IRC Logs for #crux-devel Thursday, 2014-05-22

*** frinnst_ has joined #crux-devel02:57
*** frinnst has quit IRC03:03
*** mike_k has joined #crux-devel05:34
*** mike_k has quit IRC05:36
teK_starting the final attempt for building clang, the updated way10:05
Romsterman the documentation on there site is really bad...10:14
Romsteralso not sure what todo with mesa3d in crux 3.0 do i need to patch it or will you revert llvm for crux 3.0 only teK_ ?10:14
teK_I have to read up on that matter10:16
teK_which version would you need?10:16
Romsteryeah 3.3 on that older mesa3d in crux 3.0. i'm not sure if i can patch it or we should... or just disable the expermental ati stuff off mesa3d so it don't use llvm perhaps. and let everyone migrate to 3.1 where r600 is not experiential anymore10:17
Romsterpatch mesa3d that is.10:18
teK_I'd make that decision depend on the time-line for our 3.1 release10:18
Romsterguess we need a group decision on what todo here. frinnst jue ^10:18
teK_jaeger -^10:19
Romsteras it stands mesa3d will not compile.10:19
teK_and that's no option ;)10:19
Romsterooops forgot him.10:19
Romsteris just us anymore in crux that are actually /active/10:19
teK_prologic and alan? :)10:20
teK_mike_k, dxm (Maximilian Dietrich)10:20
Romsterwell alan just does a few ports in opt and kde stuff.10:21
Romsterprologic just does some of contrib.10:21
Romsteri mean the main ones that make crux decisions. that have xorg/compat-32/opt access10:22
Romstereven those that have core.10:22
Romsterrest of contrib and other repos don't really provide much input to the direction of crux?10:23
frinnstI'd vote for not making any major changes to 3.0 and just let "everything work" as is10:23
frinnstdont bother with mesa3d/llvm changes in 3.010:23
Romsteri would then suggest downgrading llvm for 3.0 then.10:26
Romsteras it stands mesa3d has been broken for since llvm got bumped.10:26
Romsterand that's quite some time now.10:26
Romsterhonestly think who ever biumps ports should at lest try rebuilding depednent ports in the most likely used ports that depend on said port bump.10:28
Romsteri try to do this with any major changes.10:28
Romsterof a port that affects others.10:28
Romsterguys what happend to netstat in 3.1 again? is there a command to supersede it?10:49
Romsternet-tools is removed in 3.110:50
*** horrorStruck has quit IRC10:52
Romsterguess ss takes over.10:55
frinnstaye, ss11:09
teK_I vote for downgrading, too so I will do just that.11:11
prologicwho's talking about me :)12:58
*** rexich has joined #crux-devel13:24
jueI'm with frinnst, no bigger changes for 3.0, so let's downgrade llvm13:48
jaegerI have no objections14:01
*** rexich has quit IRC14:08
*** rexich has joined #crux-devel14:08
*** rexich has quit IRC14:08
*** rexich has joined #crux-devel14:15
teK_oh boy15:22
teK_one of the last xorg updates (fontsproto, gccmakedep, imake or libxfont) seriously breaks xcompmgr15:22
teK_so I went back to the old fontsproto-2.1.3 and libxfont-1.4.7 that worked before and shit still is broken15:35
teK_I cannot image gccmakede or imake to be the cause but who knows15:36
teK_ok reverting these still doesnt help15:41
teK_fuck that.15:41
jaegermaybe unrelated somehow?15:43
teK_nevermind. You know what fixed it? reinstalling feh. I'm using fluxbox:15:43
teK_OF COURSE15:43
teK_and the best thing:15:43
teK_[tek@basra][~]% grep -R feh .fluxbox || true15:43
teK_I checked that before15:43
frinnstdoes someone understand wtf is up with FS#1019 ?18:18
frinnstdoesnt look like a regular make race..18:19
teK_ok but how is moving that variable supposed to help building e2fsck?18:32
teK_I can reproduce that with -j4, too so just add -j1?18:37
frinnstreally? i struggle18:40
frinnstmy thought was that his profile was a bit fucked somehow18:40
frinnstyou get the prof_err.h: No such file or directory error too?18:40
jaegerI use -j5 on my ISO builders and build e2fsprogs quite frequently there. no issues18:46
frinnstyeah, same here.. anyways i've pushed make -j118:51
teK_frinnst: yes I do19:10
teK_jaeger: how many cores do you have?19:11
teK_I have a Quadcode i719:11
frinnstsame as tek 4 + ht19:11
teK_I do not seem to have HT19:11
teK_top shows 4 CPUs19:11
teK_ /cores19:11
teK_I hope that these are physical cores19:12
frinnsttop shows 8 for me19:12
frinnstdisabled ht in bios?19:12
teK_cpu cores: 219:12
teK_according to cpuinfo19:12
teK_this is a mobile i719:12
frinnstdont you have a 2600k?19:12
jaegerteK_: 4 on that VM19:12
frinnstsilly tek19:12
teK_yes but in my desktop, this is the laptop19:12
teK_frinnst: oh shut up19:13
teK_oh NOOO19:13
frinnstbring me a shrubbery!19:13
teK_HE SAD NI19:13
frinnstguess what i just watched19:13
frinnstlife of brian + holy grail after work19:15
frinnstepic evening19:15
teK_that's true19:15
teK_oh boy19:16
teK_=======> Building '/usr/ports/packages/e2fsprogs#1.42.10-1.pkg.tar.gz' succeeded.19:17
teK_so it's some race :-)19:17
frinnstdid you get the same missing header error as the bug reporter did?19:17
frinnstthat's so strange imo19:17
teK_I did during the first run19:21
jaegerI can'19:31
jaegercan't explain why but the caja issue I had with MATE is gone now19:31
jaegerEither they fixed it in MATE 1.8 or I built something differently this time19:31
*** sammi` has joined #crux-devel19:51
*** rexich has quit IRC20:21
frinnstso gcc 4.8.321:14
frinnstshould be safe to push to 3.1, right?21:14
jaegerprobably... we should test it but .3 versions are usually sane21:16
frinnstisnt pushing the best way to test? :)21:17
jaegerconsidering we haven't released 3.1 yet I have no objections. that can go in an rc2 as far as I'm concerned21:18
*** rexich has joined #crux-devel23:13

Generated by 2.11.0 by Marius Gedminas - find it at!