**** BEGIN LOGGING AT Wed Apr 01 02:59:57 2009 Apr 01 06:18:08 good morning Apr 01 06:19:31 you have a special nick just for #oe? Apr 01 07:35:25 good morning Apr 01 07:35:36 how do you get the stable branch ? Apr 01 08:07:18 03Jeremy Lainé  07org.openembedded.dev * r51031ca40a 10openembedded.git/recipes/ (4 files in 2 dirs): qtopia-core-*: remove obsolete qtopia tasks and images Apr 01 08:08:40 03Jeremy Lainé  07org.openembedded.dev * r27fe86c909 10openembedded.git/MAINTAINERS: MAINTAINERS: add qt4/* to my entries Apr 01 08:09:18 good morning Apr 01 08:15:42 methril|work: good morning Apr 01 08:16:05 some of the core devs tested the qemu patches? Apr 01 08:31:06 03Jeremy Lainé  07org.openembedded.dev * rb27607f52b 10openembedded.git/recipes/qt4/ (qt4-embedded_4.4.3.bb qt4-embedded_4.5.0.bb): qt4-embedded: rename qt-embedded-* to qt4-embedded-* Apr 01 08:37:21 i have booted out a angstrom image, On the LCD i see flashing of the Angstrom Login Screen which is displayed on the serial console (which now appears on the LCD) and the Anstrom GUI Login Screen, Both of the flash repeatedly. what could be wrong? Apr 01 09:03:14 anybody knows how do you get the stable branch ? Apr 01 09:03:44 mckoan, I beleive it is a branch in the git repository. Apr 01 09:04:10 git is a developmental branch Apr 01 09:04:50 somebody should tag it, i have not seen a tag Apr 01 09:05:00 if I do git clone git://git.openembedded.net/openembedded I get only the main Apr 01 09:05:11 $ git branch Apr 01 09:05:11 * org.openembedded.dev Apr 01 09:05:42 I'm not a git expert Apr 01 09:06:25 mckoan: git branch -a Apr 01 09:06:55 mckoan: to checkout just use: git co origin/stable/2009 -b stable/2009 Apr 01 09:08:24 stefan_schmidt: git checkout origin/stable/2009 -b stable/2009 Apr 01 09:08:31 stefan_schmidt: thanks Apr 01 09:08:51 mckoan: ah, right I have an alias here :) Apr 01 09:09:00 * mckoan is adding this to wiki Apr 01 09:20:55 good morning Apr 01 09:22:59 morning florian Apr 01 10:04:10 could anybody point me to a MMU-less machine based on OE (ARM7, ColdFire) ? Apr 01 10:04:49 * mckoan wonders if OE could replace a uClinux environment Apr 01 10:25:13 mckoan: Some of the blackfins iirc Apr 01 10:25:31 mckoan: That's something I wonder too. Apr 01 10:28:47 mckoan: ADSP-BF537 STAMP Board runs on uclinux Apr 01 10:38:57 hi mickey|OpenExpo Apr 01 10:40:49 hey florian Apr 01 10:46:33 03Felix Domke  07org.openembedded.dreambox * r82b3280a0c 10openembedded.git/packages/enigma2/enigma2.bb: enigma2.bb: add missing space Apr 01 10:58:00 mickey|OpenExpo: what about the qemu update? Someone tested it? Apr 01 11:46:32 mckoan: i think there is a mips mmu less machine also, not quite sure though. sh maybe? Apr 01 12:04:59 Hi, I've just created a filesystem with the latest openembedded.dev packages, with qt-embedded into it Apr 01 12:05:08 likewise: thx Apr 01 12:05:53 i've just loaded it on the board (olimex sam9-l9261), and I wanted to test the framebuffer device Apr 01 12:06:13 mckoan: Jin^eLD knows more, but I don't find him on IRC last months Apr 01 12:06:25 ~seen Jin^eLD Apr 01 12:06:29 jin^eld was last seen on IRC in channel #oe, 25d 19h 16m 13s ago, saying: 'hi woglinde '. Apr 01 12:06:44 vsr: are you using OE on STAMP Board ? Apr 01 12:09:38 So I ran the framebuffer test example from qt, and it does not draw anything at all Apr 01 12:10:29 I just keep seeing the penguin and 'Please wait: booting...' Apr 01 12:11:25 mckoan: my colleague was working on it Apr 01 12:12:42 vsr: fine thx Apr 01 12:21:15 morning Apr 01 12:36:55 im running xmms it is not displaying the time properly Apr 01 12:37:19 it is displaying some 90 minutes for a 5 minutes song Apr 01 12:39:25 bonus playtime :-D Apr 01 12:40:02 Hi all Apr 01 12:40:37 I'm new in this channel, I have a simple question to how use a prebuild toolchain Apr 01 12:40:49 I try to create a configuration for my migor sh4 board Apr 01 12:41:05 and and use the titan board file as initial startup Apr 01 12:41:29 Then I wold like to use a foreigh toolchain and add the variable in local.conf. The process Apr 01 12:41:59 stop during the compilation of libtool_cross, Apr 01 12:42:49 It try to create an executable file and test, but the the configuration variable seems to be ok Apr 01 12:43:10 I do an ack on the configure script or is there a workaround? Apr 01 12:45:51 Hi folks. Apr 01 12:46:01 I have a class which declares "addtask create_ini before do_rebuild after do_build". When bitbake executes do_rebuild, create_ini is executed. But when do_rebuild is not executed, create_ini is not executed, even if do_build is. Is it expected? Apr 01 12:54:49 im trying to run mplayer on my board, i get the message: [mov,mp4,m4a,3gp,3g2,mj2 @ 0xb22750]moov atom not found when i try to play a video file Apr 01 12:56:08 Hello. Where can i find answer for this: how do i compile msc7830.c (usb-ethernet adapter) driver for beagleboard? i have source for ubuntu, but trying compile with bitbake have "not found /linux/module.h" and others Apr 01 12:57:08 did you inherit module.bbclass Apr 01 12:57:09 ? Apr 01 12:57:24 no... Apr 01 12:57:30 then do so :-) Apr 01 12:57:36 it sets the kernel paths for module compiles Apr 01 12:57:54 classes/module.bbclass to see what it does Apr 01 12:57:58 ops sory, i found that angstrom on kernel v28 allready have it =) im happy! 2 days of searching... Apr 01 12:58:03 just put an inherit module in the .bb file Apr 01 12:58:13 thanks Apr 01 12:59:15 xora: do you know why mplayer says [mov,mp4,m4a,3gp,3g2,mj2 @ 0xb22750]moov atom not found when i try to play a video Apr 01 12:59:34 is mplayer on oe broken? Apr 01 12:59:50 vsr: it works on some platforms for sure Apr 01 13:00:52 i compiled for omap and it says LAVF_header: av_open_input_stream() failed Apr 01 13:00:55 see http://pastebin.ca/1378885 Apr 01 13:04:22 vsr: looks like that means your file is b0rk3d Apr 01 13:05:24 xora: it plays audio it wont play a video file Apr 01 13:05:39 xora: what do you mean by b0rk3d Apr 01 13:06:08 vsr: all the references to that error I find in google mean a bit of the mp4 file are missing Apr 01 13:12:43 03Angus Ainslie  07fso/milestone5.5 * r055519888c 10openembedded.git/ (conf/checksums.ini packages/hotplug2/hotplug2_0.9+1.0beta.bb): hotplug2: added new recipe and updated checksums.ini Apr 01 13:14:28 xora: i got what the problem was Apr 01 13:14:50 xora: those files were indeed broken as you said, i downloaded again and they work Apr 01 13:14:53 xora: thanks Apr 01 13:17:56 vsr: no probs Apr 01 13:34:00 03Koen Kooi  07org.openembedded.dev * rd418b37161 10openembedded.git/ (conf/checksums.ini recipes/gnome/epiphany_2.27.1.bb): Apr 01 13:34:00 epiphany: update to 2.27 Apr 01 13:34:00 * much better integration with webkit Apr 01 13:34:00 * runtime tested on beagleboard Apr 01 13:44:16 03ghost  07org.openembedded.dreambox * r845b2a7e01 10openembedded.git/conf/distro/opendreambox-1.5.conf: opendreambox-1.5.conf: set correct feed for release 25 Apr 01 13:44:16 03ghost  07org.openembedded.dreambox * r7ebca4e5f4 10openembedded.git/: Merge branch 'org.openembedded.dreambox.release_25' of git.opendreambox.org:git/opendreambox-1.5 into org.openembedded.dreambox.release_25 Apr 01 13:44:17 03Felix Domke  07org.openembedded.dreambox * rbd7346af0f 10openembedded.git/: Merge branch 'org.openembedded.dreambox.release_25' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox.release_25 Apr 01 13:44:22 03ghost  07org.openembedded.dreambox * rc65b7beab7 10openembedded.git/packages/vsftpd/vsftpd_2.0.5.bb: vsftpd_2.0.5.bb: add build hack (dont use libpcap) Apr 01 13:44:25 03ghost  07org.openembedded.dreambox * r87e9986abb 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules.bb: new dm8000 drivers Apr 01 13:44:28 03Felix Domke  07org.openembedded.dreambox * r6eea81063d 10openembedded.git/: Merge branch 'org.openembedded.dreambox.release_25' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox.release_25 Apr 01 13:44:32 03ghost  07org.openembedded.dreambox * rdc6d884c68 10openembedded.git/: Merge branch 'org.openembedded.dreambox.release_25' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Apr 01 13:44:36 03ghost  07org.openembedded.dreambox * r3ebe83d9f9 10openembedded.git/packages/gstreamer/gst-plugin-dvbmediasink.bb: gst-plugin-dvbmediasink.bb: push srcdate because of aac fixes Apr 01 13:44:39 03Felix Domke  07org.openembedded.dreambox * r60c43dafb7 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox Apr 01 13:44:47 03ghost  07org.openembedded.dreambox * r150d6d0c26 10openembedded.git/packages/vsftpd/vsftpd_2.0.5.bb: vsftpd_2.0.5.bb: add build hack (dont use libpcap) Apr 01 13:44:50 03Felix Domke  07org.openembedded.dreambox * r1fc89b8c68 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules: update for dm8000 Apr 01 13:44:53 03ghost  07org.openembedded.dreambox * r6c105a5696 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Apr 01 14:14:24 03Angus Ainslie  07fso/milestone5.5 * r1cb8c3a513 10openembedded.git/ (conf/checksums.ini packages/hotplug2/hotplug2_0.9+1.0beta.bb): hotplug2: added new recipe and updated checksums.ini Apr 01 14:14:25 03Angus Ainslie  07fso/milestone5.5 * rebce434899 10openembedded.git/: Apr 01 14:14:25 Merge branch 'fso/milestone5.5' of git@git.openembedded.net:openembedded into fso/milestone5.5 Apr 01 14:14:25 Conflicts: Apr 01 14:14:25 packages/hotplug2/hotplug2_0.9+1.0beta.bb Apr 01 14:16:49 03Angus Ainslie  07fso/milestone5.5 * ra329108336 10openembedded.git/packages/tasks/task-openmoko-feed.bb: task-openmoko-feed: add hotplug2 to the feed Apr 01 14:18:40 who is the patchwork maintiner? Apr 01 14:21:51 I do some with patchwork Apr 01 14:22:04 methril|work, what is the problem? Apr 01 14:22:28 Crofton|work: i forget my pass and i can not login in Apr 01 14:22:40 * Crofton|work beats methril|work Apr 01 14:22:48 let me see what I can do Apr 01 14:22:57 * methril|work beats methril|work Apr 01 14:24:25 * kgilmer stands back, doesn't want to get hit with a stray fist. Apr 01 14:26:09 ./msg methril|work it is changeme now Apr 01 14:26:14 ./msg methril|work Change it Apr 01 14:27:22 * methril|work reminds own one to Crofton|work Apr 01 14:27:55 <_cpo__> anybody ever tried building xserver 1.6 with oe? i always run into a libtool error Apr 01 14:29:05 03Robert Schuster  07org.openembedded.dev * r2d05226a33 10openembedded.git/recipes/llvm/llvm.inc: llvm.inc: Declare dependency of llvm on llvm-native Apr 01 14:29:23 Crofton|work: I'km not able to do it :( Apr 01 14:30:57 try again Apr 01 14:32:48 03Koen Kooi  07stable/2009 * rf341a82e30 10openembedded.git/conf/machine/omapzoom.conf: Apr 01 14:32:48 omapzoom: add xf86-input-tslib Apr 01 14:32:48 Signed-off-by: Koen Kooi Apr 01 14:32:48 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:32:50 03Koen Kooi  07stable/2009 * r844e5789b2 10openembedded.git/recipes/linux/ (3 files in 3 dirs): Apr 01 14:32:53 linux-omap: add patch to fix mmc support and update defconfig Apr 01 14:32:55 Signed-off-by: Koen Kooi Apr 01 14:32:57 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:32:59 03Koen Kooi  07stable/2009 * ree4e2374ea 10openembedded.git/recipes/webkit/webkit-gtk_svn.bb: Apr 01 14:33:02 webkit-gtk: fix PV Apr 01 14:33:04 Signed-off-by: Koen Kooi Apr 01 14:33:06 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:33:08 03woglinde  07stable/2009 * re27269a9bd 10openembedded.git/ (7 files in 3 dirs): Apr 01 14:33:11 libgsm: update to version 1.0.12 Apr 01 14:33:13 Signed-off-by: Koen Kooi Apr 01 14:33:15 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:33:17 03Koen Kooi  07stable/2009 * r8c88c5f136 10openembedded.git/conf/machine/omapzoom.conf: Apr 01 14:33:20 omapzoom: linux-omap work better on the device than linux-omapzoom, so switch to that Apr 01 14:33:26 Signed-off-by: Koen Kooi Apr 01 14:33:28 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:33:30 03Koen Kooi  07stable/2009 * r463b84c0e1 10openembedded.git/recipes/e17/places_svn.bb: Apr 01 14:33:33 places: bump PR Apr 01 14:33:35 * bumped PR to force a rebuild against updated E libs Apr 01 14:33:37 Signed-off-by: Koen Kooi Apr 01 14:33:39 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:35:27 how can i find out what is the value of SRCREV? Apr 01 14:35:47 does this variable determine which source revision needs to be pulled from the svn Apr 01 14:44:40 03Koen Kooi  07stable/2009 * rd83bbdf72c 10openembedded.git/ (conf/checksums.ini recipes/gnome/epiphany_2.27.1.bb): Apr 01 14:44:40 epiphany: update to 2.27 Apr 01 14:44:40 Signed-off-by: Koen Kooi Apr 01 14:44:40 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:57:48 03Koen Kooi  07stable/2009 * r7ee5a76ba9 10openembedded.git/recipes/linux/ (7 files in 6 dirs): Apr 01 14:57:48 linux-omap 2.6.29: disable alignment trapping completely for omap3board Apr 01 14:57:48 Signed-off-by: Koen Kooi Apr 01 14:57:48 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:57:51 03Koen Kooi  07stable/2009 * r22f65c206b 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Apr 01 14:57:53 linux-omap 2.6.29: update mmc regulator patch Apr 01 14:57:55 Signed-off-by: Koen Kooi Apr 01 14:57:57 Acked-by: Philip Balister (philip@balister.org) Apr 01 14:57:59 03Koen Kooi  07stable/2009 * rec8d0041da 10openembedded.git/recipes/initscripts/ (files/arm/alignment.sh initscripts_1.0.bb): Apr 01 14:58:02 initscripts: set alignment to 0 for armv7a Apr 01 14:58:04 Signed-off-by: Koen Kooi Apr 01 14:58:06 Acked-by: Philip Balister (philip@balister.org) Apr 01 15:03:36 morning Apr 01 15:03:53 gm Apr 01 15:11:04 are there any good file browsers with matchbox as wm Apr 01 15:18:08 hm, git pull is not doing anything for me Apr 01 15:18:19 anyone else have this too ? Apr 01 15:20:20 works now, apparently took a long time to get going Apr 01 15:31:39 03Jeremy Lainé  