**** BEGIN LOGGING AT Sun Apr 16 09:59:56 2006 Apr 16 10:14:11 FYI: a complete rebuild-from-scratch of all nslu2-linux feeds is about to begin. Apr 16 10:51:11 03rwhitby * 10unslung/make/coreutils.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:51:50 03rwhitby * 10unslung/make/diffutils.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:52:25 03rwhitby * 10unslung/make/findutils.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:52:51 03rwhitby * 10unslung/make/grep.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:53:40 03rwhitby * 10unslung/make/procps.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:54:11 03rwhitby * 10unslung/make/tar.mk: Changed the conflicts from busybox to busybox-links. Apr 16 10:54:34 03rwhitby * 10unslung/make/textutils.mk: Changed the conflicts from busybox to busybox-links. Apr 16 11:02:02 03rwhitby * 10unslung/Makefile: Updated clobber to remove nas100d too. Apr 16 12:22:46 hi Apr 16 12:23:03 i just upgrade miau and now its not working anymore Apr 16 12:23:21 doesnt open the irc port Apr 16 12:23:35 dont know if it reads the config Apr 16 12:23:50 and it stays in foreground Apr 16 12:43:53 when did the make/miau.mk change last in the cvs? Apr 16 12:44:43 Looks like you need to talk to caplink811_log Apr 16 12:45:06 date: 2006/04/13 17:55:37; author: caplink811; state: Exp; lines: +7 -7 Apr 16 12:45:07 changing to upstream version 0.6.0.2, bumped ipkg version Apr 16 12:45:19 date: 2006/04/12 20:03:27; author: caplink811; state: Exp; lines: +3 -0 Apr 16 12:45:19 beginning with work on miau-0.6.0.2, compiles like a charm, but the binary produces a lot of processor load at startup, a\ Apr 16 12:45:19 nd will do nothing, so i must have a closer look on it. Apr 16 12:47:35 ah ok Apr 16 12:50:30 that's the last two cvs entries (both from caplink811_log) Apr 16 12:50:45 the one previous to those is from Dec last year. Apr 16 13:57:08 03caplink811 * 10unslung/sources/miau/postinst: forgot to add a start parameter, now postinst works too Apr 16 15:56:25 morning, i've updated the package miau, so it reflect the new version in the upstream, the package builds cross and native, but only the native builded version runs fine. so what is the best way, go back to the old version, or mark this package as "only build native" in the optware makefile ? Apr 16 17:00:42 03bzhou * 10unslung/ (make/flip.mk Makefile): flip ready for testing Apr 16 20:32:38 03bzhou * 10unslung/sources/python/disable-ncursesw.patch: patch to explicitly disable ncursesw for those platforms with no ncursesw support Apr 16 20:50:02 03bzhou * 10unslung/make/miau.mk: made cross config.h matching native config.h Apr 16 20:52:16 03aabaker * 10unslung/make/nail.mk: nail: version upgrade, wl500g compatibility and generate ipkg control file Apr 16 20:53:25 AdamBaker, morning, what would you do in such a case ? Apr 16 20:53:29 morning, i've updated the package miau, so it reflect the new version in the upstream, the package builds cross and native, but only the native builded version runs fine. so what is the best way, go back to the old version, or mark this package as "only build native" in the optware makefile ? Apr 16 20:54:36 I'd say it depends why you have upgraded to a new version Apr 16 20:55:06 some features are new.. Apr 16 20:55:48 If the new functionality is worth the cost of builds being slower then mark it as native only if it was NSLU2 only already Apr 16 20:57:10 03aabaker * 10unslung/sources/nail/control: nail: remove control file from sources as it is now created by the Makefile Apr 16 20:57:22 but dunno, if anybody needs those features... maybe it worth to went backwards to the old version Apr 16 20:57:41 s/it/it is/ Apr 16 20:57:41 caplink811_log meant: but dunno, if anybody needs those features... maybe it