IRC Logs for #crux Monday, 2013-04-15

*** Henesy has quit IRC00:02
*** Henesy has joined #crux00:17
*** BurnZeZ has joined #crux01:13
HenesyNot much01:16
Henesybeen pretty quiet01:16
BurnZeZprologic: I could be wrong, but I think the curlftpfs build source link is broken.01:18
prologicwhy are you telling me? :)01:19
BurnZeZIt's in your repo01:19
BurnZeZAnything sourceforge is black magic, to me.01:21
Henesyworse than git? :P01:21
BurnZeZOf course01:21
BurnZeZgit is pretty easy01:21
BurnZeZAlthough it's bloated01:22
BurnZeZSo very bloated01:22
*** Henesy has quit IRC01:41
*** BurnZeZ has quit IRC01:45
*** Henesy has joined #crux01:54
*** BurnZeZ has joined #crux02:06
*** Henesy has quit IRC02:17
*** Henesy has joined #crux02:31
Romstergit is not bloated02:37
Romsterit's versatile02:37
BurnZeZAKA bloated02:37
BurnZeZUNIX philosophy is best philosophy02:37
BurnZeZgit tries to do too many things02:38
Romstergit one tool one job CMS02:38
RomsterCMS's by nature are complex.02:38
Romsterhave to remember it handles the kernel as well.02:39
BurnZeZhg does a much better job at not being bloated02:39
Romsterand rebases and git pull requests and all that.02:39
BurnZeZIt could do better, though02:39
Romstergo use hg then i like my git02:39
BurnZeZI would use hg if I was as proficient with it as I am git02:40
Romsterand you'll never be if you never use it <_<02:40
BurnZeZThat's exactly right02:40
Romstercatch 2202:40
BurnZeZMaybe I'll start using it for some side projects02:41
*** mavrick61 has quit IRC02:41
Romsterprologi_c,  uses hg ask him if you get stuck.02:42
*** mavrick61 has joined #crux02:42
GoodboxHi guys, someone here are using distcc and dmucs ?02:51
BurnZeZFirst time hearing of dmucs02:53
BurnZeZLooks neat02:53
GoodboxThat's seem to be a great tool, given i've got very differents machines02:56
*** diverse has joined #crux03:04
diverseDoes lilo support GPT?03:04
HenesyEvery time I remember macbooks use both UEFI and GPT simultaneously a small part of me cringes to death03:10
diversemacs don't use UEFI, they use their own proprietary EFI that is not standard. Anyway whats wrong with GPT?03:12
HenesyGPT is fine in its own right, but I have had trouble with bootloaders crapping out on me for seemingly no reason because GPT is in use03:13
diverseGrub2 and syslinux worked fine for me with booting into GPT from a BIOS system.03:16
diverseits a shame that lilo doesn't support that.03:17
HenesyI prefer Legacy GRUB to Grub2 personally03:19
Henesyand LILO over both03:19
diverseI want to use lilo, but it doesn't support my need for GPT.03:20
*** v33 has joined #crux03:21
v33hello hello03:22
BurnZeZIt's the worst thing ever when I find a package for something I want to install, only to realize it defies the CRUX packaging standards.03:23
Henesyv33, hallo03:24
v33how are ya henesy03:25
diversev33: you seem happy today.03:25
Henesyv33, Doing good thank you :) Only need to get my pi booting the 3.8.6 kernel on Crux-ARM and my troubles will be complete03:26
v33well, jut got back the gym...hit new records so im not complaining :P03:26
v33that and i have high hopes i can get rome total war running in wine if god be willing03:26
jaegerI've used grub2 with uefi and gpt, bit of work to get it set up but it works well03:28
diversejaeger: I am thinking of going with grub2 now, I am tired of syslinux.03:32
BurnZeZHow horrifying03:32
jaegera boot loader is a boot loader, I don't get particularly emotionally invested in them :P03:33
v33is it possible to mount a bin file to linux? its essentially an iso...03:33
jaegerIf it's in ISO format, sure03:33