07org.openembedded.dev * r54cf7b27ea 10openembedded.git/recipes/linux/ (3 files in 2 dirs): linux-2.6.27: remove unnecessary patch for boc01's PHY Apr 01 15:33:00 vsr: ROX-Filer? Apr 01 15:39:35 Tartarus, are you working with any powerpc 6xx machines? Apr 01 15:41:49 I still own some, but no, atm i'm not working with any Apr 01 15:55:44 jay7: thanks i will try that Apr 01 15:56:23 looking for a halbrowser on match box Apr 01 15:56:45 i want something equivalent of places of enlightment on matchbox Apr 01 16:01:36 Tartarus, which machines were you using? I'm not able to build meta-toolchain for lite5200, not sure that I ever could Apr 01 16:01:58 doesn't seem like anyone here using ppc anymore Apr 01 16:02:18 Ah, I've never done OE for 'em :) Apr 01 16:02:40 But, not too long ago I did one of the 82xx boards which are a 603 still, iir Apr 01 16:02:43 c Apr 01 16:03:25 there a machine conf for one of those? Apr 01 16:03:47 yeah Apr 01 16:04:01 and i meant 83, heh Apr 01 16:04:02 mpc8313e-rdb.conf mpc8323e-rdb.conf Apr 01 16:04:52 whats the diff between the ppce300c* and ppc603e? Apr 01 16:06:57 Hmm. I've been out of it too long, just assume I'm wrong for now ;) Apr 01 16:07:22 heh, ok Apr 01 16:22:55 the Gtk+ node in the .dev tree is empty. Am I not looking in the right place or is Gtk+really not supported in OE ? Apr 01 16:24:46 rkirti: "packages/gtk+" should be the right place Apr 01 16:25:09 s/packages/recipes/, no? Apr 01 16:26:08 florian, kergoth : I checked root/recipes. Apr 01 16:36:52 03Koen Kooi  07org.openembedded.dev * ra0286c180d 10openembedded.git/conf/checksums.ini: checksums: add more packages Apr 01 16:36:52 03Koen Kooi  07org.openembedded.dev * r0e26e0a6fd 10openembedded.git/recipes/gnome/gnome-desktop_2.26.0.bb: gnome-desktop: update to 2.26 Apr 01 16:37:25 i love you, git Apr 01 16:49:35 03Jeremy Lainé  07org.openembedded.dev * rb44ce1af2e 10openembedded.git/recipes/linux/ (11 files in 2 dirs): linux-2.6.29: add patches for boc01 Apr 01 16:52:51 that's odd. I just pushed a number of commits to a new branch, and CIA didn't notify Apr 01 16:53:34 I suspect it's only notifying on added commits in a given branch and just doesn't have a notifier for new branches. Apr 01 16:54:07 aye, i could see that Apr 01 16:54:28 it could be smarter though, try to determine what it was branched off of, and notify for the subsequent commits Apr 01 16:54:34 but oh well, i can see why its not done Apr 01 16:54:46 wv :) Apr 01 16:54:54 :) Apr 01 16:55:42 * kergoth 's ripping autotools.bbclass all up Apr 01 16:57:41 kergoth: can you confirm offhand checking for EVIOCGBIT(EV_ABS is the right way to detect if an event is touchscreen? Apr 01 16:57:59 pretty sure that's correct, you can take a look at input.c in tslib source Apr 01 16:58:26 thx, I was looking at detect-stylus.c Apr 01 16:58:49 then found http://www.codase.com/search/call?name=xf86ioctl Apr 01 16:59:03 where they test more bits...but one should suffice Apr 01 17:00:32 03Jeremy Lainé  07org.openembedded.dev * rd323bccf4d 10openembedded.git/recipes/hostap/hostap-daemon_0.6.9.bb: hostap-daemon-0.6.9: disable madwifi-ng support Apr 01 17:07:45 kergoth: thx again Apr 01 17:08:11 np Apr 01 17:19:13 03Chris Larson  07kergoth/master * rde1daef44b 10openembedded.git/conf/ (bitbake.conf collections.inc): (log message trimmed) Apr 01 17:19:13 bitbake.conf, collections.inc: Add COLLECTIONS mechanism w/ a sane default. Apr 01 17:19:13 Added code which takes a list of directories in COLLECTIONS, in priority Apr 01 17:19:13 order (highest to lowest), and uses those to populate BBFILES, Apr 01 17:19:13 BBFILE_COLLECTIONS, BBFILE_PATTERN_*, and BBFILE_PRIORITY_*. Apr 01 17:19:17 The default COLLECTIONS is based on BBPATH, so you can now specify a sane Apr 01 17:19:19 BBPATH, not bother setting up the collections variables or BBFILES, and Apr 01 17:19:26 03Chris Larson  07kergoth/master * rdc93678251 10openembedded.git/conf/collections.inc: Apr 01 17:19:26 collections.inc: add support for collection archives. Apr 01 17:19:26 Adds the ability to specify an archive in COLLECTIONS. Any archives of a Apr 01 17:19:28 supported format (default tar.gz/tgz/tar.X, tar.bz2/tbz/tbz2, zip/jar) will be Apr 01 17:19:32 unpacked into COLLECTIONS_UNPACKDIR and used from there. The default value of COLLECTIONS_UNPACKDIR is ${TMPDIR}/collections. Apr 01 17:19:35 Signed-off-by: Chris Larson Apr 01 17:26:47 Anyone ever done a build w/ THUMB_INTERWORK=no? Apr 01 17:26:59 I'm seeing busybox be mostly fine except for some applets (awk being one of them) Apr 01 17:27:04 trying to rule out thumb interwork now Apr 01 17:27:12 but.. it builds w/ -mno-thumb-interwork -mno-thum Apr 01 17:27:13 b Apr 01 17:27:15 so i'm confused Apr 01 17:29:07 03Koen Kooi  07org.openembedded.dev * rf2b643f923 10openembedded.git/ (2 files in 2 dirs): Apr 01 17:29:07 eds-dbus: bump SRCREV Apr 01 17:29:07 * adjust patches accordingly Apr 01 17:29:09 03Koen Kooi  07org.openembedded.dev * r612eb226ee 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 17:29:09 03Koen Kooi  07org.openembedded.dev * rd2e9587dd3 10openembedded.git/recipes/gnome/cheese_2.26.0.bb: Apr 01 17:29:12 cheese: add 2.26 Apr 01 17:29:14 * cheese it the GNOME version of Apple's photobooth Apr 01 17:33:34 * kergoth wonders if itd be useful to have more than a single "HOMEPAGE" url, for a package Apr 01 17:38:38 What does it actually get used for in OE?# Apr 01 17:43:29 either nothing, or packaging, not sure which Apr 01 17:45:20 kergoth - is the homepage mostly useful for detective work when 404 errors happen? Apr 01 17:46:29 that was part of the intent, yes. along with having a place to contact to submit patches upstream.. but really, not all projects even have homepages. itd be nice to have something more generally useful.. list contact email addresses, and list multiple useful URIS for the project, including the BTS, etc Apr 01 17:47:36 kergoth - yes; start from the ideas around the failure modes, and how such information might address it - redundancy is good. Apr 01 17:47:59 it's more than just for failures, its also documenting where to go to push patches upstream. Apr 01 17:48:33 In which case multiple homepages might be useful and shouldn't confuse software, I guess. Apr 01 17:49:26 Apr 01 17:49:48 maybe just a space separated list of contact points in an UPSTREAM_CONTACT var or somesuch Apr 01 17:49:52 * kergoth shrugs Apr 01 17:50:14 One contact per line? ;-) Apr 01 17:50:22 Web URLs can be as useful; people change over time. Apr 01 17:50:27 OpenEmbedded variables can't be multiline in a .conf. Apr 01 17:50:34 it's space separated or nothing Apr 01 17:50:43 but URLs tend to be a bit more stable (if only due to the proejct name!) Apr 01 17:50:51 broonie: web urls can change over time as well. i wouldn't add direct email addresses for individuals, more mailing lists Apr 01 17:52:48 Yeah, like I say they do tend to