IRC Logs for #crux-devel Tuesday, 2010-07-13

*** jue has joined #crux-devel06:55
*** j^2 has quit IRC08:18
*** j^2 has joined #crux-devel08:20
*** deus_ex has joined #crux-devel09:52
*** deus_ex is now known as pedja09:53
jaegerlibmpfr source url needs an update in 2.711:09
juejaeger: thanks, fixed11:51
*** frinnst has joined #crux-devel12:15
jaegergit in 2.7 has perl 5.10/5.12 and python 2.6/2.7 mismatches12:45
*** frinnst has quit IRC13:19
*** frinnst has joined #crux-devel13:21
tilmanmaybe we should have btrfsprogs for the 2.7 iso13:37
juehmm, I was kind of hoping that we announce -test1 soon ;)13:45
tilmanyeah, sorry about that13:45
juedon't worry, we should consider announcing test1 as it is and add btrfsprogs in the next step13:52
jaegerseems reasonable13:52
tilmanif you agree that btrfs support might be desirable, we also need to touch the kernel config for it13:53
tilmanannouncement set to crux-devel13:54
juesure, at all a good idea13:55
jaegerbtrfsprogs will go into opt?13:56
tilmancore i think?13:56
jaegerprobably, I just looked at where the other *fsprogs are13:56
tilmani don't think we have a fs support package that's not in core13:56
jaegerexcept ntfsprogs13:56
tilmanprobably because that one is for fuse13:57
jaegerand not useful for installing crux anyway :)13:57
jueyeah, I'd see it in core too13:58
juetilman: do you have a port?13:58
jaegerhow long before someone downloads -test1, installs it, runs 'ports -u', 'prt-get sysup', and breaks everything with 2.6's libgmp? :D13:59
tilmanjaeger: did you report a python footprint mismatch for a port other than git to me some weeks ago?13:59
tilmani thought i fixed it, and i thought that one was git...13:59
jaegerI don't remember, honestly14:00
jaegerit might have been git but it's probably been re-merged since then14:00
juetilman: it was a xorg port14:00
juexcb-something IIRC14:00
tilmanoooh, right14:01
tilmanjue: not yet, i can write one in 5 minutes14:06
juetilman: no hurry, just curious14:06
juejaeger: why that? ports (the port) is up-to-date, all rsync files sould point to the 2.7' repos14:08
jaegerjue: hrmm, I'll check again14:30
jaegerthe install from the iso had 2.6 in the .rsync files14:31
jaegermaybe the package isn't updated14:32
jaeger(on the iso, that is)14:32
jaegeryes, it's the package, damn it14:33
jaegermaybe we should mention that on the ML14:34
jaegerdid the .rsync files get updated without bumping the version? I probably had the package already built so it didn't get rebuilt14:34
tilmanno, i changed the version from 1.5-9 to 1.5-1014:36
jaegerinteresting. ports isn't one that gets updated after a -test1 install14:36
jaegerI'm confused now14:36
tilmanme too14:47
tilmanthere don't seem to be any recent btrfs-progs tarballs14:47
tilmanmmh, grammar.14:47
jaegerI'm sure it's my fault somehow since I built the ISO but I don't see where I did it =/14:48
tilmanfault schmault14:48
tilmanit's easy to spot and fix14:48
tilmanno worries ;)14:48
jaegerIt's easy to fix, it just bothers me that I can't figure out where the problem started14:50
jueadded a sentence at the bottom of TODO2714:58
jaegerI haven't yet tested the usb image build script with 2.7, will do that tonight15:01
jaegeron a side note, -test1 didn't have pkgutils in .pkg.tar.gz format15:02
juetilman: would be helpfull to have a link to ReleaseNotes2-7 in the mail?15:02
juejaeger: should be in /tools?15:02
jaegerWhat's the use-case for that? When someone wants to upgrade without booting from the ISO?15:02
jueisn't it15:02
tilmanjue: will remember to add it next time15:02
jaegeroh, maybe it was, I'll check15:02
jaegerah, it is, sorry :/15:03
jaegerI'm spaced out today, apparently15:03
tilmanshouldn't btrfs-progs install "fsck.btrfs"?15:06
tilmanso the generic fsck program can handle btrfs-s?15:07
jaegerprobably a good idea15:08
jueindeed, at least all other filesystem utils do that15:11
juethe same with mkfs.*15:11
tilmanlet's not symlink it for now15:16
tilmanshould i commit the port to core/2.7?15:25
jueyeah, go for it15:30
jaegerI'll make a -test2 with that and the fixed ports port if we like15:32
tilmanlet me first adjust the iso's kernel config15:33
tilmanare your latest changes pushed to iso.git?15:33
juewe have to add it to Makefile/ISO_PACKAGES as well15:34
jaegerthey're not pushed yet, want me to go ahead and do that?15:35
jaegerjue: yes15:35
tilmanyes please, unless you can think of a reason not to :p15:35
jaegerok, give me a few to make sure I'm synced up15:36
jaegerI've got to find a better way to do this15:40
jaegerI build the ISOs as root but commit as regular user15:41
