**** BEGIN LOGGING AT Fri Mar 11 02:59:57 2011 Mar 11 03:46:22 good evening all Mar 11 03:46:54 have a quick ?: how oe packages are managed? Mar 11 03:47:13 I have a problem with one of them... Mar 11 04:05:22 hello anybody around? Mar 11 04:10:44 anybody can answer? Mar 11 04:21:57 how do I regenerate modules.dep, what stage does that? Mar 11 04:23:47 I guess whoever can takes care of the packet maintenance, am I right? Mar 11 04:24:17 IgorK1: hit the bugzilla Mar 11 04:35:07 duffolonius: can u give me direct link, please? Mar 11 04:58:58 HI all, I ran this command MACHINE=beagleboard ./oebb.sh bitbake virtual/kernel overnight, still it's at 0% can anyone see any problem in this? http://pastebin.com/9zkZM0ug Mar 11 06:02:38 ok, on this link http://bugs.openembedded.org/post_bug.cgi bug # 5531 has been filed. Mar 11 06:02:49 thanks Mar 11 06:04:09 IgorK1, what bug has been filed? Mar 11 06:10:00 nottoodumb: the bug is about building wxwidgets with openembedded. it produces wrong symbolic link to wx-config script. it links to the file outside the OE tree which does not exist Mar 11 06:10:25 you can go and look at the bug if you want Mar 11 06:29:43 was that the reason for my command to hang? Mar 11 06:30:00 I ran this command MACHINE=beagleboard ./oebb.sh bitbake virtual/kernel overnight, still it's at 0% can anyone see any problem in this? http://pastebin.com/9zkZM0ug Mar 11 06:43:42 any help for me http://pastebin.com/9DpbQQn7 i am executing this link, http://www.kernel-labs.org/files/openembedded-guide/openembedded-guide-3.html Mar 11 06:55:14 It may be the issue with certain python versions, iirc it was in Fedora 14. Try current BitBake master branch. Mar 11 06:56:23 The fix hasn't made it into a 1.12.x release yet. Mar 11 06:57:27 am i using 1.12.x? Mar 11 07:34:28 if the # of emails about eclipse plugins for OE not working can be considered a metric, seems like OE is experiencing an influx of new users. Mar 11 07:46:18 good morning Mar 11 07:46:26 kgilmer: good to know ;-) Mar 11 07:59:11 hi mckoan Mar 11 08:00:42 hi kgilmer how is at home in japan after last night? Mar 11 08:01:06 kgilmer: I've just read the news, I'm scared Mar 11 08:01:14 tidal waves are still happening. i'm on the other coast in nagasaki so it's ok here. Mar 11 08:01:34 yeah the tv news shows some incredible destruction. Mar 11 08:01:54 not since 9/11 have i seen such shocking live tv. Mar 11 08:01:56 kgilmer: I'm watching NHK just now Mar 11 08:03:51 kgilmer: yes, me too Mar 11 08:45:18 Hi, I'm getting an issue with running bitbake 1.13.0. Mar 11 08:45:20 ${@csl_get_main_version(d)} which triggered exception OSError: [Errno 13] Permission denied Mar 11 08:45:29 Couldn't find anything on google Mar 11 08:49:21 ping florian Mar 11 08:52:31 Does bitbake run as a different user or something? Mar 11 08:52:34 Ugh Mar 11 08:52:40 no Mar 11 08:55:03 Then I guess it's me Mar 11 08:56:35 Was that bug fixed in 1.12.0? Mar 11 08:56:47 I cannot get anything to work, seems like. Mar 11 08:56:51 what? Mar 11 09:01:20 hi jineld Mar 11 09:01:22 hi ant Mar 11 09:01:46 1.12 is buggy for me, can't build a thing Mar 11 09:02:01 Function 'staging_helper' failed (see /dev/null for further information) Mar 11 09:02:39 1.13 is buggy for me too, it hangs during the initial parsing phase Mar 11 09:02:41 :/ Mar 11 09:03:38 I use bitbake 1.10.2 Mar 11 09:05:42 Heh, I'll probably have to revert to that Mar 11 09:06:04 B_Lizzard hu? Mar 11 09:06:09 here all works Mar 11 09:06:18 B_Lizzard whats your host python version? Mar 11 09:07:02 2.6.6 Mar 11 09:07:08 On Ubuntu