IRC Logs for #crux-devel Tuesday, 2010-08-31

*** sepen has joined #crux-devel02:03
*** jue has joined #crux-devel02:27
sepenI'd like to close #586, any opinions about?02:43
prologicwhat's the current (default) bootloader anyway ?02:46
sepencore/lilo vs opt/grub02:46
prologicahh ic02:47
prologicI've never used syslinux that's all02:47
prologicwasn't aware it was a lilo/grub replacement02:47
prologic(partially aware)02:47
prologicI have no issues with it being in syslinux in that case though - given the (now) single dep02:47
sepensyslinux now is able to boot btrfs partitions02:48
prologicand ext402:48
sepenyep, and I think that lilo and grub can't02:48
prologicbut ihmo that shouldn't be the reason to move it to opt :)02:49
prologicproviding an alternate boot loader02:49
prologicthe reason should be that others would like to use syslinux in place of lilo/grub02:49
prologicI myself haven't yet02:49
prologicbeen happy with lilo for 10+ years now02:49
sepenI use syslinux on usbdisk installations02:49
sepenand you?02:49
prologicI don't have any usb installations02:50
prologiclike I said, no issues with an alternate bootloader to lilo/grub being in opt (syslinux)02:50
sepenhmm If I remember correctly, Moritz, the guy who opened the ticket told me that he was using syslinux to boot his desktop02:51
prologicahh k02:52
sepenprologic, I'm also using lilo for 10+ years02:52
sepenonly boxes at office have grub installed, with password02:53
juesepen: it's your decision, you can move everything you like to opt. Currently opt is not a selection of "selection" of important ports but the repo of the opt maintainers02:55
jues/selection of//02:56
prologicI tend to agree with jue - even though I'm no opt (or even contrib - I was) maintainer :)02:56
prologicbut ... I do firmly believe opt should contain a small limited set of ports02:57
prologicso try to keep it concise guys :)02:57
juewell, that's how Per defined it ;)02:58
sepenthanks for the tip02:58
sepenjust I like to share opinions and discuss a bit too02:59
prologicofc :)03:00
juesure, that's fine03:00
sepenon the other hand I have always considered opt a good place to host most important libs03:02
teK_there's no need to keep opt small as long as its port's quality is good03:03
prologicI agree :)03:04
prologicbut just remember that one of the appealing things about CRUX and the ISO is the small footprint :)03:05
prologicsmall core/opt03:05
prologicbut ofc quality is important03:05
jueprologic: remember, not ever opt port is part of the ISO03:05
prologicactually you're quite right :)03:05
sepeni.e. opt/amule ;D03:06
juesepen: yep, we should cleanup our orphaned ports03:11
sepenor at least the last merge 2.6->2.7 should remove them before 2.7 will be release03:14
teK_there are quite some offers available (from nipul(03:27
juejust adopted libexif because libgphoto2 depends on it03:28
jueteK_: shouldn't we replace qemu/kqemu by qemu-kvm?03:36
teK_+ THEY should merge all efforts into one project.03:38
jueyep, that's right, but ...03:39
teK_of course03:40
prologicihmo amule really ought to be in contrib :)03:42
jueI agree with sepen, best would be to remove all orphaned opt ports in our 2.7 branch03:45
juebut at least we must find new maintainers for cdrkit and grub ;)03:46
teK_I use cdrkit so I could take care03:52
teK_new episode Dr. House \o/03:53
tilmanomg, srsly?04:01
tilmanpeople still use *mule?04:02
sepenI'm still using mldonkey ;D04:03
tilmanwho is our resident  python guru?04:05
tilmannipul mantained a bunch of python stuff04:05
prologic<tilman> who is our resident  python guru? <-- that would be me :)04:25
juetilman: btw, is there a specific reason why we use cryptsetup 1.0.7 and not 1.1.3 (current stable)?04:29