be a bit mroe stable though (since lots of stuff has project.org or whatever with the probjet name in it). Apr 01 17:55:23 it depends. an actively maintained project run by individuals, that's generally the case. projects maintained by distros, sadly, that's rarely the case :( Apr 01 17:55:58 In those cases your best contact is normally /dev/null anyway :) Apr 01 17:56:09 hehe, true that Apr 01 17:56:13 sadly Apr 01 17:56:32 * kergoth doesn't get why the distros cant set up public scm's and such for those things.. Apr 01 17:56:53 See the /dev/null issue ;0 Apr 01 17:57:05 Normally you've got too much slew between the distros to merge stuff. Apr 01 17:57:21 (at least, without someone motivated to work through the cruft). Apr 01 17:57:31 Besides, OE is a distro - you could do it! Apr 01 17:57:37 aye, that's why i like the vcs-pkg effort, even if they haven't gotten far yet Apr 01 17:57:46 OE doesn't maintain sources to much. almost everything in the repo comes from elsewhere Apr 01 17:58:01 Most other distros do the same. Apr 01 17:58:10 http://vcs-pkg.org/ Apr 01 17:58:13 * broonie will be vaugely surprised if vcs-pkg gets terribly far. Apr 01 17:58:18 Yeah, I'm on the list. Apr 01 17:58:36 Ubuntu are actually doing something along those lines for their stuff though it's all auto-imports and whatnot AFAICT. Apr 01 17:58:40 i like the concepts, but i think their attempt to remain vcs agnostic is going to result in nothing happening Apr 01 17:59:01 Quite, plus it's questionable if stuff like branch per feature is actually going to prove workable. Apr 01 17:59:02 if they'd just start setting up some git repositories and let various distro folk have at it, they'd probably get farther Apr 01 17:59:28 Half of the people talking are distro developers with packages in DVCSs already. Apr 01 17:59:31 if you use something like git, it's viable using something like topgit to update them and record inter-topic-branch dependencies Apr 01 17:59:33 (like me, not that I talk on the list) Apr 01 18:00:20 * kergoth wants to set up some vcs-pkg-like git repositories for individual OpenEmbedded packages, with the recipes in the repositories, and toy around with it Apr 01 18:00:20 It's possible, it's questionable if people actually want to work that way, though. Apr 01 18:00:41 kergoth - go ahead :-) Apr 01 18:00:56 well, ideally, the feature branches won't stay around indefinately, itd go upstream.. and the only long life branches would be metadata.. Apr 01 18:00:57 See my wv comments earlier :) Apr 01 18:01:08 kergoth: Meanwhile, in this universe... Apr 01 18:01:12 but that's in a happy world quite unlike the one we live in. Apr 01 18:01:14 haha Apr 01 18:01:36 nis upstream is especially annoyiny, he's there but only seems to answer mail if you criticise the network manager integration. Apr 01 18:02:04 robtow: i started on a task that could emit a git repo from a bitbake recipe, setting up all the feature branches for all the patches automatically.. the hard part is trying to record the dependencies, and also to handle SRC_URI changing over time in the OpenEmbedded git repo Apr 01 18:02:16 * kergoth never went back to it Apr 01 18:02:32 WOuld that problem be more reasonable the second time around? Apr 01 18:03:11 its doable, but like any automated setup it'd need human review.. and it was just a proof of concept anyway, i really don't know if such a setup for the sources & OpenEmbedded metadata is viable Apr 01 18:03:23 was going to use it to convert task-boot or something and poke around Apr 01 18:06:27 * kergoth has lots of crazy ideas, many of which are probably not viable :) Apr 01 18:07:07 If all of your ideas are viable, you aren't being daring enough. Apr 01 18:07:56 good point Apr 01 18:13:53 ugh, what is up with this popt recipe Apr 01 18:26:29 webkit-gtk-1.1.3+svnr42069-r0 I have problems in www.youtube.com with midori and epiphany: some strings are upside down ... very strange Apr 01 18:28:03 I have seen openembedded anouncement: I have to do git checkout -b openembedded.stable origin org.openembedded.stable ? Apr 01 18:33:26 03Frans Meulenbroeks  07org.openembedded.dev * ra2f2b1a1bc 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 18:33:37 03Frans Meulenbroeks  07org.openembedded.dev * rd67a64916e 10openembedded.git/recipes/linux/ (linux-omap-2.6.29/mmctiming.patch linux-omap_2.6.29.bb): linux-omap_2.6.29.bb: added mmctiming patch Apr 01 18:43:55 now I compile a beagleboard-demo-image with this version and I try midori and epiphany on revB6 beagleboard Apr 01 20:10:23 re Apr 01 20:10:36 jo florian Apr 01 20:12:23 hi Apr 01 20:12:55 is there some way to set the hostname of the system before building the root filesystem? Apr 01 20:30:41 jo ant Apr 01 20:32:54 hey woglinde, ant__, florian Apr 01 20:33:11 hi kergoth Apr 01 20:33:17 pplz :) Apr 01 20:34:05 hey there Apr 01 20:34:11 hi kergoth Apr 01 20:35:10 is there some way to set the hostname and root password for the embedded system before building the root filesystem? Apr 01 20:44:49 Aegiron: prepare a line in your pc and add it to /etc/passwd at its end Apr 01 20:45:05 prepare an account in your pc .. Apr 01 20:46:07 maybe there are other files similar to /etc/passwd to be modified. Apr 01 20:46:38 how do you mean, I thought it would be easy to just use the generated jffs2 file to upload, without editing files of the root filesystem and repacking thereafter Apr 01 20:47:42 like setting a variable in some image recipe or the machine's config file.. Apr 01 20:49:56 recalcati_oe: maybe an incredibly stupid question, what does the abbreviation pc stand for? Apr 01 20:54:01 Aegiron edit the basefiles Apr 01 20:54:06 in the recipe Apr 01 20:59:34 03Andrea Adami  07org.openembedded.dev * r791a9bc3f0 10openembedded.git/conf/distro/angstrom-2008.1.conf: Apr 01 20:59:34 angstrom-2008.1.conf: add PREFERRED_VERSION_psplash Apr 01 20:59:34 -Acked-by: Koen Kooi Apr 01 21:00:26 okay new xine goes in Apr 01 21:00:34 deleting the whole old stuff Apr 01 21:00:40 libxine fixed? Apr 01 21:00:50 yes Apr 01 21:00:58 great! Apr 01 21:01:03 I need it for directfb and disko Apr 01 21:01:13 2 bugs less...let me close them Apr 01 21:01:15 thx Apr 01 21:01:17 but I have to patch ffmpeg in the next step Apr 01 21:01:55 * kergoth looks into adding chkconfig to OpenEmbedded Apr 01 21:01:58 03woglinde  07org.openembedded.dev * re59631d085 10openembedded.git/recipes/libxine/ (39 files in 6 dirs): Apr 01 21:01:58 libxine: remove older verison and add the actual 1.1.16 Apr 01 21:01:58 * dont seperate x11 and fb anymore xine can handle it via plugins Apr 01 21:01:58 * therefore all recipes wich depends on libxine-fb or libxine-x11 Apr 01 21:01:58 have to changed to libxine Apr 01 21:02:05 03woglinde  07org.openembedded.dev * rfabea12c54 10openembedded.git/conf/checksums.ini: checksum.ini: add xine-1.1.16 remove older versions Apr 01 21:02:06 03woglinde  