Maverick Mar 11 09:07:20 okay that should be high enough Mar 11 09:08:14 I think the permission denied should be a red flag for my system or something though Mar 11 09:08:15 :/ Mar 11 09:08:20 B_Lizzard can you pastebin the complete error? Mar 11 09:10:13 The permission denied thing? Mar 11 09:10:15 Sure Mar 11 09:10:46 complete bitbake output Mar 11 09:10:49 after parsing Mar 11 09:10:58 or where it stops Mar 11 09:11:25 Ah, I don't have verbose output enabled, if that's possible Mar 11 09:11:40 It just hangs with the progress bar thingie Mar 11 09:12:20 hangs? Mar 11 09:12:25 Yeah, stops Mar 11 09:12:39 It responds to ^C Mar 11 09:12:47 But it just sits there Mar 11 09:12:55 ETA not updating and whatnot Mar 11 09:15:27 I just tested with a new tmp folder, and it completed the recipe parsing thing but it failed again with the staging error Mar 11 09:15:42 At least not in /dev/null this time Mar 11 09:15:51 *sigh* Mar 11 09:16:12 okay I see you dont want any help Mar 11 09:16:30 Well, having a newer bitbake would be wonderful Mar 11 09:16:43 But it doesn't matter much Mar 11 09:17:06 I only updated because I had an issue with a recipe and I'm pretty sure it's wasn't bitbake in the first place Mar 11 09:17:09 Ugh Mar 11 09:17:41 I guess I'm the only one getting these things Mar 11 09:40:11 is there a way to double check where oe takes $SERIAL_CONSOLE from it seems there are some sources for that variable ? Mar 11 09:40:26 machine conf Mar 11 09:40:47 well i change it theere but the changes are not taken .. so i assume there is a 2nd place Mar 11 09:41:14 not that I know of Mar 11 09:41:16 try grep Mar 11 09:41:20 ah Mar 11 09:41:26 found it .. stupid me Mar 11 09:41:56 can i do this in my local.conf ? Mar 11 09:42:24 look if its hardcoded Mar 11 09:42:31 = is =? not Mar 11 09:42:56 SERIAL_CONSOLE = "115200 ttyS2" is what the machine conf does Mar 11 09:43:09 when i do the same in my local.conf .. will it be overwritten ? Mar 11 09:43:17 no Mar 11 09:43:20 = is hard Mar 11 09:43:24 ah Mar 11 09:43:37 so i use =? Mar 11 09:43:39 ? Mar 11 09:43:58 or do i need += Mar 11 10:01:45 use ?= in machine config Mar 11 10:02:19 bitbake loads local.conf then machine conf then distro conf Mar 11 10:02:26 each of them can change variables Mar 11 10:02:53 so user can choose machine and distro in local.conf, then machine sets some vars which distro config can change Mar 11 10:04:47 gm hrw Mar 11 10:10:56 Jay7, you awake yet? Mar 11 10:14:05 ka6sox-away: I'm online :) Mar 11 10:14:10 morning :) Mar 11 10:14:18 hi jay7 Mar 11 10:18:53 morning...not much sleep here tonight...waiting for the wave Mar 11 10:20:36 ka6sox-away: japan or hawai? Mar 11 10:20:50 Hawaii Mar 11 10:21:41 about 20 people headed to my Sister's Mar 11 10:38:20 ka6sox-away, where does your sis live? Mar 11 10:39:55 Waikaloa Mar 11 11:01:22 ka6sox-away, that's still in hawaii.. Mar 11 11:01:52 uh huh... Mar 11 11:02:05 but she is "up country" Mar 11 11:02:05 ka6sox-away, is oe really easy to use than make cross compile? i am scared of oe..it looks scary to me Mar 11 11:02:10 oh..ok Mar 11 11:02:17 what about you? how is your place? Mar 11 11:02:28 I'll know in 5yrs Mar 11 11:02:32 er 5hrs Mar 11 11:02:46 why? Mar 11 11:02:54 thats when it gets here. Mar 11 11:03:10 be brave..things seems to be ok now Mar 11 11:04:02 what's time for you now? Mar 11 11:04:16 0300 Mar 11 11:04:57 oh..ok Mar 11 11:09:25 ka6sox-away, whom do you stay with? Mar 11 11:11:19 NotTooDumb3, bbiaf..gotta deal with things here Mar 11 11:11:31 sure.... Mar 11 12:20:37 Hi guys Mar 11 12:21:23 is it possible