**** BEGIN LOGGING AT Thu Aug 19 02:59:57 2010 Aug 19 03:06:02 jkridner|work: well, it does not use your m4. Aug 19 03:06:36 yeah, I was trying to figure out why just now. Aug 19 03:06:54 do I need to have m4 installed on the host? Aug 19 03:07:45 this doesn't seem like the right thing to do to me: export ac_cv_path_M4="${ac_cv_path_M4=/usr/bin/m4}" Aug 19 03:08:13 how is it this stuff works for anybody? :( Aug 19 03:08:57 that was pulled out of run.do_compile Aug 19 03:11:59 hmmm, why is fix_path_xtra.patch only used on native? SRC_URI_append_virtclass-native = " file://fix_path_xtra.patch" Aug 19 03:15:26 jkridner|work: this is autoconf for target right ? Aug 19 03:15:37 yeah, it is. Aug 19 03:15:52 I dont think very many build it Aug 19 03:16:16 right. took me a minute to figure that it wasn't -native. Aug 19 03:16:38 still, I don't know why removing the hard paths was only done for -native. Aug 19 03:17:29 try to enable this patch Aug 19 03:17:37 this patch is needed for cross build Aug 19 03:17:43 native should work without it too Aug 19 03:27:06 zenity fails? what *is* zenity? trying a 'bitbake autotools' to test out that patch. Aug 19 03:37:47 it is such a cheat, but I gave up and did 'sudo aptitude install m4 -y'. Aug 19 03:39:23 that particular patch doesn't do anything to remove the hardcoded path for m4. Aug 19 04:40:54 jkridner, the site files are full of stuff like ac_cv_sctp=${ac_cv_sctp=no} Aug 19 04:41:12 someone who is not me might be able to explain why Aug 19 04:41:15 i think he's concerned that its using /usr/bin/m4 not the one in staging Aug 19 04:41:21 rather than the particular syntax of the definition Aug 19 04:41:49 ah Aug 19 04:42:22 * kergoth wasn't sure at first glance either, had to read it twice :) Aug 19 04:43:17 ka6sox-work: that's not mine blog Aug 19 04:47:11 JaMa: did you look at my mail to ml Aug 19 04:52:29 about etk? yes but I was already half sleeping.. Aug 19 04:53:36 iirc I proposed to move all other etk depending stuff to obsolete too.. back then.. but haven't checked if some recipes are in newer versions rewrittern without etk (and only our DEPENDS wrongly pulls etk) Aug 19 04:54:16 khem: http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg03583.html Aug 19 04:58:38 well it would be nice if we can sort this out Aug 19 04:58:48 and I think you are best fit Aug 19 05:00:02 yup, I'll check it, 6 months old warning should be enough to move it to obsolete and remove from image (if there is one including those) Aug 19 05:00:21 I sorted only those apps I had in my images.. Aug 19 05:00:36 grg: ac_cv_sctp=${ac_cv_sctp=no} sets it if not set already Aug 19 05:00:42 and default is no Aug 19 05:00:49 JaMa: bad bad Aug 19 05:01:14 JaMa: I am putting some effort to revive world Aug 19 05:01:19 I mean bitbake world Aug 19 05:01:33 things are too much out of control Aug 19 05:01:48 broken deps wrong deps god knows Aug 19 05:02:17 or even wrong checksums or nonexistent SRC_URIs :/ Aug 19 05:02:19 khem, ok thanks. Is this bourne shell syntax, or something else? Aug 19 05:02:27 JaMa: if you could fix the recipes I mentioned in email you could get some kick ass optimizations in gcc 4.5 in return :) Aug 19 05:03:08 khem: btw I've upgraded to latest SRCREV in gcc-4.5, ok to push it? all rebuilt from scratch ok Aug 19 05:03:21 sure you have my ack Aug 19 05:03:30 btw did you check what got fixed Aug 19 05:05:21 yup read svn log, few fixes, but nothing I would know solving some real world problem we had in OE (ie -ggdb for eglibc seems still needed) Aug 19 05:07:37 ok Aug 19 05:08:01 grg: I dont have other shells but should work on bash yes Aug 19 05:08:04 others dont know Aug 19 05:10:52 and also works in dash I suppose on my ubuntu Aug 19 05:22:12 gm Aug 19 05:23:18 gm Aug 19 05:23:29 eFfeM_work: did you have time for xorg removal review? Aug 19 05:23:49 JaMa: you have my ack for xorg looked fine to me Aug 19 05:24:07 ok, thanks Aug 19 05:24:13 JaMa: btw. can I wait on you to fix the etk stuff too Aug 19 05:24:26 along with you other efl changes Aug 19 05:24:33 yup.. Aug 19 05:24:59 thx Aug 19 05:25:20 I'll check if eve works in new image (I had strange relocation issue with floorf() in webkit-efl also upgraded to latest) Aug 19 05:25:27 2 steps forward with bitbake world Aug 19 05:26:24 floorf should be in libm Aug 19 05:27:37 it is (checked in -dbg version) Aug 19 05:27:59 relocation issue at runtime ? Aug 19 05:28:10 or link time Aug 19 05:29:34 jkridner|work: reading the log: angstrom_version build fine for me, it is also weird you have a problem with do_install Aug 19 05:30:13 runtime, mmt Aug 19 05:30:36 jkridner|work: I whitnk it diidn't get to gedit yet Aug 19 05:30:40 JaMa: somehow -lm did nt get into DT_NEEDED I believe Aug 19 05:31:07 03Martin Jansa  07org.openembedded.dev * rad76b46258 10openembedded.git/recipes/xorg-lib/ (10 files in 2 dirs): Aug 19 05:31:07 pixman: move 0.18.2 to obsolete Aug 19 05:31:08 Khem Raj Aug 19 05:31:08 Signed-off-by: Martin Jansa Aug 19 05:31:09 03Martin Jansa  07org.openembedded.dev * rfa9ee53a58 10openembedded.git/recipes/xorg-data/ (xbitmaps_1.0.1.bb xcursor-themes_1.0.1.bb): Aug 19 05:31:09 xorg-data: move older not pinned versions to obsolete dir Aug 19 05:31:09 Acked-by: Khem Raj Aug 19 05:31:09 Signed-off-by: Martin Jansa Aug 19 05:31:12 03Martin Jansa  07org.openembedded.dev * ref9813dfa3 10openembedded.git/recipes/ (123 files in 6 dirs): Aug 19 05:31:12 xorg-app: move older not pinned versions to obsolete dir Aug 19 05:31:12 Acked-by: Khem Raj Aug 19 05:31:12 Signed-off-by: Martin Jansa Aug 19 05:31:12 03Martin Jansa  07org.openembedded.dev * reb8a21b00f 10openembedded.git/recipes/xorg-doc/ (6 files): Aug 19 05:31:12 xorg-doc: move older not pinned versions to obsolete dir Aug 19 05:31:13 Acked-by: Khem Raj Aug 19 05:31:13 Signed-off-by: Martin Jansa Aug 19 05:31:21 03Martin Jansa  07org.openembedded.dev * re1aac7076a 10openembedded.git/recipes/xorg-lib/ (174 files in 23 dirs): Aug 19 05:31:21 xorg-lib: move older not pinned versions to obsolete dir Aug 19 05:31:21 Acked-by: Khem Raj Aug 19 05:31:21 Signed-off-by: Martin Jansa Aug 19 05:31:36 khem: eve: relocation error: /usr/lib/libwebcore_efl.so.0: symbol floorf, version GLIBCXX_3.4.3 not defined in file libstdc++.so.6 with link time reference Aug 19 05:31:36 JaMa: thanks@ Aug 19 05:31:42 yw Aug 19 05:32:18 JaMa: add LD_PRELOAD=libm.so Aug 19 05:32:23 and see what happens Aug 19 05:32:47 03Martin Jansa  07org.openembedded.dev * r2c2b2d98a7 10openembedded.git/recipes/xorg-util/ (16 files in 5 dirs): Aug 19 05:32:47 xorg-util: move older not pinned versions to obsolete dir Aug 19 05:32:47 Acked-by: Khem Raj Aug 19 05:32:47 Signed-off-by: Martin Jansa Aug 19 05:33:00 LD_PRELOAD=/lib/libm.so.6 may be Aug 19 05:33:15 03Martin Jansa  07org.openembedded.dev * r5250895e04 10openembedded.git/recipes/ (44 files in 12 dirs): Aug 19 05:33:15 xorg-xserver: move older not pinned versions to obsolete dir Aug 19 05:33:15 Acked-by: Khem Raj Aug 19 05:33:15 Signed-off-by: Martin Jansa Aug 19 05:33:15 03Martin Jansa  07org.openembedded.dev * r7ef807aa73 10openembedded.git/recipes/xorg-driver/ (182 files in 3 dirs): Aug 19 