jaegerIf it's something else, maybe03:34
v33its a .bin file, but its essentially mounts like an iso would03:34
jaegerIf you already know it mounts, what is the question?03:35
v33would the regular mount file.bin /media work?03:36
jaegertry it and see03:36
v33i recall trying some time ago, but it wouldn't...i dont have the file completely downlaoded to try it03:36
jaegerIf it doesn't, try specifying -t iso966003:36
v33so mount file.bin -t iso9660 /media seems simple enough03:37
v33just finished downloading. lets hope it works :P03:38
*** diverse has quit IRC03:41
BurnZeZI'm not sure I quite understand how locales are handled.03:59
BurnZeZI understand you generate them yourself, but I'm not seeing the next logical step in the chain of commands.04:00
*** deus_ex has quit IRC04:07
*** Henesy has quit IRC04:22
*** pitillo has quit IRC05:06
*** pitillo has joined #crux05:08
*** BurnZeZ has quit IRC06:00
*** BurnZeZ has joined #crux06:04
BurnZeZSo how does prt-get handle conflicting package names?06:12
*** v33 has quit IRC06:30
*** diverse has joined #crux06:48
niklasweis it someone some using kde4 on crux?07:04
BurnZeZNah, I just have a couple ports installed with a same name package in both07:05
*** BurnZeZ has quit IRC07:31
*** ody has joined #crux07:39
*** Romster has quit IRC09:57
*** Romster has joined #crux09:58
*** Romster has quit IRC09:58
*** Romster has joined #crux09:58
*** ody has quit IRC10:35
*** Henesy has joined #crux11:18
*** guzzano has quit IRC11:20
*** Henesy has quit IRC11:21
*** horrorSt1uck has joined #crux11:27
*** horrorStruck has quit IRC11:30
RomsterGoodbox, distcc backend on ccache11:34
Romsternever heard of dmucs either11:35
Romsterv33 bin2iso11:38
niklaswebuild kde4 yay :P11:41
Romsterstill ironing bugs out and i got another box to setup yet too.11:42
niklasweno, Im building kde4 on my workstation..11:43
niklaswetaking aaaaloooong time..11:43
Romsteryep i don't doubt that.11:43
*** sepen has joined #crux11:46
RomsterGoodbox, there is also icecream i might give that a try too.11:47
Romsterneed more powar time to setup the other box.11:51
*** dkoby has joined #crux12:08
*** himynameisphil has quit IRC12:10
*** dkoby has quit IRC12:28
horrorSt1ucksepen: hi, if possible, it would be great to remove some bash completion overlapping with CRUX tools (pkgadd, pkgrm), thanks12:36
horrorSt1uckthose are solaris tools and trigger some weird behaviour on a CRUX box12:36
sepensure good catch!12:43
*** Kaishi has joined #crux12:45
Romsterthat still makes no sense to be other than removing all completion of those two commands.12:51
sepen see this list, I'll remote them excepts pkg-config and pkg-get12:53
sepen*err, pkg-get also comes from solaris tools12:59
horrorStrucksepen: thanks :)12:59
sepenthanks for your report13:04
cruxbot[opt.git/3.0]: bash-completion: removed package tools from other systems rather than CRUX (thanks to horrorStruck)13:06
niklaswenice :D13:06
sepenI never used a gui for git13:07
sepenwell tried gitk but not for the daily work13:07
niklaswesepen: we using github on my work..13:10
*** dkoby has joined #crux13:15
Romsterinteresting but not needed to manage git/hg13:46
*** deus_ex has joined #crux13:51
Romsterdistcc over 4 machines now.13:52
*** Kaishi has quit IRC14:00
GoodboxHi romster14:11
GoodboxIcecream ?14:11
GoodboxOh a distcc fork14:12
GoodboxI got some cross-compilation problems with distcc14:13
Romsterthough i got control of all my machines. the major issue i see is not being able to dynamicly set number of build jobs.14:13
Romstertoo few it's under used. too many it's waiting and timing out.14:13