to build policykit package without libpam? Mar 11 12:22:07 i used --without-libpam option in configure script but policykit still reports an error on -lpam Mar 11 12:28:41 any suggestion to my problem? Mar 11 13:08:23 03Koen Kooi  07org.openembedded.dev * r42b0a654ed 10openembedded.git/conf/machine/beagleboard.conf: Mar 11 13:08:23 beagleboard: switch to linux-omap Mar 11 13:08:23 * fix serial console for new kernel Mar 11 13:08:23 Signed-off-by: Koen Kooi Mar 11 13:08:26 03Koen Kooi  07org.openembedded.dev * raa5ec12bc4 10openembedded.git/recipes/sysvinit/sysvinit_2.86.bb: Mar 11 13:08:26 sysvinit: bump PR for beagleboard serial console change Mar 11 13:08:26 * a new inittab is needed, so bump PR to regenerate it Mar 11 13:08:26 Signed-off-by: Koen Kooi Mar 11 13:08:27 03Koen Kooi  07org.openembedded.dev * rf77bbb6a16 10openembedded.git/conf/machine/include/omap3.inc: Mar 11 13:08:27 omap3.inc: bump MACHINE_KERNEL_PR for beagleboard changes Mar 11 13:08:28 Signed-off-by: Koen Kooi Mar 11 13:08:30 03Koen Kooi  07org.openembedded.dev * r67a7c91a4e 10openembedded.git/conf/distro/ (angstrom-2008.1-legacy.conf angstrom-2008.1.conf): Mar 11 13:08:30 angstrom 2008.1: remove from .dev Mar 11 13:08:30 This will be maintained in the 2011.03-maintenance branch Mar 11 13:08:30 Signed-off-by: Koen Kooi Mar 11 13:08:43 03Koen Kooi  07org.openembedded.dev * rc8e096f94f 10openembedded.git/recipes/linux/ (5 files in 2 dirs): Mar 11 13:08:44 linux-omap 2.6.37: add support for beagleboard xM rev C Mar 11 13:08:44 Signed-off-by: Koen Kooi Mar 11 13:08:44 03Koen Kooi  07org.openembedded.dev * rd95d1126c0 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Mar 11 13:08:45 linux-omap 2.6.37: add patch for mmc timeout issues Mar 11 13:08:45 Signed-off-by: Koen Kooi Mar 11 13:08:46 03Koen Kooi  07org.openembedded.dev * re4c308b553 10openembedded.git/recipes/u-boot/ (64 files in 2 dirs): Mar 11 13:08:46 u-boot git: update beagleboard u-boot to 2011.03-RC Mar 11 13:08:46 Signed-off-by: Koen Kooi Mar 11 13:53:00 03Graeme Gregory  07master * r0d40ad3a8e 10openembedded.git/recipes/transmission/transmission_2.22.bb: Mar 11 13:53:00 transmission_2.22.bb : fix nls issue Mar 11 13:53:00 Inherit gettext as transmission has choice of whether to nls or not. Mar 11 13:53:00 Signed-off-by: Graeme Gregory Mar 11 13:53:13 03Lukas-David Gorris  07master * rbaaea72df4 10openembedded.git/recipes/libevent/libevent_2.0.10-stable.bb: (log message trimmed) Mar 11 13:53:13 libevent: recipe for 2.0.10-stable version Mar 11 13:53:13 > Your patches have been corrupted by MTA, please send as attachments. Mar 11 13:53:13 > Mar 11 13:53:14 > The transmission ones can have Mar 11 13:53:14 > Mar 11 13:53:15 > Acked-by: Graeme Gregory Mar 11 13:53:15 03Lukas-David Gorris  07master * re124c44f14 10openembedded.git/recipes/transmission/transmission_2.22.bb: (log message trimmed) Mar 11 13:53:16 transmission: add recipe for 2.22 version Mar 11 13:53:16 > Depends on the the previously sent libevent bump. Mar 11 13:53:17 > Mar 11 13:53:17 > From: Lukas-David Gorris Mar 11 13:53:18 > Date: Fri, 11 Mar 2011 02:00:56 +0100 Mar 11 13:53:18 > Subject: [oe] [PATCH] transmission: add recipe for 2.22 version Mar 11 13:59:49 who can grant patchwork privs? the old admin Crofton doesnt have privs on the new setup Mar 11 14:14:05 03Marco Cavallini  07master * rf91bb01cb3 10openembedded.git/conf/distro/include/kaeilos.inc: kaeilos.inc: packages upgrade Mar 11 14:14:06 03Marco Cavallini  07master * r52c6a50327 10openembedded.git/conf/distro/include/kaeilos-2009-preferred-versions.inc: kaeilos-2009-preferred-versions.inc: packages upgrade Mar 11 14:14:41 * mwester-laptop looks at the CIA-40 messages about "Your patches have been corrupted by MTA" and is a little concerned. Mar 11 14:18:29 mwester-laptop: side effect of using patchwork Mar 11 14:18:40 ask khem/koen about it Mar 11 14:18:52 I think, that khem was setting the new one Mar 11 14:30:36 morning Mar 11 14:33:02 otavio: ping Mar 11 14:33:09 hey kergoth Mar 11 14:33:19 hey Mar 11 14:33:33 kergoth_: catch you again :) Mar 11 14:33:42 * kergoth_ hides Mar 11 14:33:50 yesterday I had too much drinking, thus bb benchmarking was postposed... Mar 11 14:34:24 ant_work: hello Mar 11 14:34:28 kergoth_: have you any opinion why it is impossible to override DEPLOY_DIR or even better DEPLOY_DIR_IMAGE Mar 11 14:34:43 fa2nge Mar 11 14:34:54 kergoth_: or may be I just can't find a way Mar 11 14:35:05 otavio: pls unify your xz patch and my lzma request in a single patch Mar 11 14:35:30 ant_work: mind to point it to me? Mar 11 14:36:02 add XZ_INTEGRITY_CHECK variable -> cpio.lzma compression level in bitbake.conf Mar 11 14:36:19 while here, pls add xz compression level too Mar 11 14:36:27 Jay7: you can override them, but DEPLOY_DIR can't be moved outside of TMPDIR, because of how pstage is implemented. pstage creates ipk files relative to TMPDIR and installs the ipk files into TMPDIR as though it was a rootfs, so everything that goes into a pstage package must be relative to tmpdir Mar 11 14:36:51 otavio: this stuff needs to be overridden by machine.conf Mar 11 14:36:59 i don't see a reason that DEPLOY_DIR_IMAGE wouldn't be more open to modification, what's the behavior? Mar 11 14:37:06 kergoth_: I'm just looking for a way to get path to images built w/o any guessing and fiddling :) Mar 11 14:37:42 yeah.. ideally the deploy dir would be outside of tmpdir by *default* imo. that's build *output*, not *intermediate files* Mar 11 14:37:45 kergoth_: behaviour is simple - it's not affected by modification from local.conf Mar 11 14:37:46 but.. Mar 11 14:38:40 a quick grep will tell you why that's the case, i'm sure. someone must be overriding you Mar 11 14:38:44 * kergoth_ checks Mar 11 14:39:01 right, the distro files are setting it. Mar 11 14:39:06 use an override. Mar 11 14:39:15 DEPLOY_DIR_IMAGE_local = "${TOPDIR}/images" Mar 11 14:39:22 er Mar 11 14:39:26 DEPLOY_DIR_IMAGE_local = "${TOPDIR}/images/${MACHINE}" Mar 11 14:39:28 or whatever Mar 11 14:39:36 ah, nice Mar 11 14:39:40 thanks :) Mar 11 14:39:42 np Mar 11 14:40:00 (that's always available.. i use that override pretty often when a distro/machien file isn't using ?=) Mar 11 14:41:54 too much ways to assign vars :) Mar 11 14:42:02 not sure this is good but it works Mar 11 14:43:32 good :) Mar 11 14:44:45 ynezz: yeah, be ready for next release ;) Mar 11 14:45:15 may be even with qemu runtime testing Mar 11 14:45:19 Jay7: the problem is, a distro can't use ?= if a variable is set in bitbake.conf even if they want to, because it will never be applied, because the var was always set, just not always *by the user* Mar 11 14:45:22 Jay7: so.. Mar 11 14:45:38 Ah, that explains a lot. Mar 11 14:45:58 ?