05:33:15 xorg-driver: move older not pinned versions to obsolete dir Aug 19 05:33:16 khem: already tried and didn't help Aug 19 05:33:16 Acked-by: Khem Raj Aug 19 05:33:16 Signed-off-by: Martin Jansa Aug 19 05:33:17 03Martin Jansa  07org.openembedded.dev * r91d43ea550 10openembedded.git/recipes/xorg-proto/ (49 files): Aug 19 05:33:17 xorg-proto: move older not pinned versions to obsolete dir Aug 19 05:33:18 Acked-by: Khem Raj Aug 19 05:33:19 Signed-off-by: Martin Jansa Aug 19 05:33:19 03Martin Jansa  07org.openembedded.dev * r423252dcd9 10openembedded.git/recipes/xorg-font/ (43 files in 3 dirs): Aug 19 05:33:20 xorg-font: move older not pinned versions to obsolete dir Aug 19 05:33:20 Acked-by: Khem Raj Aug 19 05:33:21 Signed-off-by: Martin Jansa Aug 19 05:34:13 khem: and when I tried to LD_PRELOAD /lib/.debug/libm-2.12.1.so I got segmentation fault Aug 19 05:35:25 jkridner, seems gedit has an internal dep on docbook-sgml-dtd, please verify if that one is build, if not build manually then build gedit and if it works add it to the dep list of gedit Aug 19 05:40:17 grg: can you push your two patches with ACKs? Aug 19 05:40:37 JaMa, nope. I'm not a committer Aug 19 05:41:02 ah ok.. so I'll Aug 19 05:42:04 * ka6sox-away looks @ the backlog. Aug 19 05:44:04 JaMa: can you put libwebcore_efl.so.0 and libcstdc++-v3.so and libm.so.6 somehwere Aug 19 05:45:05 yup but I have to go to work now.. so maybe you'll be sleeping when I arrive there and upload it somewhere :/ Aug 19 05:45:16 you'll find it in backlog :) Aug 19 05:45:53 if it is in patchwork with acks and someone points me to the right patches I can push them Aug 19 05:46:27 (can't test them or so, so won't ack, totally assuming it is ok by the acks) Aug 19 05:51:57 03David Kozub  07org.openembedded.dev * r2ca2dea3e4 10openembedded.git/recipes/cmake/cmake-native_2.8.2.bb: Aug 19 05:51:57 cmake-native: add recipe for cmake 2.8.2. Aug 19 05:51:57 Signed-off-by: David Kozub Aug 19 05:51:57 Signed-off-by: Martin Jansa Aug 19 05:52:02 03David Kozub  07org.openembedded.dev * r585f3201eb 10openembedded.git/classes/cmake.bbclass: Aug 19 05:52:02 cmake.bbclass: add CMAKE_SYSTEM_PROCESSOR into generated toolchain file Aug 19 05:52:02 Needed for packages that check this value, e.g. recent webkit. Aug 19 05:52:02 Signed-off-by: David Kozub Aug 19 05:52:02 Signed-off-by: Martin Jansa Aug 19 05:52:03 03David Kozub  07org.openembedded.dev * rd0f8c51b16 10openembedded.git/recipes/cmake/ (cmake-2.8.2/dont-run-cross-binaries.patch cmake_2.8.2.bb): (log message trimmed) Aug 19 05:52:05 cmake-2.8.2: Add recipe for non-native cmake 2.8.2. Aug 19 05:52:05 Tested on om-gta02: Aug 19 05:52:05 * builds trivial C and C++ executable Aug 19 05:52:06 * "include(FindPkgConfig)" correctly finds the cmake module Aug 19 05:52:06 * pkg_check_modules correctly invokes pkg-config Aug 19 05:52:07 * ccmake works Aug 19 05:52:07 03Graham Gower  07org.openembedded.dev * r2e6dfcb63d 10openembedded.git/ (2 files in 2 dirs): (log message trimmed) Aug 19 05:52:08 rootfs_ipk.bbclass: Fix BAD_RECOMMENDATIONS. Aug 19 05:52:08 See http://code.google.com/p/opkg/source/detail?r=553 Aug 19 05:52:09 And previous discussions: Aug 19 05:52:10 http://www.mail-archive.com/openembedded-devel@lists.openembedded.org/msg01423.html Aug 19 05:52:57 03Jason Kridner  07org.openembedded.dev * r627570958b 10openembedded.git/recipes/angstrom/angstrom-version.bb: Aug 19 05:52:57 angstrom-version: install not mkdir Aug 19 05:52:57 Thanks Steve Sakoman for finding this. Aug 19 05:52:57 Signed-off-by: Jason Kridner Aug 19 05:55:57 Khem: Should I resubmit with r0 then? Aug 19 05:57:32 xobs: yes Aug 19 05:58:14 03Graham Gower  07org.openembedded.dev * r244ecae83f 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Aug 19 05:58:14 rootfs_ipk.bbclass: Fix BAD_RECOMMENDATIONS - fix conflict resolution Aug 19 05:58:14 Signed-off-by: Graham Gower Aug 19 05:58:14 Signed-off-by: Martin Jansa Aug 19 05:58:27 I dont like blobs as a follow up is it possible to have those tools building Aug 19 05:59:14 khem: I'll need to re-submit the bootloader as well. Aug 19 06:00:40 khem: The source is there, but in order to get it actually working it needs to be run through elftosb2, which is provided from Freescale as closed-source. Not sure what to do about that, since it's x86 Linux only. Aug 19 06:01:16 xobs, please sumbit with patch v2 in the subject line and update patchwork Aug 19 06:01:58 (that was towards the resubmit with r0, no opinion on this bootloader thing) Aug 19 06:02:53 eFfeM_work: I'll do that. I'll also put the correct subject line in this time. Should I do all three patches, or just the two that need their PR fixed? Aug 19 06:03:20 suggest to do all three, that avoids mistakes when applying Aug 19 06:06:44 gn all Aug 19 06:06:51 time to sleep Aug 19 06:08:13 nite khem Aug 19 06:19:22 eFfeM_work: Getting ready to send these out. How do I update patchwork? Aug 19 06:20:12 xobs: http://wiki.openembedded.org/index.php/Patchwork Aug 19 06:20:27 and http://patchwork.openembedded.org/project/openembedded/list/ Aug 19 06:20:41 i think you can modify the patches that you have created yourself (but I am not a wiz) Aug 19 06:22:53 Ah. I'm guessing I'll need to change the state to "Superseded". Aug 19 06:22:59 yes Aug 19 06:23:29 rwhitby: is the asusoe distro still alive ? Aug 19 06:23:37 never was Aug 19 06:24:31 should we remove ? Aug 19 06:25:06 yes Aug 19 06:25:18 can you do so or do you want me to do it ? Aug 19 06:25:37 if you are able to do it easily, please do. I don't have OE cloned anywhere at the moment. Aug 19 06:25:48 i can do easily Aug 19 06:25:57 appreciated, thanks. Aug 19 06:27:23 yw Aug 19 06:27:27 there it goes: Aug 19 06:28:10 03Frans Meulenbroeks  07org.openembedded.dev * r0ee1b49034 10openembedded.git/conf/distro/asusoe.conf: Aug 19 06:28:10 asusoe: distro removed Aug 19 06:28:10 removed after discussion on irc with, and getting permission from Aug 19 06:28:10 rwhitby who created this. Aug 19 06:28:10 Rod stated this distro was never alive Aug 19 06:28:11 Signed-off-by: Frans Meulenbroeks Aug 19 06:28:32 eFfeM_work: thx Aug 19 06:28:43 was easy :-) Aug 19 06:58:27 rwhitby, you're in Adelaide? Aug 19 06:58:36 grg: yes Aug 19 06:58:40 me too Aug 19 06:58:49 didn't realise another OEer was.. Aug 19 06:59:06 I'm not much of an OEer these days. Too busy on webos-internals.org Aug 19 06:59:25 (but webOS is built by Palm on OE, so there is still an overlap) Aug 19 06:59:27 ahr, fellow aussies Aug 19 06:59:35 * thaytan is in Albury Aug 19 07:00:05 tharvey, cool Aug 19 07:00:09 thaytan, cool Aug 19 07:00:18 :) wrong "tha" Aug 19 08:16:24 hi florian_kc Aug 19 08:17:05 good morning Aug 19 08:17:50 florian: do you know if distro/include/maemo-preferred.inc is still of any use? no one seems to use it and the file is not touched since 2006 Aug 19 08:18:04 and seems mostly your work Aug 19 08:18:45 eFfeM_work, are you pruning the whole tree? Aug 19 08:19:14 ka6sox-away: not the whole tree, but I am trying to get rid of some deadwood Aug 19 08:19:44 and the files in distro/include pin versions, barring their removal Aug 19 08:20:03 old versions of recipes Aug 19 