juetilman: it depends on util-linux-ng15:41
jaegerfor now I copy a diff to my normal user, apply it, and commit15:41
tilmanjue: yes. hmpf. forgot i have to list that one15:42
jaegerok, pushed, tell me if you see something I messed up15:46
jaeger(this push doesn't have the btrfs change)15:46
tilmanlooks good i think15:47
tilmandid you forget to add kernel/linux-2.6.34.config (and remove linux-
jaegerno, but I should have done that in the same commit15:48
jaegerinstead of the one I'm about to do15:48
* jaeger sighs15:48
jaegersame with the busybox config15:48
jaegergod damn it, wtf did git do there?15:51
jaegerit did the right thing with the kernel config but not with the busybox config. what the fuck15:51
tilmanbecause it didn't recognize it's a rename?15:52
jaegerI guess15:52
jaegerI don't understand git sometimes15:52
tilmanyou didn't remove the old busybox config file15:52
tilmanwe _could_ fix this by amending your latest commit15:53
tilmanbut then we'd break the repo kind of15:53
jaegerwhat's your preference? amend or just commit a delete?15:53
tilmancommit a delete please15:53
jaegerok, I think it's in a useful state now15:54
jaegersorry for the noise/confusion15:55
tilmannot a problem15:57
tilmanfwiw, if you would like to check that your renames worked, you could run the following after committing:15:57
tilmangit show --stat HEAD15:57
tilmanthat shows a diffstat, too15:57
tilmanin this case, it shows that you added a busybox config, and it shows that you did not remove another15:58
tilmanbtw, has anyone used CONFIG_EXT4_USE_FOR_EXT23 before?15:58
tilman"Allow the ext4 file system driver code to be used for ext2 or ext3 file system mounts."15:59
jaegeryeah, not sure why it acted differently there15:59
jaegerbecause adding the kernel worked as a rename15:59
jaegerjust goes to show I should check yet another level of git before pushing15:59
tilmanoh, i assumed you mv'ed the file15:59
jaegerI just want it to be consistent, which is apparently too much to ask16:00
jaegerwhat I did was copy the new config into the directory, remove the old one, and add the new one. git usually handles that fine16:00
tilmanyes, exactly16:02
tilmanare you sure you didn't just forget to remove the old bb config, too?16:02
tilmanalright, i pushed what i think enables btrfs support in the iso16:04
tilmanupdated kernel config, Makefile:ISO_PACKAGES, and iso/setup.dependencies16:04
*** pedja has quit IRC16:04
jaegertilman: I'm positive, the file was gone16:08
jaegerah, ok. I had those changes locally :)16:08
jaegerIs btrfs-progs (with the dash) upstream's name?16:09
jaegerjust wondering since it's different from the others16:09
tilmani was thinking about calling the port btrfsprogs16:10
tilmanbut that's kind of ugly if it's not upstream's name16:10
jaegeryeah, agreed16:10
jaegerit's fine with me, just wondered16:10
jaegerand that commit looks ok to me16:10
jaegerI'll build a -test2 with it16:11
teK_jaeger: have you changed any s-ata drivers, yet? :)16:11
jaegerno, I'm still waiting for you to tell me the ones that need changing16:12
jaeger*I* have not run into any problems16:12
teK_looking the drivers up :)16:12
jaegertilman: to answer your ext4 code for ext2/3 question, I have not16:18
jaegerI haven't even used ext4, to be honest16:18
tilmanthe only benefit is to save some space (smaller kernel image)16:20
tilmanresp fewer modules16:20
tilmanso until that becomes a problem, we can ignore that option :]16:20
jaegerI haven't tried it but if it's stable enough then I wouldn't mind16:21
teK_jaeger: got it. CONFIG_SATA_AHCI=y it's for the Intel ICH9 S-ATA Controllers16:29
teK_< looking if it's not yet enabled.16:29
tilmanit's a module in 2.7-test116:31
jaegerahci has been enabled for quite some time in my ISOs16:35
jaegerit's not in 2.6 but is in 2.6-newmedia and 2.716:36
teK_ah k :)16:37
jaegerIt gets loaded in load_sata_modules() in initramfs/init, for reference16:38
jaegerteK_: did you run into a machine on which it didn't work properly? if so, was it 2.6 vanilla or one of the updated ISOs?16:40
teK_2.6 vanilla16:48
teK_was a lenovo R50016:48
jaegerI bet it'd work fine with -updated or 2.7-test116:49
jaegerI've used -updated on a T400 for what that's worth16:50
teK_well my friend will want to upgrade to 2.7 soon (when the 64bit falvour is available) ;)16:50
teK_but i'm quite sure that it'll simply work16:51
jaegergoing home, will be back on later17:05
*** jue has quit IRC17:09
*** mavrick61 has quit IRC21:45
*** mavrick61 has joined #crux-devel21:46

Generated by 2.11.0 by Marius Gedminas - find it at!