07org.openembedded.dev * r02f4e5ae5d 10openembedded.git/ (11 files in 6 dirs): Apr 01 21:02:07 xine: update to new xine recipe Apr 01 21:02:09 * as a side node only opie-mediaplayer2 compiles out of the box Apr 01 21:02:22 do we have a shorty on how to use patchwork in the OE environment? Apr 01 21:02:34 like2wise: good question Apr 01 21:02:34 hehe now I am waiting for the failures which may come Apr 01 21:02:44 woglinde: lol, run!!! Apr 01 21:03:09 woglinde:is there anything particular where I shoudl be looking for? Apr 01 21:03:09 like2wise actually nobody really used xine in oe Apr 01 21:03:57 aegiron hm no all the recipe which depends on it besides opie-mediaplayer are too old to be used or dont build anyway in oe.dev Apr 01 21:04:10 so Apr 01 21:04:14 now comes disko Apr 01 21:04:26 then ffmpeg arm assembler fix Apr 01 21:05:32 woglinde: erh, using xine to setup I hostname and default root password does'nt appeal to me ;) Apr 01 21:06:42 03woglinde  07org.openembedded.dev * r3d25ff174b 10openembedded.git/recipes/disko/ (disko_git.bb files/header.patch files/pkgconfig.patch): disko: add disko framework to openembedded Apr 01 21:07:10 aegrion oh sorry Apr 01 21:08:42 haha, nvm :) Apr 01 21:09:03 hm Apr 01 21:09:22 sorry seems you could only edit hosts easily Apr 01 21:09:39 dont know if it is enough Apr 01 21:09:44 hm..the last psplash patch ... Apr 01 21:10:04 is this xubuntu based ? Apr 01 21:10:15 * Jay7 is testing inotify on /proc/filesystems Apr 01 21:10:18 ah Apr 01 21:10:22 03Stanislav Brabec  07org.openembedded.dev * r95f33f6c42 10openembedded.git/recipes/gtk+/gtk+.inc: gtk+.inc: Package modules .debug, remove unusable modules .la files. Apr 01 21:10:27 grep is my friend Apr 01 21:10:32 03Stanislav Brabec  07org.openembedded.dev * rdbaf04e7de 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 21:10:57 Aegiron look at base-files/base-files_3.0.14.bb Apr 01 21:12:00 hmmm Apr 01 21:12:53 Oh, my bad, I thought base files was just a generalisation for the recipes needed for a basic rootfs, I didn't expect it to be a recipe itself Apr 01 21:13:45 Just seen on an irc channel near you: 23:05 Hey timtimred - thanks for pointing me at OE the other day. That thing rocks! Apr 01 21:14:03 :-) Apr 01 21:14:29 he's right, it DOES rock indeed, I've just discovered it myself two weeks ago Apr 01 21:14:29 ~lart gmane ML Apr 01 21:14:29 * ibot grabs a large, mis-shapened log, with squirrels, and beats gmane ML until only the nuts remain ... which the squirrels run off with Apr 01 21:15:38 * ant__ thinks the old way is just better Apr 01 21:15:40 http://lists.linuxtogo.org/pipermail/openembedded-devel/ Apr 01 21:15:44 :D Apr 01 21:16:01 * timtimred is the OE pimp Apr 01 21:17:19 "pimp my OE" :) Apr 01 21:17:28 in creating my own recipes and images, i used a naming convention "blah_1.0.0.bb" for the recipe and "blah-image.bb", but recently, IMAGE_INSTALL += "blah" no longer works - i had to rename my recipe to blah.bb? Apr 01 21:17:33 with like2wise Apr 01 21:17:46 is this expected behavior? am i fowling something up...? Apr 01 21:32:38 someone with a armv5 or armv6 toolchain here, who can test my patch for ffmpeg from? http://paste.debian.net/32165/ Apr 01 21:34:57 03Stanislav Brabec  07org.openembedded.dev * rb711f9d53e 10openembedded.git/recipes/gtk+/ (gtk+_2.14.2.bb gtk+_2.16.0.bb): gtk+: Replace also gail subpackages. Apr 01 21:38:25 Jay7: that ioctl stuff is a bit though Apr 01 21:41:42 ant__: but this is normal interface for such things.. Apr 01 21:43:11 Jay7: just merge the check Apr 01 21:44:37 ant__: hehe.. nowhere to merge yet.. Apr 01 21:51:28 bad news.. inotify does not intercept any events on /proc/partitions Apr 01 21:52:04 easy way is closed Apr 01 22:11:35 did someone saw my mail about evince? Apr 01 22:13:41 03Andrea Adami  07org.openembedded.dev * r0c8dd633e2 10openembedded.git/conf/distro/angstrom-2008.1.conf: Apr 01 22:13:41 angstrom-2008.1.conf: fix typos Apr 01 22:13:41 -golden rule: don't commit if you're tired... Apr 01 22:14:12 grrr Apr 01 22:17:16 * Jay7 -> sleep Apr 01 22:24:10 trying to get gcc4.3.x working on my powerpc target - I find in the compile log that ld complains about some objects using hard float and others using soft float yet I'm not sure where in the gcc build thats setup (I don't see anything on the compiler cmdline indicating hard vs soft) - any ideas? Apr 01 22:28:45 03Chris Larson  07org.openembedded.dev * rd27e47d2d1 10openembedded.git/MAINTAINERS: Apr 01 22:28:45 MAINTAINERS: I'll grab autoconf, automake, libtool, gnu-config, and the autoconf classes. Apr 01 22:28:45 Signed-off-by: Chris Larson Apr 01 22:31:42 03Chris Larson  07org.openembedded.dev * r813a40adbd 10openembedded.git/MAINTAINERS: Apr 01 22:31:42 MAINTAINERS: I can spell my own name. Really. Apr 01 22:31:42 Signed-off-by: Chris Larson Apr 01 22:34:40 Who maintains git.openembedded.net? Apr 01 22:42:56 03Angus Ainslie  07fso/milestone5.5 * r13c723a9f0 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs.inc: bump efl version to get rid of cpu hog problem Apr 01 22:45:28 03Angus Ainslie  07fso/milestone5.5 * r0e7ad7f0a5 10openembedded.git/packages/hotplug2/ (files/hotplug2.sh hotplug2_0.9+1.0beta.bb): hotplug2: ship an init file Apr 01 22:55:46 * otavio looks interested in hotplug2 Apr 01 22:55:48 :-) Apr 01 23:02:38 whew, got chkconfig added to oe Apr 01 23:02:43 that was slightly painful Apr 01 23:08:12 hi, I've been trying to setup a default root password in base-passwd, but I was wondering: can this actually be done in the first place? Apr 01 23:09:06 I mean, the passwords are kept in the /etc/passwd file, so I just copied the encoded passwd from my current running system. Apr 01 23:10:18 if this could work, how safe is it then? Apr 01 23:12:06 or does this not matter at all, as the only person who can edit /etc/passwd is root himself, and if a person with bad intention were to be root, you'd be fucked already... Apr 01 23:14:48 So, yeah... It's best to not have root login enabled on a shipping system:) Apr 01 23:14:59 So an encoded password of just 'x' or something will stop that Apr 01 23:16:04 hmm, the C version of update-alternatives from chkconfig works fine Apr 01 23:16:05 cool Apr 01 23:16:20 * kergoth 'll have to test an actual rootfs Apr 01 23:16:31 kergoth, and there's no busybox applet version yet? ;) Apr 01 23:16:42 (yes, ive been reading the ml thread on this) Apr 01 23:17:18 * kergoth thinks itd be good to stop depending on something from handhelds cvs Apr 01 23:18:33 Tartarus: it is not save, but can it be done? I thought passwords where encoded with a random seed. can the encoded string just be copied to another system? Apr 01 23:19:37 depends on if shadow is enabled? I think you can tho Apr 01 23:29:52 hmm... bitbake can follow an RDEPENDS.. figuring out what package to build to produce that package.. but is it smart enough to handle multiple