08:20:46 okay I was about to pickup and run with something that was in nonworking/ so has that branch gone away? Aug 19 08:21:30 nonworking branch? not sure if we ever had one, we do have a recipes/nonworking dir though Aug 19 08:21:53 sorry..thats what I mean Aug 19 08:21:54 has 54 dirs in it Aug 19 08:22:04 what did you need ? Aug 19 08:22:10 freenx Aug 19 08:22:20 it is still in nonworking Aug 19 08:22:30 k Aug 19 08:22:54 it was last time I pulled but I just wanted to know if it was pruned and I"d move it Aug 19 08:23:04 (locally) Aug 19 08:23:46 i have not pruned stuff in non-working or obsolete except for a few things that had working newer versions in the tree Aug 19 08:25:14 btw anyone an idea on the performance difference between open/putc/close and fopen/fputc/fclose Aug 19 08:25:30 fopen is unbuffered Aug 19 08:25:52 * eFfeM_work wants to be able to quickly toggle a gpio pin Aug 19 09:05:31 03Martin Jansa  07org.openembedded.dev * r85d8c9c981 10openembedded.git/recipes/ (3 files in 2 dirs): Aug 19 09:05:31 exhibit: move to obsolete, depends on etk Aug 19 09:05:31 Signed-off-by: Martin Jansa Aug 19 09:05:31 03Martin Jansa  07org.openembedded.dev * rf8caf3bff7 10openembedded.git/recipes/openmoko-projects/diversity-radar_svn.bb: Aug 19 09:05:31 diversity-radar: upgrade SRCREV and drop python-etk from DEPENDS (no longer needed) Aug 19 09:05:31 Signed-off-by: Martin Jansa Aug 19 09:05:31 03Martin Jansa  07org.openembedded.dev * r394ac63be0 10openembedded.git/recipes/ (3 files in 2 dirs): Aug 19 09:05:31 assassin(-thumbnail): move to obsolete, depends on etk and no commit in last 20 months Aug 19 09:05:31 Signed-off-by: Martin Jansa Aug 19 09:05:31 03Martin Jansa  07org.openembedded.dev * r669a705046 10openembedded.git/recipes/openmoko-projects/diversity-daemon_svn.bb: Aug 19 09:05:31 diversity-daemon: use http for svn checkout, add maxrev param to patch, which no longer applies Aug 19 09:05:31 * https has expired certificate Aug 19 09:05:31 Signed-off-by: Martin Jansa Aug 19 09:05:31 03Martin Jansa  07org.openembedded.dev * r4890bc365f 10openembedded.git/recipes/e17/edje-player_svn.bb: Aug 19 09:05:31 edje-player: move to obsolete, depends on etk and newer version was merged to edje in 47344 Aug 19 09:05:31 * http://trac.enlightenment.org/e/changeset/47344 Aug 19 09:05:31 Signed-off-by: Martin Jansa Aug 19 09:05:32 03Martin Jansa  07org.openembedded.dev * ra04b449503 10openembedded.git/recipes/efl1/eflpp_svn.bb: Aug 19 09:05:32 eflpp: drop etk/ewl dependency, their bindings were removed in 44070 Aug 19 09:05:32 Signed-off-by: Martin Jansa Aug 19 09:05:32 03Martin Jansa  07org.openembedded.dev * reca3d003c9 10openembedded.git/recipes/efl1/ (evolve-native_svn.bb evolve_svn.bb): Aug 19 09:05:32 evolve: move to obsolete, depends on etk and nothing in tree is using this Aug 19 09:05:32 Signed-off-by: Martin Jansa Aug 19 09:05:32 03Martin Jansa  07org.openembedded.dev * ra34dcc7ba6 10openembedded.git/recipes/ (2 files in 2 dirs): Aug 19 09:05:32 om-settings: move to obsolete, depends on python-etk and is no longer maintained Aug 19 09:05:32 Signed-off-by: Martin Jansa Aug 19 09:05:32 03Martin Jansa  07org.openembedded.dev * rb6b32d3978 10openembedded.git/recipes/omoney/omoney_svn.bb: Aug 19 09:05:32 omoney: move to obsolete, depends on python-etk and is no longer maintained Aug 19 09:05:32 Signed-off-by: Martin Jansa Aug 19 09:05:42 03Martin Jansa  07org.openembedded.dev * rd8bdecd350 10openembedded.git/recipes/ (2 files in 2 dirs): Aug 19 09:05:42 om-locations: move to obsolete, depends on etk and no commit in last 20 months Aug 19 09:05:42 Signed-off-by: Martin Jansa Aug 19 09:05:47 03Martin Jansa  07org.openembedded.dev * r12c57719af 10openembedded.git/recipes/comic-reader/comic-reader_svn.bb: Aug 19 09:05:47 comic-reader: move to obsolete, depends on etk Aug 19 09:05:47 Signed-off-by: Martin Jansa Aug 19 09:05:47 03Martin Jansa  07org.openembedded.dev * r3ba5c14678 10openembedded.git/recipes/omview/omview_svn.bb: Aug 19 09:05:49 omview: move to obsolete, depends on epsilon and is no longer maintained Aug 19 09:05:49 Signed-off-by: Martin Jansa Aug 19 09:05:49 03Martin Jansa  07org.openembedded.dev * r0f42891604 10openembedded.git/recipes/e17/edje-editor_svn.bb: Aug 19 09:05:49 edje-editor: move to obsolete, depends on etk Aug 19 09:05:50 Signed-off-by: Martin Jansa Aug 19 09:06:15 03Martin Jansa  07org.openembedded.dev * raa9ed8d596 10openembedded.git/recipes/gpe-icons/gpe-theme-neo_git.bb: Aug 19 09:06:15 gpe-theme-neo: don't rsuggests obsolete etk-theme-neo Aug 19 09:06:15 Signed-off-by: Martin Jansa Aug 19 09:06:15 03Martin Jansa  07org.openembedded.dev * rf52e3db148 10openembedded.git/recipes/openmoko-projects/etk-theme-ninja_svn.bb: Aug 19 09:06:16 etk-theme-ninja: move to obsolete Aug 19 09:06:17 Signed-off-by: Martin Jansa Aug 19 09:56:09 florian: do you know if distro/include/maemo-preferred.inc is still of any use? no one seems to use it and the file is not touched since 2006 Aug 19 09:56:18 or do you know who can say something about it Aug 19 09:57:03 eFfeM_work: no idea Aug 19 09:58:34 florian: it is not referred to in any distro file, you made it jul 1, 20005 and Micked made somee changes in aug 29, 2006 Aug 19 09:58:57 s/Micked/mickeyl/ Aug 19 09:59:25 eFfeM_work: might be possible, but i don't know who might be using it currently Aug 19 09:59:41 florian: grep says no one is (at least not in-tree) Aug 19 09:59:53 we have too many of those orphaned files around Aug 19 10:00:20 but these don't hurt anyone... Aug 19 10:00:20 mickey|office: do you have an opinion on this? Aug 19 10:00:36 Hi eFfeM_work, In which scenarion one will convert FILESDIR to FILESPATHPKG Aug 19 10:01:55 florian: they do hurt as they pin recipes; if they are using legacy staging they should be converted or removed; as the pinnings are to old versions, I am in favour of removing, at least I have no desire to spent time on converting something that is not needed Aug 19 10:02:30 Noor: in lots of cases FILESDIR can just go, FILESPATHPKG is only useful if patches come from other dirs Aug 19 10:02:46 e.g. X_0.bb will look for patches in dirs X-0, X and files Aug 19 10:03:04 but if it also needs patches from say X-1 or Y-0 then FILESPATHPKG need to be used Aug 19 10:03:07 If they are not used they don't hurt even if they use legacy staging. Aug 19 10:03:15 that means if patches are not coming from different dir we can remove the FILEDIR var Aug 19 10:03:22 Noor, yes Aug 19 10:03:28 thanks Aug 19 10:03:35 florian: i would like to get rid of legacy staging altogether Aug 19 10:04:12 Noor, simplest thing is to remove FILESDIR, then do a bake of the recipe (or recipes if it is in an .inc file), if the recipes still build no change is needed Aug 19 10:04:43 fine Aug 19 10:05:19 eFfeM_work: soure would be good but i would be absolutely happy if we would not need it any more. but removin somethign is a different thing - it means that we loose the information how to build some piece of software. Aug 19 10:05:21 florian: and yeah, somehow it disturbs me having lots of old crap around (that is personal, I know), but imho every recipe in our tree should be buildable Aug 