IRC Logs for #crux-devel Thursday, 2008-10-16

*** pitillo has joined #crux-devel00:07
*** Rotwang has left #crux-devel00:19
*** Rotwang has joined #crux-devel00:20
*** sepen has joined #crux-devel01:26
sepencptn, please could you take a look when having a piece of time? FS#36003:30
cptnsure03:31
cptnno objections from my side :-)03:31
cptnnot sure about the uf + um thing yet03:32
cptnbut I agree that i make sense addressing this03:33
sepenwell its not a big deal, but it could be fixed easily03:39
cptnyeah, it also makes the code harder to read though :-)03:40
cptnmaybe we could just exit with an error if both options are given03:40
cptnsince I can't imagine a situation where you'd do both at the same time03:41
cptnit's actually kinda dangerous03:41
cptnif you build a package -> fp missmatch03:42
cptnthen edit the patch, and run -uf -um03:42
cptnyou'll end up with an inconsistent port03:42
cptn-um should always require a rebuild03:42
cptnsince you changed files used during the build03:43
cptn-uf aleays happens after a build03:43
sepenwell, I mad a sample patch03:43
cptn*always03:43
cptnsure, it's good you reported it03:43
cptnI just thinking what might be the best solution03:44
sepenyeah, also would be good to see others pkgutils related bugs before bump out the 2.5 release03:45
cptnagreed03:45
sepenjust we could take some minutes of the next meeting for closing/modifying/etc. these bugs03:45
cptnsounds like a good idea03:46
sepen:)03:46
*** fetid has joined #crux-devel03:57
*** sepen has quit IRC04:22
*** fetid has quit IRC04:22
*** fetid has joined #crux-devel04:26
*** sepen has joined #crux-devel05:02
*** jue has joined #crux-devel05:10
*** fetid has quit IRC05:18
jueany news wrt test1-iso?05:39
sepenjue, I know that pitillo tests the complete installation, and also Ill try to test the upgrade method at night05:47
juefine, thanks05:47
sepenalso I've another offtopic question about ck4up, seems that after sfg changes we need to add 'package_id=' to our sourceforge lines05:48
sepenexample: amsn  md5  @SFID@54091&package_id=48692  @TAR@05:48
juehmm, I wrote about that to the ML sometimes ago05:48
sepenwhere @SFID@    http://sourceforge.net/project/showfiles.php?group_id=05:48
sepenohh so sorry05:49
juebut yeah, the man-page needs an update05:49
sepenjue, Im reading the sample conf in ck4up, so its fine forme05:49
sepenyep05:49
juehttp://lists.crux.nu/pipermail/crux-devel/2008-August/003726.html05:50
sepenfor what I know ck4up is listed as a freebsd port too, so would be nice to communicate it to freebsd comunity too05:51
sepenI really like ck4up, atm using it on my obsd box really fine05:52
sepen:)05:53
juenice to here that you like it :-)05:53
sepenalso I need a regexp tutorial near ck4up  muahhhahaa05:56
juebtw, did you solved the gdbm issues?05:57
sepenno05:58
sepenbut I think my dbm files are right05:58
sepenjue, I was playing with dbm files for compare these queries with /usr/ports files06:00
sepenI think it could be useful too06:00
sepenalso I'm using dbmtools for querying database files with no results06:03
sepenhttp://mikeux.dyndns.org/~sepen/files/dbmtools/06:03
sepen$ gdbmdump contrib.dbm06:05
sepengdbmdump: can't fetch "": Item not found06:05
juehmm, can you upload contrib.dbm somewhere, so I can try it?06:06
sepenand the dbm file is the one I'm using for checking my contrib port updates (without problems)06:06
sepenyep06:06
sepenhttp://mikeux.dyndns.org/~sepen/files/contrib.dbm06:07
sepenthanks in advance06:07
jueindeed, seems broken06:14
sepenhmm06:14
sepenjue, ck4up --cleandb uses the default ~/.ck4up/ck4up.dbm file right?06:16
jueyep06:16
sepenso would be nice to append the dbm file as a new argument to -c|--cleandb06:16
aonjue: do you get an md5sum mismatch with python 2.6 in the 2.5 branch?06:17
jueif you specify a different one with -f that will be usesd06:17
sepen/usr/bin/ck4up:186:in `size': lseek error (GDBMFatalError)06:17
sepen        from /usr/bin/ck4up:186:in `clean_db'06:17
sepen        from /usr/bin/ck4up:26706:17
sepenjue, ok using the -f switch its possible06:18
juebbl06:21
sepenok me too06:21
*** Rotwang has quit IRC06:42
*** Rotwang has joined #crux-devel06:44
*** jtnl has joined #crux-devel06:57
juere07:19
*** jtnl is now known as n00b07:20
*** n00b is now known as jtnl07:22
juesepen: yes, got a md5sum error too. They forgot to remove some .svn dirs07:23
juesepen: thanks, fixed now07:34
sepenhmm np07:41
sepenjue, aon says the md5 related suff07:43
sepenxDDD07:43
jue:)07:45
sepens/says/said07:45
jues***, as I feared, we run into trouble with libcap. The first port I found, who link against it, is coreutils.08:07
juewell, seems we have to move libcap/libattr to core08:10
sepenO_o08:11
aondamn09:27
pitillotalking about branches, moving core/opt rsync files from 2.4 to 2.5 in the test iso, is enought to start playing a bit?09:27
aonfeel free to move them, unless you wanna work around it09:28
pitilloatm I modified core to 2.5 and making a sysup09:28
aoni guess having them in core is better09:28
pitilloaon, is you need someone to check something, I can try it09:29
