IRC Logs for #crux-devel Sunday, 2013-08-04

*** Romster has quit IRC00:54
*** Romster has joined #crux-devel01:00
*** heroux has joined #crux-devel01:26
*** _mavrick61 has quit IRC02:25
*** mavrick61 has joined #crux-devel02:26
jueRomster: IIRC you are using lvm2, right?09:49
Romsteryes i am actively using lvm2 and mdadm09:51
Romsteri'm also a helper in #lvm and #device-mapper09:51
jueyould be helpful if you could give latest version 2.02.99 a try09:51
jue*would09:52
Romsterjue, any chance jaeger's patch to libgmp can be applied?09:52
Romsteri can test that.09:52
jueoops, which patch?09:53
Romsterhttp://jaeger.morpheus.net/linux/crux/files/gmp32fix.zip09:53
Romsterit's in there it introduces a stub for gmp.h so i can fix up the paths for 32bit libgmp and nettle09:53
Romsterjaeger tested it on the tool chain i tested it with gnutls-32 test suite and wine iexplore https that tok me soem time to debug but foudn out 32bit header in libgmp-32 is needed for nettle-32.09:54
jueah, I see, it's similar to the patch we have for python09:54
Romsteryeah same issue with type differences.09:55
Romsterjaeger, wasn't happy where i stuck the libgmp-32 headers currently.09:55
Romstereven though that also works ok.09:55
Romsterwhile were on the subject of lvm2 what sort of changes were you thinking of for CRUX 3.1?09:56
jueof course I've no objections for the gmp changes, but it's jaeger decision to apply it09:56
Romsterjaeger got core access? or he'll let you know to apply it?09:57
frinnstuh, of course he has :)09:57
jueRomster: no idea, but if you have ideas how to improve the port09:57
Romsterok he should just push it then frinnst agrees too?09:57
Romsterwell you were talking about moving stuff out of init or something?09:58
Romsterpossibly having a init.d directory to place such add ons?09:58
frinnstlibgmp looks sane enough09:59
Romsterfrankly cryptosetup lvm2 and mdadm are the only ones that benefit of some init code to mount/activate.09:59
Romsterbut order of that will vary from user to user.09:59
Romsteror is it all ok as is.10:00
frinnstRomster: these can probably be moved to udev10:00
frinnstas with btrfs scan10:00
Romsterah so a udev rule then?10:00
jueRomster: we have a TODO31 entry for that, feel free to add ideas there10:00
frinnstyeah its in the todo for 3.110:00
jue:)10:00
Romsterbut is that wise will we be using udev in the future?10:00
frinnstfor 3.1 atleast10:01
Romsterwe don't ever plan to touch systemd right? we would go to mdev/udev fork/static nodes again?10:01
jueRomster: I don't see a alternative for udev currently, but we have the option to switch to eudev10:01
Romsteror will we be forced to systemd eventually? i see LFS is trialing it.10:01
frinnsthttps://gist.github.com/anonymous/6149936 <- my btrfs udev rule10:02
juethe biggest problem isn't finally the node generation, but the udev library which is a must have for other progs like xorg-server etc.10:03
Romsterit seems everyone else has switched to systemd without even any backlash?10:03
frinnstmd is already handled by udev, only lvm is in /etc/rc10:03
Romsterhmm lvm2 can be done with udev.10:03
Romsteri'll have to test that though.10:03
frinnstmm, would be brilliant if you could give it a go. i dont use lvm10:04
Romsterbut in future i fear we will be forced to follow the zealots with systemd. unless eudev is viable.10:04
jueFYI, running eudev on my laptop since some months, no issues so far10:04
Romsteris that the udev fork gentoo?10:04
Romsteror some other fork10:04
jueyep, it's in my private repo10:05
jueI mean a port for it10:05
Romstereep i talked to that dev only one working dev on it that's a professor at some uni school that's got very little time for eudev.10:05
*** deus_ex has joined #crux-devel10:05
Romsterso far ti's just udev with systemd bits striped out of it.10:06
jueIMO that's enough ;)10:06
Romsteri don't see it being viable currently.10:06
juesure, they just released version 1.210:06
Romsteroh hmm perhaps that guy has more time for it than he told me. the last time he said he hasn't got much time for it and will only track udev changes minus systemd stuff.10:07
Romsterin future he did mention a rewrite.10:07
Romsterbut that requires more effort.10:07
Romsterand what happens to the kernel side of udev?10:08
Romsterif that's changed to suit systemd, where does that leve eudev?10:08
Romsters/leve/leave10:08
Romsterthis issue has been concerning me for a long time now.10:09
jueas long as eudev follows systemd/udev I don't see such a problem10:09
Romsterfrinnst, cool i can probably knock up a udev rule for lvm2 quite easily using your example.10:09
Romsterjue, i just hope that is the case.10:10
Romsterdo you plan to use eudev on crux 3.1?10:10
Romsteralso we have nto had a meeting in a long time. do we have one coming up to plan for crux 3.1?10:10
jueI think we might consider eudev for 3.110:11
jueRomster: look at TODO31, there's everything we have so far10:11
Romstercompat-32.git10:12
Romsterevery time i push i get this non-fatal message: error: unable to update ./info/refs+10:12
Romsteri can edit that to include my changes that i need todo?10:13
jueFYI, just added an entry for eudev to TODO3110:14
Romsterlike boost that i've postboned (not making the same mistake i did with icu)10:14
Romsteryeah i see that, didn't know you just added it.10:14
*** jue has quit IRC10:27
*** jue has joined #crux-devel10:33
*** nrxtx has joined #crux-devel14:28
nrxtxhi14:28
Romsterlo14:56
*** jue has quit IRC16:01
*** nrxtx has quit IRC16:06
*** jue has joined #crux-devel16:12
*** felixos has quit IRC19:17
*** c0x` has quit IRC22:14
*** c0x has joined #crux-devel22:15
*** deus_ex has quit IRC23:56

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!