Romsteroh you have to be really really careful how you setup for cross-compilation14:13
GoodboxDid you tried dmucs ?14:13
Romsternope, been busy setting up another computer.14:14
Romstergot 3 dual cores for moosefs testing. and for distcc use.14:14
GoodboxYes cross-compilation might be a trade14:14
Romsterany anything else that comes to mind.14:14
Romsterthere is a few ways it can be done.14:14
Romsterchroots different distccd servers restrict the port you connect to other nodes.14:15
Romsteror you could have system wide multiple gcc/g++ compilers but use alias names like gcc-foo gcc-bar and set CC CXX for each.14:16
Romsteri'm keeping the same system on all so i wont hit any issues hopefully.14:16
Romsterdefently need a fast server as the master else it's not using all the nodes effectively.14:17
GoodboxI've builded a 2.6 kernel from i686 and x86_64 boxes, it was working perfectly with i686-pc-linux-gnu-gcc alias,14:18
GoodboxI added other i686 boxes, and that's fails, I think i'm gonna build only with i686 boxes14:19
Romsteri have a 32bit and multilib chroots i've used distcc over that fine.14:19
Romsteractaully i use ccache in fornt of distcc, i did however change ccache's hashing so it wont conflict as i share the same ccache cache between all on my desktop.14:19
Romsteri686 i built 3.0 ports for i686 crux 3.014:20
Romsternot made a iso though14:21
Romsterall these boxes i got are multilib.14:21
RomsterGoodbox, the issue is the gcc binutils glibc and other tool chain parts need to be identical on each machine.14:22
Romsterelse it'll fail.14:22
GoodboxOwh effectively versions might be differents14:23
Romstermore than likely.14:23
Goodboxi'm gonna chroot an squeeze i686 like on other machines14:24
Romsterif you can't update them you could build your own tool chain and use the alias for that tool chain on all the boxes.14:24
Romsteror a chroot that'll work.14:24
GoodboxOh yes, i thougt crosstool-ng something like that14:24
Romsterbut if oyu got more than one distccd running on that box you'll have to be careful on port restrictions too.14:25
Romsteri'm starting to get a small sizeable cluster to toy around with and iron the bugs out of distcc.14:26
Goodboxyou mean forwarding the distcc port inside the chroot ?14:26
Romsterno i mean if oyu run a system distccd you need to run one in the chroot but you need to use a different port.14:27
Romsterso you don't use the wrong distccd14:27
RomsterDISTCC_HOSTS="--randomize arkanoid/4 chunkserver1/2 master/2 metalogger/2"14:31
Romsteris what i got now.14:31
Goodboxthe --randomize reduces inactive time on the boxes ?14:33
Romsterspreads the load out it's mentioned in man distcc14:33
Romsterthan togo from a b c d it'll spread the load out evenly.14:34
Romsterelse c and d may not get much load.14:34
Romsteryou would usually list DISTCC_HOSTS from fastest to slowest.14:34
GoodboxHmm can be useful, however I set up dmucs who receive load average from distcc servers14:36
Romsteri might try that14:36
Romsterone step at a time14:36
Romsteri only just installed the 3rd box this evening.14:37
GoodboxYes, cause in my case, i've very differents machines14:37
Romster2 2GHz and 1 2.6GHz core 2's and my 3.4GHz quadcore on so far.14:38
Romstergot another 2.4GHz quadcore not setup with crux yet.14:38
Romsteri'm trying to get the same hardware, at lest no older than a core 2.14:39
Romsterbrb kitchen14:39
GoodboxI got an intel i3 four cores, three old amd-athlon64 two cores14:40
Goodboxbon appetit !14:40
Romsterchromium is a good test on distcc14:54
Romster613 ports14:57
Romsterdamn it's hard to cut back on ports.14:57