packages emitting that package? can you express preferred rproviders? Apr 01 23:29:54 * kergoth scratches head Apr 01 23:31:02 * kergoth wonders if it can handle RPROVIDES, also Apr 01 23:33:56 * kergoth 'll have to talk to richard tomorrow Apr 01 23:35:42 ok... I'm git-challenged... can't figure out how to switch to the stable/2009 branch - don't I just do a 'git checkout stable/2009' ? Apr 01 23:35:44 Tartarus: just found out it is possible for sure. it's a default feature of openslug Apr 01 23:36:51 tharvey: you can checkout origin/stable/2009, but thatll be stuck to a specific commit, so you wont be able to commit. its more likely that you want to create a local branch that tracks stable, and check that out: git checkout -b stable origin/stable/2009 or so Apr 01 23:38:14 can't seem to get the hang of git - so by tracking stable how is that different from just doing an occasional 'git pull' to sync up? Apr 01 23:38:27 that *is* doing a git pull to sync up. Apr 01 23:38:31 you cant pull if you aren't on a local branch Apr 01 23:38:53 it happens that a clone creates a local branch that tracks the main upstream branch for you Apr 01 23:39:00 you have to do that yourself for any other branches Apr 01 23:39:15 do a git branch; then do a git branch -r Apr 01 23:39:29 you want a branch in the former that tracks a branch in the latter Apr 01 23:40:25 ah... ok the shortcut clone was taking for me then was misleading me about what really was happening - I've read all kinds of guides about git and just can't seem to grasp it... svn is easy compared Apr 01 23:40:32 for me to understand that is Apr 01 23:40:56 clone is basically git init, then git remote add, then git fetch, then makes a local branch tracking the main upstream branch and checks it out. Apr 01 23:41:06 you can do those steps yourself, and it gets you to the same point Apr 01 23:41:47 think of it this way. if you could commit to the remote branches, how would you be able to figure out the difference between what's upstream and what you've done? you'd no longer have a ref pointing at current upstream Apr 01 23:41:57 the remote branches are always whats upstream, you never mess with them directly Apr 01 23:42:08 instead you make a local branch that merges from the upstream, and pushes to the upstream, and work there Apr 01 23:46:38 is there a proper way to clean all packages, or do I just remove the tmp folder? Apr 01 23:47:03 removing tmp is the best route Apr 01 23:47:05 kergoth, thx - still missing some of the concepts but the more I work with oe and kernel the more I'll get it hopefully Apr 01 23:47:09 np Apr 01 23:47:56 Aegiron, I find that certain changes to conf require cleaning tmp - its painful but the right thing to do - I don't trust the deps much Apr 01 23:49:00 deps are fine, but they don't do everything for you. there are things that are quite difficult to keep track of, ilke knowing when dependent packages should be rebuilt when a lib is rebuilt... should really only happen when the ABI changes, but OpenEmbedded doesn't keep track of that Apr 01 23:49:27 tharvey: I just copied base-files and base-passwd to my local overlay and changed them there, and when rebuilding them i specifically needed to point to the files itself Apr 01 23:49:31 * kergoth thinks about what itd take to add that Apr 01 23:49:37 Aegiron: what do you men? Apr 01 23:49:38 mean? Apr 01 23:50:54 well, I edited those 2 recipes in order to set a custom hostame and a default root password, and to do such things, I never go messing about in the local copy of oe.dev Apr 01 23:51:10 instead, i copy them to an overlay and work from there Apr 01 23:51:33 but after cleanin base-files and base-passwd Apr 01 23:52:14 Aegiron, you can clean the recipes that use those files... bitbake -cclean Apr 01 23:52:49 where it gets a little messy is if those local files are used in say a task, that is part of an image, then you need to rebuild the task and the image Apr 01 23:53:00 Aegiron: so what's the problem? Apr 01 23:54:33 well, nothing anymore :) Apr 01 23:56:13 In belgium its already 2am, so rebuilding everything (in case it would be required), should'nt be much of an issue Apr 01 23:56:54 03Jeremy Lainé  07kergoth/autotools * rd323bccf4d 10openembedded.git/recipes/hostap/hostap-daemon_0.6.9.bb: hostap-daemon-0.6.9: disable madwifi-ng support Apr 01 23:56:55 03Koen Kooi  07kergoth/autotools * rd2e9587dd3 10openembedded.git/recipes/gnome/cheese_2.26.0.bb: Apr 01 23:56:55 cheese: add 2.26 Apr 01 23:56:55 * cheese it the GNOME version of Apple's photobooth Apr 01 23:56:57 03Koen Kooi  07kergoth/autotools * rf2b643f923 10openembedded.git/ (2 files in 2 dirs): Apr 01 23:57:00 eds-dbus: bump SRCREV Apr 01 23:57:01 * kergoth sighs Apr 01 23:57:02 * adjust patches accordingly Apr 01 23:57:04 03Koen Kooi  07kergoth/autotools * r612eb226ee 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 23:57:07 03Jeremy Lainé  07kergoth/autotools * rb44ce1af2e 10openembedded.git/recipes/linux/ (11 files in 2 dirs): linux-2.6.29: add patches for boc01 Apr 01 23:57:07 stupid CIA notification script Apr 01 23:57:12 03Frans Meulenbroeks  07kergoth/autotools * rd67a64916e 10openembedded.git/recipes/linux/ (linux-omap-2.6.29/mmctiming.patch linux-omap_2.6.29.bb): linux-omap_2.6.29.bb: added mmctiming patch Apr 01 23:57:15 03woglinde  07kergoth/autotools * rfabea12c54 10openembedded.git/conf/checksums.ini: checksum.ini: add xine-1.1.16 remove older versions Apr 01 23:57:20 03Andrea Adami  07kergoth/autotools * r791a9bc3f0 10openembedded.git/conf/distro/angstrom-2008.1.conf: Apr 01 23:57:23 angstrom-2008.1.conf: add PREFERRED_VERSION_psplash Apr 01 23:57:25 -Acked-by: Koen Kooi Apr 01 23:57:27 03Stanislav Brabec  07kergoth/autotools * r95f33f6c42 10openembedded.git/recipes/gtk+/gtk+.inc: gtk+.inc: Package modules .debug, remove unusable modules .la files. Apr 01 23:57:30 03Stanislav Brabec  07kergoth/autotools * rb711f9d53e 10openembedded.git/recipes/gtk+/ (gtk+_2.14.2.bb gtk+_2.16.0.bb): gtk+: Replace also gail subpackages. Apr 01 23:57:37 03woglinde  07kergoth/autotools * r3d25ff174b 10openembedded.git/recipes/disko/ (disko_git.bb files/header.patch files/pkgconfig.patch): disko: add disko framework to openembedded Apr 01 23:57:38 * kergoth thinks he needs to talk to whoever is maintaining git.openembedded.net to fix it up Apr 01 23:57:40 03Stanislav Brabec  07kergoth/autotools * rdbaf04e7de 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 23:57:49 03woglinde  07kergoth/autotools * re59631d085 10openembedded.git/recipes/libxine/ (39 files in 6 dirs): Apr 01 23:57:52 libxine: remove older verison and add the actual 1.1.16 Apr 01 23:57:54 * dont seperate x11 and fb anymore xine can handle it via plugins Apr 01 23:57:56 * therefore all recipes wich depends on libxine-fb or libxine-x11 Apr 01 23:57:58 have to changed to libxine Apr 01 23:58:00 03Frans Meulenbroeks  07kergoth/autotools * ra2f2b1a1bc 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Apr 01 23:58:05 03Andrea Adami  07kergoth/autotools * r0c8dd633e2 10openembedded.git/conf/distro/angstrom-2008.1.conf: Apr 01 23:58:08 angstrom-2008.1.conf: fix typos Apr 01 23:58:09 * kergoth shakes head Apr 01 23:58:10 -golden rule: don't commit if you're tired... Apr 01 23:58:12 03Chris Larson  07kergoth/autotools * r813a40adbd 10openembedded.git/MAINTAINERS: Apr 01 23:58:16 MAINTAINERS: I can spell my own name. Really. Apr 01 23:58:18 Signed-off-by: Chris Larson Apr 01 23:58:22 03Chris Larson  07kergoth/autotools * rd27e47d2d1 10openembedded.git/MAINTAINERS: Apr 01 23:58:24 MAINTAINERS: I'll grab autoconf, automake, libtool, gnu-config, and the autoconf classes. Apr 01 23:58:26 (24 lines omitted) Apr 01 23:58:59 kergoth, re git: I think I'm understanding more. 