19 10:05:52 in fact the information is still in git but not easy to find any more Aug 19 10:06:11 florian: no one seems to be using it, but I am happy to move to the obsolete dir if that is preferred Aug 19 10:06:41 eFfeM_work: apart from very obvious cases i would prefer this Aug 19 10:07:03 but that's my personal opinion Aug 19 10:07:55 The main problem is that we as OE developers have no way to find out what actually gets used and what not. Aug 19 10:08:28 in-tree we do, outside we don't Aug 19 10:08:54 eFfeM_work, I know someone who had a n800 or n810 and that used a maemo compat thing Aug 19 10:09:04 hi btw Aug 19 10:09:21 maybe it's chinhook-compat or similar Aug 19 10:09:26 not sure Aug 19 10:09:45 ah ok no distro files uses it Aug 19 10:09:50 then sorry for the noise Aug 19 10:09:55 a 10 line distro/include file that is not used by a distro imho does not belong in our tree. either I want to see the distro in oe too, or they should maintain their own pinnings (which are in that inc file) Aug 19 10:12:09 and a recipe that is not maintained and blocks progress also is not too desirable Aug 19 10:12:14 imho Aug 19 10:12:54 kergoth: i got a mail from this guy: http://github.com/dettus he sent me this file for tslib: http://www.dettus.net/input-elo.c , I told him to send it to the ML or to you, i hope that's ok. Aug 19 10:14:35 Noor, peeking at FILESDIR, it might be that for native recipes FILESDIR or FILESPATHPKG is needed to go to the non-native patches Aug 19 10:15:16 a simple solution is to substitute FILESDIR by FILESPATHPKG, a better solution is to merge the native and nonnative recipes and get rid of FILESDIR completely Aug 19 10:15:21 I added qt4-embedded to my image, and that made it quite large. Can something be done to include only the parts of Qt that my application actually uses? Aug 19 10:17:22 tasslehoff: there is no autodetection for that. one way to start is to just install your application Aug 19 10:17:33 tasslehoff: and then install plugins and fonts in the image as well Aug 19 10:18:28 zecke, has tinderbox been tested @ all on Lucid? Aug 19 10:21:52 zecke: just try to launch it, and picup the errors as I go, you mean? Aug 19 10:24:02 hm, barebox has a file://defconfig" line in it but no dir with files in it Aug 19 10:26:40 03Frans Meulenbroeks  07org.openembedded.dev * r42715655fb 10openembedded.git/recipes/barebox/barebox_git.bb: (log message trimmed) Aug 19 10:26:40 barebox_git removed superfluous FILESPATHPKG Aug 19 10:26:40 removed Aug 19 10:26:40 FILESPATHPKG =. "barebox-git:" Aug 19 10:26:40 This is the default so it is not needed. Aug 19 10:26:41 Apart from that no dir barebox-git (or any other files dir) exists Aug 19 10:26:41 (which is odd as there is a file:// line) Aug 19 10:33:31 how to safely -force-reinstall eglibc? Aug 19 10:33:51 current opkg just removes it and then fails to install it (because libc.so is missing) Aug 19 10:33:59 anyone an idea what is wrong with: Aug 19 10:34:00 ERROR: too many values to unpack while parsing /home/frans/workspace/openembedded.git/recipes/freesmartphone/fso-gpsd_git.bb Aug 19 10:34:00 NOTE: :too many values to unpack while evaluating: Aug 19 10:34:00 ${@bb.fetch.get_srcrev(d)} Aug 19 10:34:31 then I've tried to unpack .ipk manually and unpack data.tar.gz to / but it resulted to kernel panic (no idea why) Aug 19 10:34:33 and then I get three more "too many values" errors (also for lots of other recipes) Aug 19 10:44:28 ka6sox-away: most likely it has no special lucid certification Aug 19 10:44:57 khem: relocation issue went away with rebuild from scratch and reflash... Aug 19 10:44:59 tasslehoff: yeah, I also do ls qt*plug* and then you will see image formats, decorations, input plugins Aug 19 10:45:43 zecke, great...then I can upgrade melo. Aug 19 10:46:14 ka6sox-away: I can pick-up the pieces Aug 19 10:47:04 zecke, I'll try not to make any pieces. Aug 19 10:47:20 zecke: thanks Aug 19 10:55:32 I can imagine that pulling in the demos take a bit of space :) Aug 19 11:00:58 zecke: my first error is missing libQtGuiE.so.4. kinda fundamental :). do you have a clever way of finding the ipk I need for a given missing library? Aug 19 11:05:30 tasslehoff: then something is wrong. Your application should have linked to the libraries and they should have been detected automatically Aug 19 11:05:49 tasslehoff: in general there is a *filelist* in each feed directory, so you can find the package for the file Aug 19 11:08:04 zecke: I compile a rootfs that has qt4-embedded, and compile my app using an sdk. So my app relies on the libs being in the rootfs it's packaged with. Aug 19 11:10:27 strange, dbus glib says: Aug 19 11:10:28 dbus/dbus-glib_0.86.bb:do_install_virtclass-native_prepend() Aug 19 11:10:40 postfix recipe says: Aug 19 11:10:41 postfix/postfix.inc:do_install_prepend_virtclass-native() { Aug 19 11:10:49 note the location of the prepend Aug 19 11:11:25 what is the right solution ? Aug 19 11:12:28 as a hack to build gedit: sudo aptitude install gnome-doc-utils -y (I'm getting quite frustrated with the lack of isolation in OE recipes) Aug 19 11:13:09 yeah recipes are not too well contained, pstaging made this more explicit Aug 19 11:13:25 tasslehoff: ah okay, yeah the best way then would be to either use the filelist or to build a dummy app linking to Qt libraries in OE and then take the dependencies Aug 19 11:13:37 but better have gnome-doc-utils in gedit ad dependency Aug 19 11:15:41 eFfeM_work: the path is hard-coded to the host directory. Aug 19 11:15:49 yuk Aug 19 11:15:55 eFfeM_work: you'd need a patch to fix that as well. Aug 19 11:16:08 yes i understand Aug 19 11:16:18 * eFfeM_work always stayed away from gnome stuff Aug 19 11:16:33 I'm just getting exhausted and want to see the end before going back and filling in the potholes. Aug 19 11:16:45 * jkridner|work guesses that is how things got this way. Aug 19 11:17:48 jkridner|work: can't you assign koen to doing the image? he is very good in it! Aug 19 11:18:03 he's busy with exams. Aug 19 11:18:31 ah ok, the august reruns Aug 19 11:18:39 he has Narcissus, but it hasn't been completing for me and I'm concerned that no one would be able to reproduce it from the sources. Aug 19 11:19:09 * jkridner|work checks my latest Narcissus attempt. Aug 19 11:19:25 * jkridner|work isn't good at using 3rd person when using /me Aug 19 11:19:27 i understand (and appreciate your dedication to this, that's why I also decided to kick off the demo -image build earlier this week) Aug 19 11:19:55 no one likes to get code they can't reproduce. Aug 19 11:20:24 well, no one here anyway. Aug 19 11:20:24 nope, that's why you might have seen some rants from me on quality and old recipes etc. Aug 19 11:20:40 mix-n-match does not work Aug 19 11:21:16 If I understood how to use the host-side opkg-cl stuff, I think I would be doing a bit more with it. Aug 19 11:21:32 * eFfeM_work never really peeked into it Aug 19 11:28:40 hmmm.... no recipe for dos2unix. Aug 19 11:29:04 jkridner|work: tofrodos Aug 19 11:31:18 I guess update-alternatives wouldn't work native. Aug 19 11:31:56 is dos2unix a low enough hanging item that we assume it to be on the