= just means "set if the variable is unset" Mar 11 14:46:01 yep, read the order of includes in bitbake.conf Mar 11 14:46:04 kergoth_: may be local.conf just should be processed after distro/machine configs? Mar 11 14:46:15 if not already :) Mar 11 14:46:16 kergoth there is a bug with tar_stash and downloading Mar 11 14:46:34 when I cancel the fetch via crtl + c Mar 11 14:46:50 (or the fetch timeouts?) Mar 11 14:46:52 the md5 sum is generated from what was downloaded Mar 11 14:47:22 and do_fetch is marked as done Mar 11 14:47:24 Jay7: not possible. Mar 11 14:47:29 local.conf *sets* DISTRO/MACHINE Mar 11 14:47:35 bitbake isn't a time machine ;) Mar 11 14:47:37 kergoth_: ah, sure Mar 11 14:47:42 its theoretically possible in the long term Mar 11 14:47:43 :) Mar 11 14:47:47 by leveraging our ast nodes Mar 11 14:47:52 see what nodes set what, rearrange things Mar 11 14:47:56 but not today Mar 11 14:48:10 well, anyway, overriding is working Mar 11 14:48:15 it's enough for me :) Mar 11 14:48:28 the problem is of course that DISTRO could theoretically reference some other variable which was set in the distro.conf.. gets awfully circular :) Mar 11 14:48:37 * kergoth_ nods Mar 11 14:48:45 _local was always intended to be the override equivalent of local.conf Mar 11 14:48:53 ant_work: I will take a look into that and post a patch for you to take look, ok? Mar 11 14:49:00 if you think of our metadata as layered, those are just two different mechanisms of implementing the layering Mar 11 14:49:10 one via include file ordering, one via overrides processing Mar 11 14:49:17 both from least specific to most specific Mar 11 14:49:26 (with the distro/machine thing the exception for technical reasons) Mar 11 14:49:28 * Jay7 can't imagine bitbake internals even :) Mar 11 14:49:40 it's not about bitbake's internals so much as the design Mar 11 14:50:10 the idea was that our metadata was bound to have both more general information and more specific information, and it was important that machine info be used in preference to architecture info, in preference to global info Mar 11 14:50:35 but it was clear that we needed both conditionals for recipes and the like *and* the same layering in the configuration, hence having both Mar 11 14:50:48 i think it makes it slightly less confusing if you think of it in this way Mar 11 14:51:05 which probably isn't saying much, really.. Mar 11 14:51:10 but still :) Mar 11 14:51:37 kergoth_: btw, what is your opinion on runtime testing implementation? should it be bitbake plugin as Yocto have or better to have it separated? Mar 11 14:52:10 haven't looked into how yocto does it at all yet, have a link? Mar 11 14:52:17 i know they do, but haven't checked it out Mar 11 14:52:29 haven't under hands right now Mar 11 14:52:44 it is bbclass in yocto's metadata Mar 11 14:52:50 runtime testing? Mar 11 14:52:52 ah, okay Mar 11 14:52:58 ynezz: yes, with qemu Mar 11 14:53:05 ah Mar 11 14:53:25 they are starting qemu after image built and do some tests Mar 11 14:53:44 e.g. scp file into VM, run something, etc Mar 11 14:53:45 good Mar 11 14:55:14 * Jay7 is trying to create kde's plasma workspace for OE things Mar 11 14:55:53 * kergoth_ wonders if pycharm is unborked on his linux box yet Mar 11 14:56:15 whats pycharm? Mar 11 14:56:21 magic Mar 11 14:56:25 hehe Mar 11 14:56:50 it's a python ide, actually pretty nice. i don't use it much at all yet, but want to try it more Mar 11 14:57:11 can highlight non-portable python constructs given a range of versions you want to support, etc Mar 11 14:57:29 03Tom Rini  07master * r01e463f247 10openembedded.git/conf/distro/include/kaeilos-2009-preferred-versions.inc: Mar 11 14:57:29 kaeilos-2009-preferred-versions.inc: Lock policykit Mar 11 14:57:29 glib-2.0 is locked to a version that will not work with policykit Mar 11 14:57:29 newer than 0.97. Mar 11 14:57:29 Signed-off-by: Tom Rini Mar 11 14:57:30 03Tom Rini  07master * r64204cdb62 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: Mar 11 14:57:30 