'git branch' shows local branches and you only want to work on a local branch. 'git branch -r' shows all branches in the remote repo you cloned that you 'could' create a local branch of if you want? Apr 01 23:59:11 s/you cloned/you fetched/ Apr 01 23:59:18 you can add any number of remotes to a repository Apr 01 23:59:23 not necessarily just the upstream OpenEmbedded one Apr 01 23:59:29 but otherwise, yes Apr 01 23:59:29 so basically git-remote adds branches from remote reops - which you can add all kinds of various branches from various repos Apr 01 23:59:51 yep, it creates remote branches for all the branches at the upstream git repository Apr 02 00:00:06 ya... I was definitely confused by the simplification that git clone was doing for me Apr 02 00:00:44 so when you say 'track a branch' that simply means that you've done the equiv of a 'git-remote add' for it right? and what does tracking it do for you? all that means is it knows where to pull from for that branch right? Apr 02 00:00:59 nope, -cclean did'nt exactly solve the issue, so I'll rebuild the whole bunch then :) Apr 02 00:01:23 Aegiron, it may be that you have something setup wrong then Apr 02 00:02:43 that could always be the case, but it did work for other stuff in the overlay, so I expect that a total rebuild will solve things Apr 02 00:02:57 Aegiron, if I understand you correctly you want to override something in the base-files recipe, for example 'issue' - so you create your own collection (overlay) and create a recipes/base-files/base-files/ containing files right? Apr 02 00:03:24 tharvey: the remote branches are needed so you have a consistent environment, you always know what the upstream branches are at. they always track upstream. when you create a local branch with --track (i think thats default now), it sets it up so that the local branch tracks the parent branch, where "tracks" is defined as setting up the .git/config so that a git pull in that branch will merge from that parent. Apr 02 00:03:32 see git help branch, under --track Apr 02 00:04:05 so changes flow from upstream to the git "remote branch", which in turn gets merged or rebased to your local tracking branch, where you do your work Apr 02 00:05:04 * kergoth 's sure there's better ways of describing it, but .. Apr 02 00:05:07 tharvey: in almost, except that I am not creating new files, for the moment I just copy the full base-files dir to the overlay and then edit it there Apr 02 00:05:31 Aegiron: are you using bitbake's collections support to ensure it always prefers the versions from your overlay? Apr 02 00:05:39 jup Apr 02 00:05:43 cool Apr 02 00:06:01 Aegiron, I setup the same thing a while back (with an older oe snapshot) and at the time I found that I had to copy the entire recipe... not just provide the files used in that recipe Apr 02 00:06:06 if all you're doing is changing something that referenced via file://, you can do that without even needing an overlay, actually Apr 02 00:06:40 at the time I found that the 'local file fetcher' did not use collections if that makes sense... I believe thats been changed/fixed now? Apr 02 00:06:59 * tharvey waits for kergoth to explain Apr 02 00:07:02 its easier to manipulate filespath now, and file:// searches filespath Apr 02 00:07:10 yes, I know, but I started doing things like this a few days ago when qt embedded 4.5 came into the repo Apr 02 00:07:29 Hmmm Apr 02 00:07:37 Is there an overrides catch-all for all arm targets? Apr 02 00:07:41 kergoth, filespath? you mean BBPATH? Apr 02 00:07:44 no Apr 02 00:07:47 Or do I need to enumerate armv4...armv7a? Apr 02 00:07:53 FILESPATH is a : separated list of locations which file:// looks in Apr 02 00:08:10 and I thought it would be subject to changes, so I decided to keep my own changes and to not mind the repo's Apr 02 00:08:13 by default it looks under package name, package name - package version, files, etc under FILE_DIRNAME (the location of the recipe), and then under override specific dirs uner that Apr 02 00:08:20 ah yes... I recall now. Some time ago, FILESPATH was not iterated over within collections - is that not the case anymroe? Apr 02 00:08:56 if you mkdir build/files; FILESPATHBASE =. "${TOPDIR}/files:", then mkdir build/files/busybox; cp -a openembedded/recipes/busybox/busybox-foo/defconfig files/busybox/, it'll use the local one over the one int he recipe's location Apr 02 00:09:00 as an example Apr 02 00:09:15 (assuming TOPDIR is build/ in that example) Apr 02 00:11:54 kergoth, do you recall when that feature was added? thats what I was looking for when I originally snapshotted oe back in May 08. I recently sync'd up (couple of weeks back) Apr 02 00:12:02 using that would greatly clean up my overlay tree... Apr 02 00:12:05 i committed it a couple weeks ago Apr 02 00:12:07 :P Apr 02 00:12:12 nice Apr 02 00:12:22 commit ecfad70aae Apr 02 00:12:31 see the diff, it's pretty straightforward Apr 02 00:12:48 wow, it took me a while to figure out what exactly you meant Apr 02 00:13:25 ya, I'm not sure I understand it yet myself heh Apr 02 00:13:31 still swimming over the git info... Apr 02 00:13:40 * kergoth chuckles Apr 02 00:13:50 oh, you guys might like this collections thing i did Apr 02 00:13:56 might make things a bit simpler to set up Apr 02 00:14:16 check out the top two commits on the kergoth/master branch Apr 02 00:14:29 kergoth, have you had time to document this up in OE manual? it could really use some nice examples on collections and overlays Apr 02 00:14:35 but that only works when specifying different files for the same packages? Apr 02 00:14:44 what about editing the bb file itself? Apr 02 00:14:58 Aegiron, for that you want to use a collection Apr 02 00:15:05 you have to use collections for actual recipe modifications, unless you want to screw around with pn-foo overrides Apr 02 00:15:17 SRC_URI_append_pn-busybox = " file://foobar" Apr 02 00:15:18 :P Apr 02 00:15:26 but that could get ugly fast Apr 02 00:15:42 Updating the manual would be a good idea. Apr 02 00:15:51 I found