host? Aug 19 11:33:09 eFfeM_work, should I assume that something that is in nonworking and hasn Aug 19 11:33:17 it is not on my host (ubuntu 10.04) Aug 19 11:33:26 't been touched since 15apr is pretty much not being worked? Aug 19 11:33:49 eFfeM_work: it is used in ti-msp430-chronos Aug 19 11:33:54 it = dos2unix Aug 19 11:35:33 * jkridner|work changes the recipe to use 'fromdos'. Aug 19 11:35:42 seems Ubuntu doesn't even have dos2unix anymore, as you say. Aug 19 11:40:22 ka6sox-away: what was not working since april 15 (personally I'd say it is too early to call it non-working, especially if it is not raised) Aug 19 11:40:33 and if it is something from me, let me know and I'll peek into it Aug 19 11:44:00 its not yours... Aug 19 11:44:47 last "status" I saw was 13Apr and last commit was 15Apr Aug 19 11:45:16 ka6sox-away: generally if it is not working, I'd say the way to contact is the person who did the last commit (not being a generic one like e.g. JaMa for the adding of the checksums) Aug 19 11:46:23 ka6sox-away: which recipe are you trying to resuscitate? Aug 19 11:46:26 okay it looks like it was added in march then JaMa added the checksums in April. but the bb looks awfully small Aug 19 11:46:31 freenx Aug 19 11:46:37 (and its siblings) Aug 19 11:47:44 ka6sox-away: the jama added checksums on april 12, the previous was the packages->recipes rename which as on mar 17, 2009 Aug 19 11:48:31 okay...I saw a page that talked about needing a "manual test" Aug 19 11:49:34 http://permalink.gmane.org/gmane.comp.handhelds.openembedded/31962 Aug 19 11:52:15 hmmm... how do I put tools in the path for OE to use? Aug 19 11:52:39 hi eFfeM_work, when merging native and non-native recipes, if there is a different do_install in non-native version than that in the native version, should i still merge them and use do_install_virtclass-native or leave native and non-native as seperate recipes? Aug 19 11:52:46 I was able to build tofrodos-native fine, but they don't seem to be in the path for ti-msp430-chronos install to use. Aug 19 11:53:06 jkridner|work: odd Aug 19 11:53:44 well, it is a new recipe, since the target recipe wouldn't work -native. Aug 19 11:54:53 it should show up in tmp-angstrom_2008_1/sysroots/x86_64-linux/usr/bin/, right? Aug 19 11:54:58 fahad_usman: if the install is different first see if doing the standard do_install for a target installs the same things (or more) than the native recipe installs, if that is the case use the std do_install; it is ok if the target do_install installs a few more files for target Aug 19 11:55:09 jkridner|work: have you added inherit native ? Aug 19 11:55:14 no. Aug 19 11:55:39 if you created a separate native recipe then you should do that Aug 19 11:55:39 I thought the -native in the recipe did that for me. Aug 19 11:55:44 k. Aug 19 11:55:44 no Aug 19 11:56:31 fahad_usman: try to use the std do_install, if there are issues with it let us know and someone can look at it Aug 19 11:56:59 eFfeM_work: but what if the non-native do_install installs something that's neither installed by the standard do_install, nor the native do_install Aug 19 11:57:00 and if possible after merging a native recipe try to build recipes that depend on that native recipe Aug 19 11:57:42 fahad_usman: that would be very strange, because the standard do_install function is used but it is just installed to a different dir Aug 19 12:02:34 hmm.... do_populate_sysroot failed. :( Aug 19 12:02:43 eFfeM_work, when the packages changed names did the history before that just disappear? Aug 19 12:03:11 ka6sox-away: partly, i think cgit still has them Aug 19 12:04:05 * ka6sox-away remembers the Monotone to Git transition. Aug 19 12:05:07 ka6sox-away: Use git log --follow Aug 19 12:05:20 broonie, thanks Aug 19 12:06:22 hm it doesn't i'd swear I saw older logs Aug 19 12:08:21 I froze the monotone dbs along with an instance *just* to be able to read them if necessary. Aug 19 12:08:53 hm, don't see them, the info is proably still in git somewere, git blame has info on older dates Aug 19 12:10:44 ka6sox-away: you wanted this: 2005-09-07packages/nonworking/freenx/: add freenx suite to nonworkingKoen Kooi Aug 19 12:11:18 yes, it seems to have been sitting there for the last 5 yrs. Aug 19 12:11:20 you can browse this for pre-renaming: http://cgit.openembedded.org/cgit.cgi/openembedded/tree/?id=fa6cd5a3b993f16c27de4ff82b42684516d433ba Aug 19 12:13:21 okay they are up to 3.4.0 now so I'll see if that can build. Aug 19 12:13:34 gl Aug 19 12:14:11 if it was easy somebody might have done it already. Aug 19 12:29:09 03Martin Jansa  07org.openembedded.dev * ra1f89a780e 10openembedded.git/recipes/xalan-j/xalan-j_2.7.1.bb: Aug 19 12:29:09 xalan-j: fix non-native DEPENDS Aug 19 12:29:09 Signed-off-by: Martin Jansa Aug 19 12:36:31 pb_: RP__: mickey|office: ping? Aug 19 12:43:19 I get this error on git-native: /usr/lib/libgnutls.so.26: undefined reference to `gcry_cipher_setkey@GCRYPT_1.2' Aug 19 12:44:08 tasslehoff: oooh... Aug 19 12:44:38 tasslehoff: what happens if you remove the libgcrpt from the native staging? Aug 19 12:45:34 zecke: translate that to oe-rookie language please :). remove a file from the staging-directory? Aug 19 12:46:15 tasslehoff: did you try to rebuild git-native first? Aug 19 12:46:17 tasslehoff: I have added libgcrypt-native to the gtk-native build Aug 19 12:47:10 I'll try rebuilding now Aug 19 12:47:18 tasslehoff: and now libgnutls.so.26 looks like it is using OEs libgcrypt... or such. :} Aug 19 12:50:21 zecke: JaMa|wrk: clean + compile seems to do the trick Aug 19 13:08:10 zecke: pong Aug 19 13:10:32 RP__: I think we need to discuss something. :) Aug 19 13:19:43 zecke: yes Aug 19 13:23:43 there is a PR in inc file named INC_PR and in recipe file PR is defined as PR = ${INC_PR}.0 Aug 19 13:24:08 I am only modifying inc file so i need to pump PR on inc file or the recipe file Aug 19 13:24:16 Noor, yes, if you modify the .inc file bump INC_PR, if you modify the bb file modify the digit in PR Aug 19 13:24:27 OK Aug 19 13:24:47 and where should I right NATIVE_INSTALL_WORKS in recipe or in inc Aug 19 13:24:55 inc is also native Aug 19 13:25:11 typically the inc contains common stuff for all versions Aug 19 13:25:19 so if it applies to all versions put it in the inc Aug 19 13:25:30 OK Aug 19 13:25:32 and if there is only one recipe and an inc, also put it in the inc Aug 19 13:29:04 mickey|office: ping? Aug 19 13:33:04 zecke, he seem very busy, he responded me this morning,but I wasn't in front of the computer,after I pinged him many times and I didn't get a response yet Aug 19 13:43:38 zecke: pong, what'sup? Aug 19 13:45:05 ok, I have a hack in place for tofrodos that I'll submit as soon as the build finishes, but gnome-games is my next challeng. Aug 19 13:45:18 mickey|office: please check the mail. :) Aug 19 13:45:52 no mail yet Aug 19 13:47:37 argh. tofrodos still didn't patch cleanly? Aug 19 13:48:35 oh, I have the patch level wrong. :( Aug 19 13:48:55 ah, there Aug 19 13:51:25 where's our tsc room again? Aug 19 13:56:05 *sigh* why wouldn't ${D} be set on a native build? Aug 19 14:00:39 mickey|office: ##tsc Aug 19 14:06:57 ah Aug 19 