tilmanno. i'll update04:29
juethanks, works flawless on my netbook04:30
*** tnut has joined #crux-devel05:03
*** prologic has quit IRC05:26
*** sepen has quit IRC06:47
*** tnut has left #crux-devel07:09
*** kreed has joined #crux-devel07:50
*** tnut has joined #crux-devel08:12
*** jaeger has quit IRC08:16
*** jaeger has joined #crux-devel08:23
*** aubic has quit IRC11:14
*** aubic has joined #crux-devel11:27
*** jue has quit IRC12:20
*** aubic has quit IRC12:20
*** jue has joined #crux-devel12:32
*** kreed has quit IRC12:39
frinnstdoes anybody else not have sourceforge specified in /etc/hosts ?14:33
frinnstim getting tired of replacing with (always seems to work btw)14:33
*** prologic has joined #crux-devel14:39
juewhenever I have to touch a SF port, I change the url to the new one which is usually something like ->
jaegernice, I'll have to try that14:47
tilmanfrinnst: download or downloads?14:48
tilmanafaik will resolve to the same ips as dl.sf.net14:49
jaegerit does, yes14:53 is just a CNAME for downloads.sourceforge.net14:57
juefrinnst: I don't have it in /etc/hosts currently15:00
jaegerinteresting... if it's a cname they must be doing some kind of local named-base vhost config15:02
jueare there any issues left with 2.7?15:20
tilmanx is behaving a little weird on i945 ;)15:20
tilmanthat a release blocker though15:21
juehmm, with server 1.9.0?15:23
*** aubic has joined #crux-devel15:25
tilmani'm currently bisecting the server15:26
tilmanmaybe it can be fixed soonish15:27
juemy netbook works nice with 1.8.1, I can do the update to 1.9.0 tomorrow, anything I can do/test?15:28
tilmanproblems i'm seeing:15:28
tilman* my urxvt doesn't draw its cursor properly wrt a focused/unfocused window state15:28
tilman* switching between virtual desktops causes the last focused window not to be drawn (might be intel-specific)15:29
teK_ <- could be of interest15:30
juewith openbox?15:30
teK_(no connection to your xorg-problem)15:30
jueteK_: you've tk installed?15:31
teK_I'm aluding to the missing files15:31
teK_the new comes from tk, I guess, yes15:32
tilmanpkginfo -i|grep tk ? :p15:32
teK_why are there files _missing_15:33
pitilloteK_: 64b?15:37
juedidn't you get that with 2.6.5?15:39
teK_dunno. shall I test it?15:39
tilmanjue: btw, in the 2.7 branch, the Description for opt/python says "... version 2.6" but the port is version 2.715:40
juethanks, will fix it :)15:40
jueteK_: yes15:41
prologictilman, btw ... feel free to throw anything python related my way :)16:32
tilmanbtw, i tracked down the x bug16:33
tilmanunfortunately it seems it's not easily fixed ;D16:33
prologicthe x bug ?16:35
prologicI think I saw you mentionining something about it - but not quite up to speed with it :)16:35
teK_jue: still missing and in python 2.6.516:36
prologiceh ?16:36
prologicyou shoudln't be missing those16:37
teK_yet I am16:37
prologic$ pkginfo -l python | grep dl\.so16:38
teK_you know.. that's the issue we're talking about16:38
teK_22:30 < teK_> <- could be of interest16:38
teK_what does do?16:38
prologicdynamlic loading of C extension modules I believe16:38
prologicif I'm not mistaken16:38
tilmani'd inspect the makefiles/stuff and see what the dependencies for are16:39
prologica lot of non-pure python stuff just won't work without it I'm fairly sure16:39
teK_so '' will not be built, when:16:39
teK_- sys.maxint isn't 0x7fffffff (perhaps on some 64-Bit archs)16:39
teK_- dlfcn.h can't be found16:39
prologicstrange that I'm hearing of this now though16:39
prologicoh wait16:40
prologicare you on 64bit crux ?16:40
tilmanteK_: huh16:40
teK_prologic: I am16:40
tilmanteK_: run python, then punch in: import sys (newline), sys.maxint16:40
tilmanwhat does it say?16:40
teK_*punching the snake*16:41