angstrom-2008-preferred-versions.inc: Lock policykit Mar 11 14:57:31 glib-2.0 is locked to a version that will not work with policykit Mar 11 14:57:31 newer than 0.97. Mar 11 14:57:32 Acked-by: Koen Kooi Mar 11 14:57:32 Signed-off-by: Tom Rini Mar 11 14:57:35 03Tom Rini  07master * rbd8768ee3d 10openembedded.git/recipes/policykit/ (policykit-gnome_0.100.bb policykit_0.100.bb): Mar 11 14:57:35 policykit / policykit-gnome: Add 0.100 Mar 11 14:57:35 This version supports shadow or pam auth so we key that off of the Mar 11 14:57:35 pam DISTRO_FEATURE. This version is default now that distros Mar 11 14:57:35 which wouldn't be able to build this due to dependencies lock down Mar 11 15:01:23 kergoth_, funny retweet Mar 11 15:01:30 I ahve often wondered the same thing Mar 11 15:01:34 he crofton Mar 11 15:02:04 * kergoth_ too :) Mar 11 15:02:15 if i were an eye doctor, i sure as hell would Mar 11 15:02:21 have to find some entertainment somewhere Mar 11 15:02:23 hehe Mar 11 15:05:00 * kergoth_ mutters Mar 11 15:09:45 another useful thing is possibility to suspend bitbake build.. Mar 11 15:09:58 what do you mean? Mar 11 15:10:00 now I'm doing this via lxc-freeze/lxc-unfreeze Mar 11 15:10:06 ah Mar 11 15:10:13 but this suspending/resuming container Mar 11 15:10:15 maybe SIGSTOP/SIGCONT? Mar 11 15:10:21 i don't know how it would affect it Mar 11 15:10:22 hmm Mar 11 15:10:24 * kergoth_ ponders Mar 11 15:10:53 kergoth_: it should be sent to every process started by bitbake Mar 11 15:11:07 not sure they are all in same process group Mar 11 15:11:16 yeah, i was just thinking about that :) Mar 11 15:11:29 anyway, good idea to try one time :) Mar 11 15:11:33 worst case could leverage cgroups or something Mar 11 15:11:34 i dunno Mar 11 15:11:37 yeah, something to think about for the future Mar 11 15:11:54 (we need a list of these things somewhere) Mar 11 15:12:09 it's useful on desktop as well Mar 11 15:12:15 on workstation even Mar 11 15:15:27 hrw: xfce is nice, but i still <3 tiled window managers :D Mar 11 15:20:00 Is someone available to review and apply the patches I sent some time ago? Please skip patch 4 since it is going to be updated with ant_work request. Mar 11 15:20:56 is patchwork up to date? Mar 11 15:21:20 otavio: I looked at them earlier but didnt really understand them enough Mar 11 15:23:07 XorA: I can clarify any of them if required Mar 11 15:23:28 Tartarus: have you asked it to me? Mar 11 15:23:41 otavio, hm? Mar 11 15:23:48 Tartarus: about the patchwork Mar 11 15:23:56 I'm asking, are all of your patches listed in patchwork and ones that you don't want applied marked correctly? Mar 11 15:24:59 Tartarus: Just looked at patchwork and no patches from me seems there. Mar 11 15:25:27 http://patches.openembedded.org/project/oe/list/ Mar 11 15:25:33 it moved address Mar 11 15:25:38 ahhhhhhhhhhhhhh Mar 11 15:26:01 03Tom Rini  07master * r60ef5a3b2d 10openembedded.git/recipes/guile/guile-native.inc: Mar 11 15:26:02 guile-native: Reorder inherits to fix gettext Mar 11 15:26:02 Signed-off-by: Tom Rini Mar 11 15:26:14 03Tom Rini  07master * r2a9a834ce4 10openembedded.git/classes/gettext.bbclass: Mar 11 15:26:14 gettext.bbclass: Add virtclass DEPENDS prepends Mar 11 15:26:14 With BBCLASSEXTEND recipes which also modify DEPENDS_virtclass-native Mar 11 15:26:14 gettext.bbclass is otherwise not working and gettext-native isn't Mar 11 15:26:15 added. So we need to have a DEPENDS_virtclass-$foo_prepend to Mar 11 15:26:15 make sure GETTEXT_DEPENDS gets in. Mar 11 15:26:16 Signed-off-by: Tom Rini Mar 11 15:26:16 03Tom Rini  07master * r9537596e06 10openembedded.git/recipes/classpath/classpath-native.inc: Mar 11 15:26:17 classpath-native: Reorder inherits to fix gettext Mar 11 15:26:17 Signed-off-by: Tom Rini Mar 11 15:28:27 XorA: are the users imported? Mar 11 15:28:30 otavio: probably means you dont have the permissions to edit the status either Mar 11 15:28:34 Tartarus: that commit message was quite confusing. of course, i can't think of a good way to describe it offhand, but .. i really hope no one tries to go back and figure out why we're setting those Mar 11 15:28:35 hehe Mar 11 15:28:35 otavio: it seems they werent Mar 11 15:28:38 XorA: hehe Mar 11 15:28:42 so ... Mar 11 15:28:48 Tartarus: so they're not updated heheh Mar 11 15:28:54 otavio: I had to create new account this morning Mar 11 15:29:35 XorA: ok Mar 11 15:29:41 * otavio created 'otavio' account. Mar 11 15:29:52 Waiting for 'rights' Mar 11 15:29:56 * XorA doesnt actually know much about it, waiting for khem to wake up Mar 11 15:31:34 kergoth_: I have to choose one of tiled one someday Mar 11 15:32:35 XorA, when you talk with khem ahve him bump rights on Crofton Mar 11 15:33:34 * XorA wouldnt trust that Crofton dude ;-) Mar 11 15:34:30 Tartarus: OK. I updated them. The ones that are there seem fine. Please take a look. Mar 11 15:34:58 * otavio out to lunch Mar 11 15:36:33 otavio, depending on how the weekend goes here i'm going to try and do a patchwork weekend again Mar 11 15:49:24 * pb__ stabs gstreamermm Mar 11 15:49:54 and glibmm Mar 11 15:50:19 I think it was a bad idea to try to write this program in C++. Mar 11 15:50:31 hm.. Mar 11 15:50:39 * Jay7 is reading mail about bugzilla Mar 11 15:51:08 is there any tracker which have something like patchwork and issue tracker integrated? Mar 11 15:52:18 not that I can think of, though it wouldn't be all that hard to graft patchwork functionality onto bugzilla Mar 11 15:54:17 have new patches automatically filed as "enhancement" tickets or some such, doesn't sound like such a bad idea to me. Mar 11 15:54:50 pb__: that was my thought, + if you actually mentioned BUG #XXXX it could file it under that bug # Mar 11 15:55:04 yeah, that would be neat Mar 11 15:55:13 we just need a coder now :-) Mar 11 15:56:15 seems to be a feature all commercial bug/ticketing systems have, but none of the FOSS ones Mar 11 15:56:29 I guess I could probably write the code for that. I'll have a bit of a think about it. Mar 11 15:56:32 yeah, indeed Mar 11 15:56:47 How about JIRA? Mar 11 15:56:57 We could probably qualify for one of the free licenses Mar 11 15:57:48 * XorA uses JIRA in another job Mar 11 15:58:11 * Tartarus uses JIRA at work and for Jenkins and filing boxee box issues Mar 11 15:58:57 Might be worth a look. I've never used it. Mar 11 16:00:56 There's not a patchwork plugin yet, but I'm sure someone could whip one up once we figured out how we wanted it to work Mar 11 16:01:21 (since there's already a lot of "key mentioned in other resource HERE, so post to the issue like this..." Mar 11 16:03:13 Tartarus: it already has mail box support, so possibly just need patchwork to do correct munging then forward the email Mar 11 16:03:44 yeah Mar 11 16:07:28 * Jay7 have quick look over Trac Mar 11 16:08:05 it have integration with Bitten (thing like Hudson/Buildbot) Mar 11 16:08:20 but have no way to create tickets via email Mar 11 16:08:24 I used to admin trac, its horrible to maintain Mar 11 16:08:35 yeah, have heard that Mar 11 16:09:02 starts off all cool, but then turns out all the plugin maintainers like to used all the uber new features of latest