14:07:24 hmm, can't connect from here :D Aug 19 14:07:29 throws me out immediately Aug 19 14:07:40 mickey|office: need to register for your nick Aug 19 16:05:07 hi, any progress with build for dream ? it displays framebuffer snapshot and restarts when booting up. GNUtoo ? Aug 19 16:10:25 here's one that fails I'd expect more people to be building: qt4-x11-free-gles Aug 19 16:15:49 anyone in Cambrudge want a quick job Aug 19 16:15:50 http://old.nabble.com/GumStix-help-needed-in-Cambridge-today-tt29479535.html#a29479535 Aug 19 16:34:11 Crofton: heh Aug 19 16:54:03 bitbake qt4-x11-free-gles (4.6.3) failed with a missing library for EGL, but 'bitbake -k qt4-x11-free-gles; bitbake -c clean qt4-x11-free-gles; bitbake qt4-x11-free-gles' runs a *whole* lot further (and still fails). Aug 19 16:54:40 oh, wait, nevermind. it runs a lot longer, but the failure is in the same place. :( Aug 19 16:55:06 does nothing satisfy virtual/egl? Aug 19 17:13:00 there's always git grep to find out :) Aug 19 17:40:58 kergoth_: around ? Aug 19 17:41:30 kergoth_: I want to chop a given path at a keyword how to do it Aug 19 17:42:15 suppose a/b/src/c/foo.c I want to get only c/foo.c which means all before that gets chopped off Aug 19 17:42:45 * khem looks at os.path Aug 19 17:46:27 kergoth_: I want to chop a given path at a keyword how to do it suppose a/b/src/c/foo.c I want to get only c/foo.c which means all before that gets chopped off Aug 19 17:47:19 you can do a slice if you have the length of the prefix. you can split by "/" and slice that to remove X components, or do whatever you want with that list, or.. Aug 19 17:47:49 if you want to find the 'src' and cut off there, split by "/" and use .index(), or you can just do a search of the string Aug 19 17:48:17 there are two dirs in different trees I want to compare them say /a/src/b.c /b/src/b.c Aug 19 17:48:31 its copy of same src Aug 19 17:48:40 but may be modified Aug 19 17:49:09 pardon me I want to check if file exists Aug 19 17:49:20 in the mirror tree or not Aug 19 17:49:31 the common point is src Aug 19 17:49:37 below src the dir structure is same Aug 19 17:50:25 could always just do a "/a/src/b.c".find("/src/b.c") and cut at the index it gives you Aug 19 17:50:27 lots of ways Aug 19 17:51:43 meh Aug 19 18:24:22 re Aug 19 19:45:53 anyone here have a freerunner? Aug 19 19:46:15 i'm wondering what they labeled the SD slot... "System" or "HardDisk" or something like that Aug 19 20:22:32 03Frans Meulenbroeks  07org.openembedded.dev * rc12662b614 10openembedded.git/recipes/boost/boost.inc: Aug 19 20:22:33 Revert "boost.inc: added dependency on cmake-native" Aug 19 20:22:33 This reverts commit 1409e393ee38fa5dbfc2c99eb5fec4da07183bc9. Aug 19 20:22:33 inherit cmake already adds a depends on cmake-native Aug 19 20:22:33 so as discussed on the OE mailing list reverting the change Aug 19 20:22:33 next patch will bump PR to avoid that it is lowered by the revert Aug 19 20:22:33 Signed-off-by: Frans Meulenbroeks Aug 19 20:22:36 03Frans Meulenbroeks  07org.openembedded.dev * raa6d8dab70 10openembedded.git/recipes/libgpod/usbmuxd_1.0.3.bb: Aug 19 20:22:36 Revert "usbmuxd: added dependency on cmake-native" Aug 19 20:22:36 This reverts commit 0f9da77ecee5318dfae05ef2d00139036447eea7. Aug 19 20:22:36 inherit cmake already adds a depends on cmake-native Aug 19 20:22:36 so as discussed on the OE mailing list reverting the change Aug 19 20:22:37 next patch will bump PR to avoid that it is lowered by the revert Aug 19 20:22:37 Signed-off-by: Frans Meulenbroeks Aug 19 20:22:38 03Frans Meulenbroeks  07org.openembedded.dev * r423fb49fa0 10openembedded.git/recipes/libgpod/usbmuxd_1.0.3.bb: Aug 19 20:22:41 hm: Aug 19 20:22:42 * pkg_init_from_file: Malformed package file /home/frans/oe/pstage//angstromglibc/staging-qt4-x11-free-armv7a-angstrom-linux-gnueabi_4.6.3-r18.2_armv7a-linux-gnueabi.ipk. Aug 19 20:24:05 hey eFfeM Aug 19 20:24:12 hi ant Aug 19 20:24:21 how are you ? Aug 19 20:24:47 eh..Im surviving to 'mother+father in law inspection' Aug 19 20:24:56 :-) Aug 19 20:24:59 at my age... Aug 19 20:25:43 be polite..s/inspection/visit/ (eyes on my shoulder :) Aug 19 20:25:50 i'm nearing the age that I am the one needing to do the inspection :-) Aug 19 20:26:00 eek Aug 19 20:26:23 ant_: heh my inlaws goes well with me Aug 19 20:26:26 I started slowly, lot of training before baby ;) Aug 19 20:26:37 father at 39 Aug 19 20:26:50 Aug 19 20:26:55 my mom inspects and measures my cleanliness and getting up early and eat healthy kind Aug 19 20:27:23 khem :-) Aug 19 20:27:35 moms natural behaviour Aug 19 20:27:55 in laws are always impressed with my personality :) Aug 19 20:28:01 so I am like a king Aug 19 20:28:03 eh Aug 19 20:28:14 but I am like a mouse when mom arrives Aug 19 20:28:20 I have this effect only with grandma's :/ Aug 19 20:28:42 heh my grandma is coolest of all she doesnt give a damn about what you do Aug 19 20:32:59 btw khem, is this needed ? Aug 19 20:33:00 gcc-4.4.1.inc:FILESPATHPKG .= ":gcc-$PV" Aug 19 20:33:12 I see it only for 4.4 and 4.1.1 Aug 19 20:33:23 sry 4.1.0 Aug 19 20:33:49 khem, back on OE topic, I'd have to rename e.g klibc-utils-static. What do you suggest? kexec-tools-static-klibc? plain kexec-tools-klibc? Aug 19 20:34:04 err.. you see Aug 19 20:34:25 both kexec-tools and klibc-utils Aug 19 20:34:28 and more... Aug 19 20:36:42 eFfeM: I think no Aug 19 20:37:25 ant_: kexec-tools-klibc Aug 19 20:37:33 ant_: and another recipe for klcc Aug 19 20:37:42 and klibc-utils? Aug 19 20:37:51 we should not try to package klcc with klibc Aug 19 20:38:00 because klibs is target/machine recipe Aug 19 20:38:05 and klcc is a cross tool Aug 19 20:38:13 well its a perl script Aug 19 20:38:18 yes Aug 19 20:38:25 but meant to be installed in cross tools Aug 19 20:38:29 aside from current conversation: gnome-power-manager requires libtool. Aug 19 20:38:44 but I dislike the need of a patch when it could build w/out before Aug 19 20:39:06 something is missing/differenty with new staging Aug 19 20:39:16 ant_: the new staging is more accurate Aug 19 20:39:22 well, I think it refuses any cross stuff Aug 19 20:39:42 ant_: we were doing it wrong when we fudged installing klcc into cross dir when building a target recipe Aug 19 20:39:45 thats uncalled Aug 19 20:39:52 yes thats correct Aug 19 20:39:55 ok, original sin then Aug 19 20:40:15 should have been two recipes for crosscompilation Aug 19 20:40:38 jkridner, tried to build beagleboard-demo-image again, I'm now stuck at gnome-games and iputils; the latter needs sgmlspl but sgmlspl-native does not seem to stage that Aug 19 20:40:42 still it was somehow working before .... Aug 19 20:40:56 yeah, gnome-games is a problem for me too. Aug 19 20:41:22 it complains about scrollkeeper Aug 19 20:41:33 khem, old do_stage could stage sysroots for both host and target Aug 19 20:41:39 probably a bug... Aug 19 20:41:53 jkridner|work1: do you have sgmlspl in sysroots ? Aug 19 20:42:18 khem, btw, I think some gcc- recipes are doing the same, I mean staging for two archs Aug 19 20:42:29 initial iirc Aug 19 20:43:05 yes Aug 19 20:43:30 intermediate too Aug 19 20:44:59 hm.. target sysroot is empty.. just