pitillo(I was trying to look for the python issue you told before)09:29
pitillonot sure if I am not enought patient but the sysup stalled in the man-pages update, in the last sentence of the Pkgfile, the touch one, I will try to let it a bit more time09:31
aonhmm09:31
pitillonow was built fine09:55
pitillonext opt09:55
pitillono diff on opt, moved to xorg09:57
tilmanhi10:24
pitilloyo tilman10:25
*** sepen has quit IRC10:31
*** pedja has quit IRC10:36
jueaon: IMO it's at least the less painfull way to go, but will wait until we have other opinions11:24
aonyeah11:24
juefrom tilman for example11:26
aonyup11:26
tilmanopinion about what?11:26
jueabout the libcap/libattr issue11:26
tilmanthe question is whether to move those to core?11:27
jueyep11:27
tilmando it!11:28
jueok :-)11:29
aon:D11:32
juetilman: have you tried the test1 iso?11:32
tilmanno, didn't get to it yet11:33
jue.oO and the pending udev release?11:38
tilmandidn't see that. hang on.11:38
*** sepen has joined #crux-devel12:16
*** sepen has quit IRC12:31
*** sepen has joined #crux-devel12:40
*** sepen has quit IRC12:54
jaegerI just downloaded the test1 iso myself, looking forward to checking it out12:58
*** sepen has joined #crux-devel13:19
cptnhey13:29
sepenhi cptn13:29
jaegerhrmm... for me, the test1 iso just repeatedly says "Abort installation?"13:34
jueoops13:42
sepenjaeger, do you have the last log? in your server?13:43
jaegerfor what? the 2.5-test1 install?13:48
*** Rotwang has left #crux-devel13:48
juejaeger: strange, just tested on a second laptop, works for me13:59
sepenjaeger, just I don't have the url to your logs :)14:00
*** Rotwang has joined #crux-devel14:01
juesepen: you mean the irc logs? -> http://crux.morpheus.net/irc/14:01
sepenjue, no no I ear about jaeger's iso logs14:02
sepenor maybe I'm wrong14:02
jaegerI don't understand what you're asking. Do you want the log from jue's test1 setup script?14:06
sepenohh yes14:06
sepenexcuse me so14:06
sepenwould be interesting for reading it14:06
jaegerdoesn't seem to keep one14:08
sepenok14:09
jaegernot getting far enough for the package logs, etc.14:11
*** Rotwang has left #crux-devel14:14
sepenwell imho could be interesting some kind of stuff for getting log/debug messages14:14
sepenI mean in the building process and also when booting the iso for debug14:14
*** Rotwang has joined #crux-devel14:16
juewell, I was too fast wrt libcap/libattr, libcap support in coreutils comes with 7.0, which is in beta now but I had installed, and can be disabled via configure option.14:17
*** treach has joined #crux-devel14:17
jues fresh build of all core ports with installed libcap/libattr shows that only libarchive links against libattr14:18
cptndamn, I keep messing up my git checkout :-(14:28
cptn"warning: 4 lines add whitespace errors"14:29
cptnin gegl, which I never touched myself14:30
juetilman: seems that should add --disable-xattr to libarchive14:32
jaegerooh, got successfully through stage1 this time, on xfs instead of reiserfs15:10
juejaeger: have you tried test1 on a second box?15:13
jaegernot yet, I will when I get home15:13
juethe "Abort installation" loop came up if you run setup?15:14
jaegeryes. I started setup, selected all 3 collections, didn't choose to manually alter the selections15:16
jaegerit checks dependencies and then the repated abort screen comes up15:16
jaegerselecting yes exits the setup and no repeats15:16
cptnI've just updated my laptop15:19
cptnworked fine so far :-)15:19
cptnexcept that it overwrote jw/blackbox with opt/blackbox ;-)15:19
*** DarkNekros has joined #crux-devel15:21
juejaeger: you did/tried a fresh install15:24
jue?15:24
jaegeryes, that was a clean install, not upgrade15:31
cptnnspr wasn't injected for firefox15:37
cptnI guess this is mainly caused because I didn't update firefox earlier15:37
cptnthen: "firefox: symbol lookup error: /usr/lib/libcairo.so.2: undefined symbol: pixman_region32_init"15:38
juehmm, nspr is listed for firefox in setup.dependencies15:43
cptnmmmh, for some reason a number of xorg packages wasn't correctly updated either15:45
cptn nor xorg-libpciaccess15:48
cptnnot sure what I did wrong... just ran an regular update15:48
cptnno manual intervention15:48
jueeither we have a bug in setup or the whole dependency stuff doesn't work in the expected way15:51
cptnor I did something wrong :-)15:53
juefor sure not, unfortunately15:55
*** jue has quit IRC16:05
*** jtnl has quit IRC16:12
*** jtnl has joined #crux-devel16:13
cptnmmmh, well16:20
cptnI guess I see the problem16:20
cptnthe update action only looks at core and opt16:20
cptnnot xorg16:20
cptntilman: it might make sense to inject xorg-libpciaccess; however only if a user has xorg installed...16:23
*** DarkNekros has quit IRC17:08
*** Rotwang has quit IRC17:19
*** jtnl has quit IRC17:34
*** treach has quit IRC17:45
*** jtnl has joined #crux-devel17:54
*** jtnl has quit IRC18:31
*** fetid has joined #crux-devel20:34
*** sepen has quit IRC20:43
jaegerfuck it, I give up22:46
jaegerI'm too stupid to build an ISO, apparently22:46
*** Rotwang has joined #crux-devel23:04

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