is worth to went backwards to the old version Apr 16 20:58:25 caplink811_log, i just checked in a version of miau.mk that generates config.h exactly matches native build Apr 16 20:58:41 maybe you can test it again? Apr 16 20:59:50 yes, i'll do that tomorrow, do you think, it depends on config.h, i assume, it was a problem with paths.patch.. Apr 16 21:00:55 i have no idea about miau, but it's always a good idea to have the cross config.h matching native one Apr 16 21:01:09 :) i'll try Apr 16 21:01:47 I've just checked and miau is in the wl500g feed so a native only package would reduce the supported platforms Apr 16 21:02:34 If a working package is currently built cross, and an upgrade can only be built native, then please do not commit the upgrade. Apr 16 21:02:36 03aabaker * 10unslung/Makefile: nail: mark as ready for testing and remove from WL500G_BROKEN Apr 16 21:02:54 k Apr 16 21:05:28 looking at the paths.patch and paths.patch.0.6, does not seem to have anything to do with cross compilation Apr 16 21:06:48 yeah, but something went wrong, gdb isn't realy my friend, but it seems, that the process, which allocate space for variables stook somewhere Apr 16 21:07:15 when i used a cross compiled binary Apr 16 21:07:36 maybe try strace native-built-binary vs strace cross-built-binary? Apr 16 21:08:04 no difference there Apr 16 21:15:25 eno-away, nice idea, it seems that it works now :) Apr 16 21:15:37 cool, everybody's happy Apr 16 21:15:45 yep :) Apr 16 21:26:25 btw, is it neccessary to increase the ipk-version, to reflect the last changes, or will the package build automagically (with the last changes in it)? Apr 16 21:26:49 you always need to increment ipkg_version Apr 16 21:26:57 k, thx Apr 16 21:29:13 but eno-away have done it, thx :) Apr 16 23:23:46 optware/nslu2 fails to rebuild php. Apr 16 23:24:12 copying over the old build directory to enable the feeds rebuild to continue ... Apr 16 23:24:26 (the old builds/php directory that is) Apr 16 23:25:13 If someone could do a php optware/nslu2 compile from scratch and see if it builds, that would be helpful. Apr 17 00:49:58 rwhitby, very likely libxml2 & limxslt upgrade broke php, adding -pthread to PHP_LDFLAGS probably will fix the build Apr 17 00:50:48 just like cvs diff -r1.2 -r1.3 make/php-fcgi.mk did Apr 17 00:52:02 i'll do a clean build to make sure Apr 17 00:52:29 thx Apr 17 00:52:42 rebuild is up to cups Apr 17 02:02:05 optware/nslu2 fails to build freeradius Apr 17 02:02:27 can't find -lssl, but openssl is definitely staged Apr 17 02:23:17 rebuild is up to groff Apr 17 02:23:58 03bzhou * 10unslung/make/php.mk: added -pthread to LDFLAGS to unbreak the build Apr 17 02:29:55 So far this is what I've been seeing break on my builds of optware as well. Samba will be breaking, too, I bet Apr 17 02:31:53 I've changed the autobuild script and the distro Makefiles to pass the -k flag to make and bitbake, so that the good packages are built and released if a single package doesn't build. Apr 17 02:32:27 Probably a good idea; I've never yet managed to build all of optware in one go. Apr 17 02:33:05 it happens about once every six months :-) Apr 17 02:34:28 Wow - with the rate the oe stuff changes, I'm surprised it happens that often! Seems every time I try to build optware, something in the toolchain, or something in bitbake has changed... Apr 17 02:50:13 well, it happens that often cause that's how often I do a full rebuild ... Apr 17 02:50:21 (and fix all the problems) Apr 17 05:49:00 03bzhou * 10unslung/make/py-moin.mk: upstream upgrade to 1.5.3 Apr 17 07:18:56 Any provisions when packages will be back? It looks like nudi autobuild does not copy from build dir anymore. Apr 17 08:12:53 oleo: it' Apr 17 08:12:57 s still building **** ENDING LOGGING AT Mon Apr 17 09:59:56 2006