empty dirs Aug 19 20:45:08 doh Aug 19 20:46:22 ok, staging-gcc-cross contains ssysroots for both Aug 19 20:46:22 ant_: yes gcc is mess Aug 19 20:46:42 staging-gcc-cross-armv5te-angstrom-linux-gnueab~.3.3-r16.1_i686-linux-armv5te-linux-gnueabi.ipk│ 62291K│Aug 18 21:01 Aug 19 20:46:47 ideally it should be divided into cross and target recipes Aug 19 20:46:53 ah Aug 19 20:46:56 where all crt and runtime stuff is packaged in libc Aug 19 20:47:06 03Noor Ahsan  07org.openembedded.dev * r3112262e6a 10openembedded.git/recipes/sdcc/sdcc-native_2.8.0.bb: Aug 19 20:47:07 sdcc-native: convert to new style staging, remove 'do_stage()' Aug 19 20:47:07 * Remove do_stage() function from the recipe. do_install function of autotools.bbclass replaces the do-stage function of the recipe. Aug 19 20:47:07 * Bump PR to "r2" Aug 19 20:47:07 Signed-off-by: Noor Ahsan Aug 19 20:47:13 ah, ok, I see..like klibc + klcc Aug 19 20:47:17 03Noor Ahsan  07org.openembedded.dev * r9c6ce0950e 10openembedded.git/recipes/sdcc/sdcc-native_2.8.0.bb: Aug 19 20:47:17 sdcc-native: Add bison-native and flex-native to DEPENDS Aug 19 20:47:17 * configure script was giving error "Cannot find required program bison." and "Cannot find required program flex.". So added bison-native and flex-native to DEPENDS list. Aug 19 20:47:17 * Bump PR to "r1". Aug 19 20:47:17 Signed-off-by: Noor Ahsan Aug 19 20:47:18 yes Aug 19 20:47:30 eFfeM: find . | grep sgmlspl is still running... Aug 19 20:48:55 jkridner try find tmp/sysroots | grep sgmlspl Aug 19 20:49:40 I did it under sysrootfs... not there. Aug 19 20:49:53 strainge iputils says it needs it Aug 19 20:50:06 03Noor Ahsan  07org.openembedded.dev * rbd143786bc 10openembedded.git/recipes/rpm2cpio/rpm2cpio-native_1.2_2.bb: Aug 19 20:50:06 rpm2cpio-native: convert to new style staging, remove 'do_stage()' Aug 19 20:50:06 * Replace do_stage() with do_install from the recipe and replace all occurrences of ${STAGING_BINDIR} with ${D}{bindir}. Aug 19 20:50:06 * Add NATIVE_INSTALL_WORKS = "1" Aug 19 20:50:06 * Bump PR to "r1" Aug 19 20:50:06 Signed-off-by: Noor Ahsan Aug 19 20:50:07 I've started building -k to try to get a more complete list faster. didn't understand what that was before a couple hours ago. Aug 19 20:50:08 03Noor Ahsan  07org.openembedded.dev * r8697a82eae 10openembedded.git/recipes/rpm2cpio/rpm2cpio-native_1.2_2.bb: Aug 19 20:50:08 rpm2cpio-native: Run oe-stylize.py Aug 19 20:50:08 * Run oe-stylize.py script on recipe and modified the recipe according to the output of oe-stylize.py. Aug 19 20:50:08 Signed-off-by: Noor Ahsan Aug 19 20:53:18 jkridner|work1: I also got that far with -k Aug 19 21:02:55 03Frans Meulenbroeks  07org.openembedded.dev * rfa89522c68 10openembedded.git/recipes/gawk/ (gawk-native_3.1.4.bb gawk_3.1.4.bb): Aug 19 21:02:55 gawk: removed legacy staging, converted to BBCLASSEXTEND = "native" Aug 19 21:02:55 triggered by the work of Fahad Usman Aug 19 21:02:55 tested by building gnome/gvfs which depends on gawk-native Aug 19 21:02:55 Signed-off-by: Frans Meulenbroeks Aug 19 21:13:01 03Martin Jansa  07org.openembedded.dev * rcede256fbd 10openembedded.git/conf/distro/include/shr-autorev.inc: Aug 19 21:13:01 shr: drop AUTOREV for ventura and emtooth Aug 19 21:13:01 Signed-off-by: Martin Jansa Aug 19 21:15:52 any way to get tinderbox to show all the pages on one page? Aug 19 21:18:33 03Noor Ahsan  07org.openembedded.dev * r98319485f9 10openembedded.git/recipes/sgmlspl/sgmlspl-native_1.03ii.bb: Aug 19 21:18:33 sgmlspl-native: convert to new style staging, remove do_stage() Aug 19 21:18:33 * Remove do_stage() from the recipe. The cpan_do_install () replaces the do_stage () of the recipe. Aug 19 21:18:33 * Bump PR to "r1" Aug 19 21:18:33 Signed-off-by: Noor Ahsan Aug 19 21:18:36 03Frans Meulenbroeks  07org.openembedded.dev * r4511027a40 10openembedded.git/recipes/sgmlspl/sgmlspl-native_1.03ii.bb: Aug 19 21:18:36 sgmlspl-native: removed PACKAGES Aug 19 21:18:36 the regular package with the sgmlspl binary was not generated Aug 19 21:18:36 so the binary was not staged. This fixes that Aug 19 21:18:36 Signed-off-by: Frans Meulenbroeks Aug 19 21:18:36 03Frans Meulenbroeks  07org.openembedded.dev * r4fc7d465d6 10openembedded.git/recipes/iputils/iputils_20071127.bb: Aug 19 21:18:37 iputils: added dependency on sgmlspl-native Aug 19 21:18:37 Signed-off-by: Frans Meulenbroeks Aug 19 21:19:02 jkridner|work1: with these two patches iputils builds for me Aug 19 21:19:07 * eFfeM is calling it a day Aug 19 21:19:09 have fun! Aug 19 21:19:20 thanks for the patch! Aug 19 21:19:27 that's one less blocker! Aug 19 21:19:56 yw Aug 19 21:28:44 03Andrea Adami  07org.openembedded.dev * r46e7caf119 10openembedded.git/recipes/zaurus-utils/nandlogical-klibc-static_1.0.0.bb: Aug 19 21:28:44 nandlogical-klibc-static: rename recipe to nandlogical-klibc Aug 19 21:28:44 * avoid conflicts with -static suffix Aug 19 21:28:45 03Andrea Adami  07org.openembedded.dev * re531bb9f91 10openembedded.git/recipes/kexecboot/ (kexecboot.inc kexecboot_git.bb): kexecboot: follow kexec-tools-klibc renaming. Bump PR. Aug 19 21:28:47 03Andrea Adami  07org.openembedded.dev * ra3a8cadc1f 10openembedded.git/recipes/kexec-tools/ (7 files in 2 dirs): Aug 19 21:28:47 kexec-tools-klibc-static: rename recipe to kexec-tools-klibc Aug 19 21:28:47 * avoid conflicts with -static suffix Aug 19 21:28:47 * update package names Aug 19 21:28:47 * bump PR Aug 19 21:29:26 03Martin Jansa  07org.openembedded.dev * rdfb57804d1 10openembedded.git/conf/distro/include/sane-toolchain-eglibc.inc: eglibc: fix typo in FULL_OPTIMIZATION override Aug 19 21:39:58 03Andrea Adami  07org.openembedded.dev * r8785b4bb6f 10openembedded.git/recipes/klibc/klibc-utils-static_1.5.19.bb: Aug 19 21:39:58 klibc-utils-static: rename recipe to klibc-static-utils Aug 19 21:39:58 * avoid conflicts with -static suffix Aug 19 21:39:58 * rename packages Aug 19 21:39:58 * bump INC.PR Aug 19 21:41:03 03Tom Rini  07org.openembedded.dev * rc1363b8821 10openembedded.git/recipes/base-files/ (base-files/filesystems base-files_3.0.14.bb): Aug 19 21:41:03 base-files: Prefer ext4 > ext3 > ext2 Aug 19 21:41:03 Have /etc/filesystems try and mount ext4 then ext3 then ext2 Aug 19 21:41:03 Acked-by: Frans Meulenbroeks Aug 19 21:41:03 Signed-off-by: Tom Rini Aug 19 22:48:33 jkridner: ping Aug 19 22:50:23 03Jason Kridner  07org.openembedded.dev * r3e6ef3533a 10openembedded.git/recipes/images/beagleboard-test-image.bb: Aug 19 22:50:23 beagleboard-test-image: add g_ether kernel module Aug 19 22:50:23 Signed-off-by: Jason Kridner Aug 19 22:50:23 Signed-off-by: Denys Dmytriyenko Aug 19 22:50:24 03Jason Kridner  07org.openembedded.dev * r3987cc1b63 10openembedded.git/recipes/esc/esc-gst_6.bb: Aug 19 22:50:24 esc-gst: added revision 6 Aug 19 22:50:25 Signed-off-by: Jason Kridner Aug 19 22:50:25 Signed-off-by: Denys Dmytriyenko Aug 19 23:45:46 03Jason Kridner  07org.openembedded.dev * r3095531ac5 10openembedded.git/recipes/esc/esc-media_5.bb: Aug 19 23:45:46 esc-media: updated revision Aug 19 23:45:46 Added startup.wav (missed it the first time). Corrected the MD5SUM. Aug 19 23:45:46 Signed-off-by: Jason Kridner Aug 19 23:45:46 Signed-off-by: Denys Dmytriyenko Aug 19 23:48:10 03Jason Kridner  07org.openembedded.dev * rc379aed93a 10openembedded.git/recipes/libidl/libidl-native_0.8.13.bb: (log message trimmed) Aug 19 23:48:10 libidl-native: add dependency on bison-native Aug 19 23:48:10 bison-native now depends on flex-native, so we should be good there. Aug 19 23:48:10 The error that would show up without bison: Aug 19 23:48:10 checking for bison... no Aug 19 23:48:10 checking for byacc... no Aug 19 23:48:10 checking for yacc... no Aug 19 23:49:54 03Jason Kridner  07org.openembedded.dev * rf7635a8467 10openembedded.git/conf/machine/beagleboard.conf: Aug 19 23:49:54 beagleboard.conf: use qt4-x11-free-gles Aug 19 23:49:54 Signed-off-by: Jason Kridner Aug 19 23:49:54 Signed-off-by: Denys Dmytriyenko Aug 19 23:50:24 thanks denix Aug 19 23:51:25 03Jason Kridner  07org.openembedded.dev * rc323496195 10openembedded.git/recipes/base-files/ (base-files/beagleboard/profile base-files_3.0.14.bb): (log message trimmed) Aug 19 23:51:25 base-files: beagleboard: profile: use nano Aug 19 23:51:25 If nano is available and EDITOR isn't otherwise set, then use it. Aug 19 23:51:25 Advanced users who would prefer vi will know how to set EDITOR anyway. Aug 19 23:51:25 I can clearly see myself being frustrated typing 'export EDITOR=vi' Aug 19 23:51:26 every time I start a BeagleBoard test ramdisk image, but better me have Aug 19 23:51:27 that than to try to explain to someone fresh how to use vi. Aug 19 23:53:51 jkridner: you are welcome. I think that's it - the rest are not very clear to me or need more feedback :) Aug 19 23:55:41 right... I think going to version numbers for the task-beagleboard-demo and beagleboard-demo-image is the right thing, but I'll need to read some examples to see how to do it. Aug 19 23:57:06 did I not submit the patch to ti-msp430-chronos, denix? perhaps I just didn't summarize it? Aug 19 23:58:07 denix: if you haven't left yet, can you look at patches 2716 and 2717 as well? Aug 19 23:59:52 why tofrodos? seems very ugly... Aug 20 00:01:12 well, I thought the idea was to isolate the host as much as possible... Aug 20 00:01:19 I don't see dos2unix in OE.... Aug 20 00:01:59 tofrodos was right there.... though it took me about 2 hours to get it to build for native. Aug 20 00:02:43 * jkridner isn't sure why building libtool-native wasn't good enough to get gnome-power-manager to find libtool in its configure step. Aug 20 00:03:19 * jkridner expects 'sudo aptitude install libtool -y' to work, but not be as nice a solution for OE. Aug 20 00:05:40 jkridner, how funky is your build box? Aug 20 00:05:47 and/or local.conf + ASSUME_PROVIDED ? Aug 20 00:05:49 Ubuntu 10.04 Aug 20 00:06:25 I used Koen's oebb.sh to setup local.conf and haven't added any of my own ASSUME_PROVIDEDs. Aug 20 00:06:46 * Tartarus googles oebb.sh Aug 20 00:07:48 interesting Aug 20 00:08:02 can we stop tossing this oebb around? :) Aug 20 00:08:06 * Tartarus recalls when the openmoko makefile was the worst thing ever, but assume koen has made amends for that :) Aug 20 00:08:31 * Tartarus has of course had a script to make local.conf for years, so sees this as a good thing Aug 20 00:08:43 Tartarus: ironical, isn't it? :) Aug 20 00:09:26 everyone has a setup script... I tried to keep mine very minimalistic Aug 20 00:09:57 having building part in a separate scripts :) Aug 20 00:10:29 yeah, mine just does tmpdir (not called tmp) + a local.conf in there + a sh and *csh env script Aug 20 00:11:01 So, back on topic.. Aug 20 00:11:18 jkridner, haven't tried u10.04 myself yet, but this is all crazy sounding Aug 20 00:11:26 Lemme go find one thing oebb.sh doesn't do Aug 20 00:11:31 denix: the more minimalistic the better, but having something similar helps too. Aug 20 00:11:52 jkridner: similar to what? Aug 20 00:12:09 jkridner, try installing make gcc g++ patch help2man diffstat texinfo texi2html cvs subversion libexpat-dev gawk libbonobo2-common libncurses5-dev Aug 20 00:12:17 and ubuntu-minimal ubuntu-standard too, if you're up for it Aug 20 00:12:21 Tartarus, I already prodded koen over that :) Aug 20 00:12:30 denix: I find that people manage to setup OE in very odd ways that make it not work. Aug 20 00:12:30 Crofton, k :) Aug 20 00:13:39 jkridner: right, but does it mean koen's way is the right one? :))) Aug 20 00:14:15 * jkridner compares that list to http://gitorious.org/beagleboard-validation/scripts/blobs/master/ec2build.sh#line198 Aug 20 00:14:32 * Tartarus does too Aug 20 00:14:36 anyway, oebb was conceived by _don_ and later adapted by koen Aug 20 00:14:46 ah yes chrpath :( Aug 20 00:14:59 * Tartarus hasn't updated the script yet, but it's on his list Aug 20 00:14:59 Tartarus: speaking of which! Aug 20 00:15:27 I think some of that is covered by ubuntu-minimal + ubuntu-standard Aug 20 00:15:38 we go w/ those packages as they help to ensure a common baseline Aug 20 00:15:43 assuming the customer says yes, install that :) Aug 20 00:16:04 Tartarus, are you still fiddling with Canadian Cross Aug 20 00:16:29 Crofton, personally no, but I know someone who is Aug 20 00:16:34 Tartarus: if chrpath is not in one of the system paths, but still in PATH, then relocatable.bbclass throws a weird exception about not being able to find the binary. had to set CHRPATH_BIN with absolute path Aug 20 00:16:35 She's -> <- close to posting on the ML Aug 20 00:16:46 kergoth and I have been helping to groom the patch set a bit more Aug 20 00:17:01 it's a bit funky since the nativesdk / crosssdk switch made it a little harder Aug 20 00:17:12 cool Aug 20 00:17:24 denix, strange... Aug 20 00:17:34 denix, I'd still rather link w/ $ORIGIN and chrpath the hard ones Aug 20 00:17:42 But it's more complex than what we do today for little gain Aug 20 00:18:27 (There's 2 diff escape sequences, one for makefile by itself builds, one for autotools, and gcc/binutils/gdb still need a mangling or a very crazy 3rd escape sequence) Aug 20 00:20:47 * kergoth wants to find whatever moron chose $ORIGIN as the name Aug 20 00:24:06 git blame ? Aug 20 00:25:32 no, i mean in binutils Aug 20 00:25:33 heh Aug 20 00:25:38 well, are they on git now? Aug 20 00:25:39 heh Aug 20 00:25:44 * kergoth is behind in toolchain bits Aug 20 00:25:49 Well, it's not a GNU thing Aug 20 00:26:02 pretty sure anyhow Aug 20 00:26:09 found some stuff about it for solaris too Aug 20 00:26:24 oh, right Aug 20 00:26:25 http://docs.sun.com/app/docs/doc/805-3050/6j2tng0mg?l=ko&a=view Aug 20 00:26:50 okay, so the moron worked at sun, i guess? Aug 20 00:27:02 still needs a good cluebat to the side of the head Aug 20 00:31:36 he was probably writing a zone file... Aug 20 00:32:13 and on that note, the guy who came up with the zone file syntax needs a 2x4 on the side of the head as well Aug 20 01:36:19 03Chris Larson  07org.openembedded.dev * rb8b2ca6b0d 10openembedded.git/classes/amend.bbclass: Aug 20 01:36:19 amend.bbclass: work with all versions of bitbake Aug 20 01:36:19 Latest bitbake uses a set for __depends, rather than a list, so handle that Aug 20 01:36:19 when we add the nonexistent files. Aug 20 01:36:19 Signed-off-by: Chris Larson **** ENDING LOGGING AT Fri Aug 20 02:59:56 2010