IRC Logs for #crux-devel Saturday, 2014-02-15

*** mavrick61 has quit IRC03:58
*** mavrick61 has joined #crux-devel03:59
Romsteri need to test eudev with my lvm mdadm setup.04:58
*** Romster has quit IRC08:39
*** Romster has joined #crux-devel08:40
*** deus_ex has quit IRC11:49
*** deus_ex1 has joined #crux-devel11:49
*** rexich has joined #crux-devel13:07
*** rexich has quit IRC14:12
*** deus_ex has joined #crux-devel14:56
*** deus_ex1 has quit IRC14:58
jaegergood morning15:03
*** deus_ex has quit IRC15:20
*** deus_ex has joined #crux-devel15:23
juejaeger: what's your idea for a test release? Is 2 weeks a realistic time frame?15:28
*** deus_ex has quit IRC15:28
*** deus_ex has joined #crux-devel15:29
jaegerThat's probably fine, I think. I should quiet the git hooks, forgot about that yesterday15:32
*** deus_ex has quit IRC15:38
jueok, waiting for that, will merge the core branch afterwards15:53
jaegerLooking at them now, will start with core15:58
jaegerperl isn't my strongest language but I think this is simple enough: next if $ref == 3.1;16:02
jaegerok, I think they're ready. If you get any errors let me know and I'll investigate16:10
jueok, thanks16:10
jaegerno problem. looks like the core merge was ok16:14
jueyeah, looks fine16:15
juewill now commit the sysvinit/util-linux/procps changes16:15
jaegerAnything in particular you'd like me to work on?16:15
juewell, one thing that is a bit problematic is the new hwdb of eudev; currently we need a post-install action to create the binary db from the sources16:19
jueit might be possible to do this within the port?16:19
jaegerIf it can't be done withing build() perhaps we can ship a starter version16:19
jueyep, something like this16:20
frinnstwee, stuff's happening16:37
frinnsti'll do some tuneup on xorg 3.116:37
juegood :)16:49
jueFTR, I merged the 3.0 branch into 3.1, we should probably keep this direction for future merges to avoid problems16:51
juemore precise, I did the following:16:52
jue git checkout 3.116:52
jue git pull . 3.016:52
frinnstdo we have rsync hooks for 3.1? it would help i think16:53
jaegerI enabled them16:53
jaegerif I missed any let me know16:53
jaegerI think I did core, opt, xorg, contrib, compat-3216:53
jaegerI didn't mess with the outlier repos like kde, etc.16:53
frinnstUpdating file list from
frinnstrsync: change_dir "/crux-3.1/core" (in ports) failed: No such file or directory (2)16:55
jaegercore/3.1 exists, will check if there's a missing symlink or something16:55
jaegerah, crux-3.1/* will be symlinks to the others. sec, will fix that16:55
jaegerok, try now16:57
frinnst <- my glibc 2.19 port16:57
frinnstdid we decide on the version?16:58
frinnstjaeger: yeah works16:58
jaegerIf 2.19 is considered stable upstream I'm not adverse to testing it16:58
frinnstshould i add my port then?16:59
frinnstand gcc too, perhaps16:59
jaegerNo objections from me. jue?17:00
jueyeah, sure. We are waiting for it :)17:02
juefrinnst: binutils?17:02
frinnsti've been running with 2.24"17:04
frinnstnot really kept up to date with that17:04
frinnstThe latest release of GNU binutils is 2.24.17:04
frinnsti'll push that too17:05
frinnstI dont see perl anywhere on the todo. i have a 5.18 port ready, if you want to try it17:09
frinnstah, now i see it17:10
jueit's in the postponed section, running the test-suite right now17:10
juebut it, please17:11
frinnstfound a perl problem with openssl:
juelet's add the patch to openssl17:20
juebtw, the perl test-suite just finished with "All tests successful."17:21
juemarked perl as done in TODO3117:23
frinnstnoticed that the kernelheaders for glibc have a missing $source17:55
frinnsti've uploaded the tarball to my homedir (/home/rehabdoll/), if someone with root-privs can move it to distfiles17:56
jaegerThe ISO's currently using 3.10.28 for what that's worth. I don't think that's any issue for the headers, only if --with-kernel is set too high17:58
jaegerer, --enable-kernel17:59
jaegerI kinda like the idea of using a longterm kernel on the ISO because it usually works with all the extra stuff like vmware tools, virtualbox additions, zfs, etc.17:59
frinnstyeah, i just used the latest kernel when generating the headers18:01
frinnst3.10 is probably a better bet for use with glibc18:01
frinnstfeel free to change it18:02
frinnstafk for a while18:02
frinnstcool, i get a kernel too old when building glibc on a 3.10.25 vm18:09
jaegercurrently it has --enabled-kernel set to 3.1218:09
frinnstplease change :)18:09
frinnstafk again18:09
jaegerheh, ok18:09
frinnstoops, seems like i fucked up the merge of mesa3d19:24
juelooks like we've done quite some work today, thanks guys :)19:25
frinnsteudev is missing from core. are there still work to be done on it?19:27
juenot really, will add it19:27
juethe post-install thing is a 'issue' but jaeger already mentioned a fix for it19:28
jueAFK now, back tomorrow19:35
*** deus_ex has joined #crux-devel19:48
frinnstlibjpeg -> libjpeg-turbo will require quite a few ports to change the dependency line in the Pkgfile20:14
frinnstopt still seems a little broken:20:16
frinnstWriting objects: 100% (4/4), 396 bytes | 0 bytes/s, done.20:16
frinnstTotal 4 (delta 3), reused 0 (delta 0)20:16
frinnst  File "hooks/post-receive", line 8420:16
frinnst    if $branch == "3.1":20:16
frinnst       ^20:16
frinnstSyntaxError: invalid syntax20:16
jaegeroops, mixed my perl and python. sec20:23
jaegerfixed, I think20:24
frinnstbah, lots of conflicts in opt20:37
frinnsthopefully everything is ok now21:00
jaegernice :)21:15
jaegerdoing some builds here21:15
prologicmike won't mind if I update pip in contrib will he? :)22:12
prologicit's a) out-of-date and b) broken22:13
frinnstsend a mail and ask first22:19
prologicsure :)22:22
frinnstjue: i removed opt/whois in 3.1 fyi22:26
frinnstsome more glibc / gcc cleanups22:59

Generated by 2.11.0 by Marius Gedminas - find it at!