IRC Logs for #crux-devel Tuesday, 2009-06-30

*** sepen has joined #crux-devel01:03
*** jue has joined #crux-devel06:16
jueFYI if you are using the 2.6 branch ->
jueteK_: you've read my mail to crux-devel wrt lvm?07:05
teK_yeah I did. I just had a look at your commits which look reasonable07:12
jueok, thanks07:13
teK_the rc.conf-part could really be left out, I think07:13
teK_so an overall +1 from me :)07:13
jueif you have additions etc., please let me know07:13
teK_the reason I did not try to include lvm on my test iso (with dmcrypt) was that I thought it'd be a nobrainer to do so :)07:14
teK_I think this is fine07:14
juewell, with the changes from today to e2fsprogs and util-linux-ng we are almost done for 2.6 IMO07:16
jueexcept xz support for pkgmk07:17
jueand a decision regarding rsyslog07:18
teK_so I will have to provide a (possible) patch for dmcrypt better now than later, right?07:19
teK_I know your objections regarding this topic ;)07:19
jueabsolute, if we have an official -test1 it's too late07:20
teK_I will git clone 2.6 tonight and have a look :)07:21
aonIf this is not possible, please let the terms unnecessarily overwrite07:34
aonindividual ports with support -nls"07:34
aoni wonder what that means07:34
jueaon: where does this come from?07:38
aoncrux@ more specifically07:41
sepenaha, it's a lot of work to do, isn't as easy as remove --disable-nls lines in Pkgfiles ;D07:43
juesaw it now, but dosn't understand it07:45
sepenwell IMHO he would like to have native language support by applying individual patches to Pkgfiles07:48
juesepen: almost, seems that he wants to keep the docs as well07:48
sepenI started something similar sometime ago to keep only spanish doc, but it's really a lot of work07:49
sepenI discarded this idea07:49
nipuLi think that was something i brought up (nls)07:50
nipuLrather than disable nls in the pkgfile, simple add an install rule for /usr/share/locale07:50
sepenwell I can't understand the line 'Collier ...'07:52
aonbut the packages will be larger07:52
aondisk space is expensive! :|07:52
sepenaon, lol07:52
nipuLyeah i know, my last hdd cost about 20ยข/GB07:53
juehmm, and the footprint doesn't show exactly what you have installed07:53
jueat all I think that people who wants localized programms needs the docs in their language as well07:54
juee.g. man-pages etc.07:54
sepenit could be possible with some patches for pkgmk07:55
sepenbut -1 imho07:55
nipuLusing install rules rather than explicitly deleting them has it's advantages07:55
nipuLit lets the end user define what is "junk"07:56
nipuLthe same could be applied to documentation07:56
aonin some cases i'd like info pages but no way in hell will i install localized manpages :)08:02
sepenyou can maintain the foo-info port in your private repo or locally08:02
aonyes, i'm aware of that08:04
aongtg home ->08:04
nipuLso why was the install rule implemented if no one wants to use it?08:04
juethe info-page issue was already discussed but deferred because of the needed post-install/pre-remove actions IIRC08:04
juenipuL: good question, dunno remember why08:06
nipuLinstall/remove hook rules for pkgadd.conf could solve that08:07
nipuLi'd personally like such a feature08:08
nipuLfor a range of uses08:08
juefor what especially?08:09
nipuLit could replace the need for a number of install/remove scripts08:10
nipuLit would make maintaining haskell  packages a lot easier08:10
nipuLthat or add more oppurtunties to run helper scripts in prt-get08:14
nipuLpre-remove for example08:14
juenipuL: would be great if you write down such ideas somewhere, you've seen the PkgutilsIdeas wiki-page?08:17
sepencalled %preun by fedora's spec files08:17
nipuLremind me tomorrow, i'm on the netbook atm, it's not fun typing for extended periods of time on it08:22
jueok :)08:22
juenipuL: found it, the initial reason for the addition of INSTALL to pkgadd was indeed the info-page discussion08:59
juepoint 3)08:59
jueTBH I don't remember why we never carried on with that discussion, maybe we recongnized that we need a install/remove hook that calls install-info09:17
juetilman probably knows more09:18
*** Rotwang has joined #crux-devel09:34
*** sepen has quit IRC10:20
rehabdollfirefox 3.5 seems to be out11:59
strawoi, nice :>11:59
*** mike_k has joined #crux-devel12:29
tilmanjue: exactly: we would need the hook stuff to properly update info page indices or something12:34
juetilman: ok12:54
juetilman: I guess you are not going to implement hooks in current pkgutils, so this has to wait until pkgutil6 aka CRUX 3.012:56
tilmanyes :(12:56
*** treach has joined #crux-devel13:19
*** sepen has joined #crux-devel13:23
*** mike_k has quit IRC16:04
teK_is there a crux-2.6.iso yet?16:19
jueonly an unofficial one ->
teK_and new_iso-buildref.txt really is the recommended way to build the iso?16:29
juedunno, never used it16:30
jueIIRC it is from Matt16:31
teK_imho it's unusable, last time I built the ISO I did it free-style, iirc. :\16:32
juewell, the makefile provides all you need if you have the ports-tree in place16:34
*** sepen has quit IRC16:36
jueteK_: hmm, looks not unusabel to me16:43
*** jue has quit IRC16:57
*** Rotwang has quit IRC17:01
jaegerI don't think it's up to date but it shouldn't be far off17:04
teK_I did not mean to be mean :)17:31
rehabdollutil-linux-ng 2.16-rc1 fails without -j1 for me17:35
jaegerteK_: didn't take it that way :)17:49
jaegerI wonder if I can replace the failing 160G drive with a 250G drive and later replace the other 160 with a 250 and grow the array at that point18:07
jaegerpresumably the first step is no problem if I use the same partition size on the larger drive18:08
teK_with LVM?18:09
teK_puh :)18:11
jaegerI found someone's blog post describing the same thing I want to do, looks like it just works like you'd expect18:12
jaegerI suppose I could easily test this in a VM, too18:13
teK_well, you have a backup[tm], don't you ;)18:14
teK_rehabdoll: is right,  util-linux-ng fails without -j1, here, too.18:26
rehabdollofcourse im right18:28
rehabdollim always right18:28
*** treach has quit IRC19:49

Generated by 2.11.0 by Marius Gedminas - find it at!