tilmanoh well, no idea whether that's expected16:42
prologic>>> sys.maxint16:42
prologicobviously 32bit here16:42
prologic>>> 2**64 / 216:42
prologicit's expected16:42
prologicsys.maxint (afaik) is 2**x / 2 where x is the bit width of your arch16:42
prologicsame thing :)16:43
tilmanjust because it matches 2**63 -1 doesn't mean it's expected to match sys.maxint ;)16:43
prologic2**63 or 2**64 / 216:43
tilmanso it's essentially the max signed _long_ type of the system16:43
teK_tilman: yeah; but it really looks like 2's complement for 64 Bit ints16:43
tilmannot _int_16:43
prologicthat's true16:43
tilmanfucked confused inverted fucked up python shit16:43
prologicbut on a 64bit machine that's what sys.maxint should be16:43
prologichowever I failt to see why that has anything to do with the missing not being compiled16:44
tilmanteK_: so, issue solved. what about the deps for imageop?16:44
teK_grep-ing for it16:47
prologicThe above instructions worked well. However, I still had a couple of modules that could not be built - namely "dl" and "imageop". This was cleared up by the README in that "imageop" soes not work on 64 bit platforms and "dl" dynamic loading does not work via dlopen(). disabled both from building and finished with the make install. I also tried the launchpad instructions but the platform I was building on was hardy and not intrepid.16:47
prologicignore it16:48
prologic(at least on 64bit machines)16:48
teK_maybe one could create a 64bit overlay port16:48
tilmanthat seems to be the right thing to do16:48
teK_yes sir16:48
prologicsomeone remind me to buy and build a new desktop - a pure 64bit one :)16:49
prologicmight still keep this old one lying around though just for the hell of it :)16:49
teK_ah! I wanted to test nmap+openssl for 2.7.16:49
prologicI prefer virtualbox :)16:49
prologicvirtualbox rocks :)16:50
prologichave it on my mbp too16:50
prologichey my mbp is 64bit actually16:50
prologicI just realized d'uh16:50
prologicQ: Are we shipping out officially supproted 64bit CRUX ISOs for 2.7+ ?16:50
teK_don't think so :\16:51
frinnstonly my broken iso's :)16:52
teK_what's the matter? time? man power?16:52
frinnstthey should hopefully work without issues by the time 2.7 gets rolled out16:52
teK_what problems do you have?16:53
prologicyes what's the problem dude ?16:53
prologiceven I have 2 64bit machines here - 1 of which isn't utilized as one16:53
frinnstnone really, just need to build and test more16:54
frinnstusually too tired to do much when i get home from work16:54
teK_I will checkout i686 rc2, first16:54
prologicfrinnst, wheres's the iso atm ? got a link on somewhere ?16:56
prologicI'll do some testing on my mbp under virtualbox for now16:56
* teK_ just discovered pkgmk_ignore_new.. yay!17:00
prologicfrinnst, mind putting a link to this on the crux wiki ?17:04
frinnstoff to bed, but feel free17:07
prologictilman, can you see if I still have an account on the crux wiki ?17:12
prologicI was certain I had one17:12
prologicor can someone check / :)17:14
teK_trying right now17:15
teK_I forgot where it stores its users17:15
prologicnot sure if I had an account called prologic or JamesMills17:15
teK_JamesMills17:20 haha nice pun17:28
prologicany idea what my password is ?17:30
prologicor can you reset it / send it to my email ? :)17:30
teK_karma++ *g*17:35
teK_jue: I'm not quite sure which qemu revision + fork to take as qemu-kvm 0.12.5 (current) won't boot 2.6.35 kernels as I recently found ot17:59
teK_I will give this some more thought18:26
*** mavrick61 has quit IRC21:35
*** mavrick61 has joined #crux-devel21:36
*** prologic has quit IRC22:08
*** prologic has joined #crux-devel22:19
*** prologic has quit IRC22:55

Generated by 2.11.0 by Marius Gedminas - find it at!