*** JanC_test has quit IRC15:02
sepenuname -m on a chroot won't work15:07
sepenalso you'll need to enable CONFIG_IA32_EMULATION in your kernel configuration15:09
jaegerIt would work with  setarch if you mean a 32-bit chroot in a 64-bit system15:13
jaegerI've used that to build ISOs in the past15:13
*** JanC_test has joined #crux15:14
sepenI used to write a little uname script and mv uname{,.orig}15:15
jaegertry "setarch i686 uname -m"15:15
jaegeror "setarch i686 chroot /path/to/chroot /bin/bash"15:15
sepenthat rocks thanks!15:15
jaegernp :)15:16
sepenI'm still building packages for my old i585 on a safe-crux15:16
sepenthat will help to me alot15:16
sepenwhois cody? (or l00nix?)15:20
*** guzzano has joined #crux15:21
horrorStruckyou can use linux32 to save some keystrokes15:23
*** dkoby has quit IRC15:24
Romstercrux chroot multilib 3.0# uname -m15:30
Romstercrux chroot i686 2.8# uname -m15:30
Romsteri made it work -_-15:30
Romster/bin/uname "$@"|sed -e's/x86_64/i686/'15:31
Romstersetarch why did i not see that.15:31
Romsteri'll have to try that later.15:33
Romsterpreparing for bed.15:33
Romsteri wont be able to match jaeger's machine but i'll surely get something impressive.15:50
tilmanhas probably been spammed in here already, but it's just too cool :P16:44
niklaswe :=)16:54
*** guzzano has quit IRC17:16
*** guzzano has joined #crux17:19
sepentilman: from hackaday ;D17:20
*** mike_k has joined #crux17:27
*** tilman has quit IRC17:30
*** tilman has joined #crux17:32
*** guzzano has quit IRC17:39
rmullAnybody else getting a segfault from mplayer 2013-03-22 with this command: mplayer -playlist
sepenworked fine here17:58
rmulldang, okay18:00
rmullwonder what's going on18:00
rmullthanks sepen18:00
sepengenerate a core dump, then 'gdb /usr/bin/mplayer /path/to/core' > bt18:00
sepenor try 'strace <command>'18:01
rmullhmm, need to figure out how to generate core dump. strace show a recvfrom() reading DNS data, then a close(), then the SIGSEGV18:15
sepenalso you should rebuild mplayer with debuging symbols and don't use 'strip' command18:18
*** guzzano has joined #crux18:19
rmullI seem to have disabled ELF_CORE in my kernel. I think it may be preventing my coredumps.18:23
rmullI'll start with that18:24
rmullthanks for the pointers sepen18:24
rmullInteresting... Fuduntu couldn't cope with systemd and gtk3 dependencies and is about to undergo major churn18:29
horrorStrucksounds like a late april's fool joke... gtk3 and systemd appear and suddenly the distro has to die because of that. super fucking weird blog post/decision.18:46
sepenlol, "The light version uses about 3-4 GB less hard drive space, depending on architecture." (fuduntu release)18:49
teK_rmull: trying.19:45
teK_plays for me19:48
sepenooops sorry, wrong win20:07
*** guzzano has quit IRC20:32
*** BurnZeZ has joined #crux20:45
BurnZeZWrong channel20:50
*** guzzano has joined #crux20:57
*** guzzano has quit IRC21:31
*** mike_k has quit IRC21:34
*** BurnZeZ has quit IRC21:35
*** timcowchip has joined #crux22:09
*** timcowchip has quit IRC22:11
*** timcowchip has joined #crux22:17
*** timcowchip has quit IRC22:17
*** timcowchip has joined #crux22:23
*** timcowchip has quit IRC22:24
*** comfree has joined #crux22:33
*** BurnZeZ has joined #crux22:37
BurnZeZSo what GUI web browsers would you guys recommend?22:44
BurnZeZI currently use dwb, but I'm looking for something with less bloat.22:44
rmullBurnZeZ: I use xombrero. Don't know about dwb though22:59
rmullxombrero is a webkit thing22:59
BurnZeZSo is dwb22:59
BurnZeZThey seem similar22:59

Generated by 2.11.0 by Marius Gedminas - find it at!