that in my use of OE I had a lot of cases where I wanted to overide files:// - netbase (/etc/network/interfaces), basefiles (/etc/issue), etc etc - now I can do that without copying the entire recipe Apr 02 00:16:35 I'll give it a shot Apr 02 00:16:38 tharvey: my intent was to use it to make something like the kernel.bbclass's menuconfig task, but one that saves the new config into a build area files dir, so those changes would persist Apr 02 00:16:42 but i haven't finished that part yet Apr 02 00:17:07 03Tom Rini  07org.openembedded.dev * r1f33c1fb4d 10openembedded.git/recipes/ltp/ (ltp-20090131/no_epoll_create2_mips.patch ltp_20090131.bb): Apr 02 00:17:07 ltp: Fix building on ARM. Apr 02 00:17:07 The no_epoll_create2_mips.patch also applies for ARM, and probably Apr 02 00:17:07 a few others still (that I haven't tried). Rename to no_epoll_create2.patch Apr 02 00:17:09 and apply on arm/armeb. Apr 02 00:17:11 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * r55a31b3d72 10openembedded.git/packages/libogg/ (libtheora-1.0/libtheora-1.0-no-docs.patch libtheora_1.0.bb): Apr 02 00:17:14 Do not build libtheora-1.0 docs or examples Apr 02 00:17:16 since this will run a cross-compiled utility Apr 02 00:17:18 (vp3huff) on the host Apr 02 00:17:20 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * r8ef8f32bf4 10openembedded.git/packages/perl/perl-native_5.8.8.bb: Do not build perl package ODBM_File in perl-native, since it breaks the build Apr 02 00:17:23 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * rcd6e0709f2 10openembedded.git/conf/machine/ (6 files): Set default at91bootstrap version to 2.10 in at91 board files Apr 02 00:17:26 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * r0b91593c03 10openembedded.git/ (14 files in 7 dirs): Add support for busybox-1.13.3 Apr 02 00:17:33 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * r3979740c9d 10openembedded.git/packages/gpsd/gpsd.inc: Avoid parallel make for gpsd Apr 02 00:17:36 03Ulf Samuelsson  07refs/remotes/origin/ulf/linux * rab2fe0712a 10openembedded.git/packages/busybox/busybox_1.13.3.bb: Fix broken recipe for busybox-1.13.3 Apr 02 00:17:37 on that note... I think its time to head home and feed the kids - night Apr 02 00:18:01 the collections stuff, it lets you remove BBFILES and all the BBFILE_ stuff from your config, and just specify COLLECTIONS = "/my/overlay /path/to/openembedded" and it automatmically sets up all the rest based on it, assuming it's in priority order, highest to lowest Apr 02 00:20:30 i had it organised like this ftm: BBFILE_COLLECTIONS = "openembedded tapango" Apr 02 00:20:30 BBFILE_PATTERN_openembedded = "${OE_HOME}/org.openembedded.dev" Apr 02 00:20:30 BBFILE_PRIORITY_openembedded = "0" Apr 02 00:20:30 BBFILE_PATTERN_tapango = "${OE_HOME}/tapango.dev" Apr 02 00:20:31 BBFILE_PRIORITY_tapango = "1" Apr 02 00:20:47 * kergoth nods Apr 02 00:21:03 is that still valid, or can this get shorter as well? Apr 02 00:21:30 thats the best way right now, until i get this stuff I'm working on merged Apr 02 00:21:51 once this is merged, you'll be able to set that up with a single variable: COLLECTIONS = "${OE_HOME}/tapango.dev ${OE_HOME}/org.openembedded.dev" Apr 02 00:22:16 lol, just needed a dictionary to find out what 'nods' meant :p Apr 02 00:22:34 hehe Apr 02 00:23:31 kergoth, before I go... one more question: looks like your changes are not in org.embedded.dev? Apr 02 00:23:44 does the BBPATH env var still stand tho? Apr 02 00:23:50 the filespath stuff is there, the collections stuff is in kergoth/master for now Apr 02 00:23:50 regarding the pulling filespath out of base.bbclass Apr 02 00:24:00 tharvey: thats in org.openembedded.dev Apr 02 00:24:15 doh... I just switched to stable/2009 and its not there yet Apr 02 00:24:22 Aegiron: yeah, BBPATH isn't going away anytime soon, its needed for bitbake to find bitbake.conf, which everything else pulls in :) Apr 02 00:24:36 ok... really out now... thx Apr 02 00:24:52 erm, which pulls in everything else, i mean. Apr 02 00:24:59 god, its getting late in the day, i can't think anymore Apr 02 00:25:36 hehe, late, its already 2:25am here :) Apr 02 00:26:33 hmm, dammit Apr 02 00:28:57 why does bbpath have to point at the build dir as well as the org.oe.dev dir? Apr 02 00:29:43 that's how it finds your conf/local.conf Apr 02 00:29:46 it's not psychic :) Apr 02 00:29:51 the conf file is located in build/conf/local.conf and bbpath=${oedir}/build Apr 02 00:29:56 yes Apr 02 00:30:01 bitbake.conf includes conf/local.conf Apr 02 00:30:07 it searches BBPATH when a relative path is included/required Apr 02 00:30:23 aaah, ok, I see Apr 02 00:30:45 bitbake.conf is the one it cannot find Apr 02 00:30:59 i did oversee that one Apr 02 00:31:00 bitbake only directly loads bitbake.conf Apr 02 00:31:04 conf/bitbake.conf, to be exact Apr 02 00:31:08 bitbake.conf in OpenEmbedded loads all the others Apr 02 00:31:16 see the include & require lines in conf/bitbake.conf Apr 02 00:35:51 what the hell, I think I just screwed it all up :) Apr 02 00:36:39 the sanity checkerr is complaining again :p Apr 02 00:38:20 Hmm Apr 02 00:38:27 ok, my bad, typos in bbpath Apr 02 00:38:29 I assume adding upstream BB patches should still go past the ml, yeah? Apr 02 00:38:46 bitbake-dev@lists.berlios.de for bitbake patches Apr 02 00:40:01 kergoth: it seems I still need to specify bbfiles in local.conf even when using collections Apr 02 00:40:22 yes, you can only avoid it if you have the two commits i pointed you to on my branch Apr 02 00:40:31 :P Apr 02 00:42:31 oww, owkey, I overlooked that small part as well Apr 02 00:42:50 anyways, thanks a lot for the help, I'am of to bed now Apr 02 00:44:25 np Apr 02 01:00:26 Damnit Apr 02 01:00:48 Can anyone else confirm that ltp, when packaged with ipkg, actually installs everything from the archive? Apr 02 01:01:10 * kergoth hopes the sheevaplugs ship this month Apr 02 02:53:50 03Mike Westerhof  07org.openembedded.dev * r20aa856376 10openembedded.git/ (3 files in 3 dirs): Apr 02 02:53:50 linux-ixp4xx: add 2.6.29 kernel recipe, defconfig for NSLU2, and bump Apr 02 02:53:50 up the sane srcrev for the ixp4xx kernel patches from nslu2-linux.org Apr 02 02:57:06 anyone know if my collections.inc patches hit the OpenEmbedded devel list? was testing git send-email, not sure if it went through Apr 02 02:58:05 03Mike Westerhof  07org.openembedded.dev * r27d8a050c8 10openembedded.git/recipes/openssl/ (openssl-native_0.9.8g.bb openssl-native_0.9.8j.bb): Apr 02 02:58:05 openssl: make the default -native recipe match Apr 02 02:58:05 DEFAULT_PREFERENCE = -1 is removed from the -native recipes; this Apr 02 02:58:05 caused problems with tools linking against different versions of Apr 02 02:58:05 the openssl libs, resulting in obscure build failures (e.g. Apr 02 02:58:09 python-native "import md5" would fail). **** ENDING LOGGING AT Thu Apr 02 02:59:57 2009