patch the final result to avoid the regenreation hickups
18:15 < mpp> in this case it fails on a fresh build host
18:15 < mpp> so no point in patching the .am
18:15 < rxr> you mean the regen from the .am file?
18:15 < mpp> yes
18:16 < rxr> well, appears to build on my tests, but I know this autojunk is nightmare, which is why I most often patch the final file
18:16 < rxr> guess aldas did that patch
18:16 < rxr> http://rene.rebe.de/2010-02-16/darn-how-i-hate-this-gnuauto-junk/
18:16 < mpp> http://pastebin.com/FkrqBGY1
18:16 < mpp> indeed
18:17 < mpp> but it's okay as long as it is that trivial - not to encourage though
18:17 < rxr> you just removed the whole abi check thing?
18:17 < mpp> yes sure
18:17 < mpp> i did research on that topic
18:17 < rxr> I've never seen that before
18:18 < mpp> for now it does what it's supposed
18:18 < rxr> but certainly only causes troubel for us
18:18 < mpp> to
18:18 < mpp> removing it
18:18 < rxr> either the compiler works or it doens't - certainy not for mysql configure to decide
18:18 < mpp> ?
18:18 < mpp> you are right
18:18 < mpp> but that only occurs on gcc 4.5
18:19 < mpp> eaccelerator needs to be upgraded to 0.9.6
18:19 < mpp> it fails on gcc 4.5
18:19 < mpp> i double and triple chekced that
18:21 < rxr> damn I think your paste is tab mangled?
18:22 < mpp> ahm..
18:22 < mpp> sorry
18:22 < mpp> it's just my git diff
18:25 < mpp> have a nice sunday - bbtmrrw
18:25 -!- mpp [~mpp@2001:470:9a41:107:c1ce:ce52:5b18:7d4d] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]]
18:25 < rxr> u2
18:26 < CIA-30> rene * r36485 /trunk/package/www/eaccelerator/eaccelerator.desc:
18:26 < CIA-30> Martin Papadopoulos :
18:26 < CIA-30> * updated eaccelerator (0.9.5.2 -> 0.9.6)
--- Log closed Mon May 31 00:00:03 2010