Python +1 so really hard to support on servers Mar 11 16:09:58 just place it into jail alone! :) Mar 11 16:10:23 I did consider that, but then that becuase yet another server to maintain Mar 11 16:10:23 well.. I can do some review wrt TTS Mar 11 16:10:56 yeah, I used trac before as well and I was not very keen on it. Mar 11 16:11:53 redmine seems good but this is ruby Mar 11 16:12:00 another language to know :) Mar 11 16:15:19 http://git.collabora.co.uk/?p=user/cassidy/git-bugzilla Mar 11 16:16:50 looks like might be useful to use to knock something together, basically a git-send-email replacement Mar 11 16:18:24 Jay7, I use redmine...it okay for what it is... Mar 11 16:18:48 Jay7, what do you need it to do? Mar 11 16:19:02 ka6sox-away: read OE ML ;) Mar 11 16:19:08 wrt bugzilla Mar 11 16:19:22 oh, not a bugzilla fan Mar 11 16:19:29 hehe Mar 11 16:20:37 * XorA is quite happy to work with bugzilla if we can come up with a OE workflow that works Mar 11 16:20:53 reason bugzilla ended up dead was the workflow didnt work obviously Mar 11 16:21:17 XorA, understood...like any tool it takes maintainence and it fell out of use. Mar 11 16:21:26 and its essentially unmaintained. Mar 11 16:21:39 ka6sox-away: people really liked the patchwork way I think Mar 11 16:22:18 yes, and as Stefan mentioned its out of the workflow... Mar 11 16:22:31 (or was that you? Mar 11 16:22:32 ) Mar 11 16:22:35 anyways... Mar 11 16:23:07 I use redmine (but not in the way that it was intended) and for our needs its great Mar 11 16:25:11 its really a question of how we want folks to report "bugs" and whether if they do anyone will actually SEE it. Mar 11 16:28:29 Tartarus: because email is running so slow, admining JIRA is a doddle, I have two on my servers Mar 11 16:35:32 yeah, it does look like something is a bit wrong with the mailing list at the moment. I'll have a look at that when I get home. Mar 11 16:35:41 it seems to be delaying mail by about 20 mins or so for some reason. Mar 11 16:59:09 https://spreadsheets.google.com/ccc?key=0AhY8lmfkCabTdC1NdzJlQnpWYU9BcnFlN2tKU0taNlE&authkey=CPG4l4QJ Mar 11 16:59:22 short comparison of TTS Mar 11 17:00:20 (excerpted from wikipedia) Mar 11 17:03:34 * Jay7 afk Mar 11 17:23:41 Jay7, very interesting Mar 11 17:23:52 that needs to be on our wiki as a link Mar 11 17:25:12 ka6sox-surfin, how is your weather now? Mar 11 17:25:51 well..i'm still standing Mar 11 17:26:42 the wave is now in Las Vegas so we are good. Mar 11 18:04:46 hi obi Mar 11 19:17:37 ouch http://fpaste.org/8RWM Mar 11 19:19:57 isn't it related to that recent gettext breakage? Mar 11 19:21:37 er Mar 11 19:21:55 your metadata revision isn't uptream Mar 11 19:22:20 ok, I'll reproduce it with upstream Mar 11 19:23:14 or at least let us know what the upstream has you started with is Mar 11 19:23:15 *hash Mar 11 19:23:25 rebased few mins ago Mar 11 19:24:12 * Tartarus tries angstrom 2010.x + antgw100 + uclibc Mar 11 19:24:26 http://fpaste.org/bdGH Mar 11 19:25:37 as you can see, just few patches on top of OE Mar 11 19:25:59 yeah Mar 11 19:26:01 it should be fine Mar 11 19:26:20 um Mar 11 19:26:34 lets try console-image Mar 11 19:27:03 there we go, sigh Mar 11 19:27:47 so you can reproduce it? Mar 11 19:31:14 yeah Mar 11 19:34:46 Well, this is crazy Mar 11 19:34:49 But, I think I see Mar 11 19:36:17 good, because I didn't :) Mar 11 19:38:14 how on earth this worked before I do not know Mar 11 19:38:32 libiconv.inc inherit's gettext Mar 11 19:38:40 all of the real recipes say "Do this, we can't inherit gettext" Mar 11 19:39:52 * mwester-laptop gets the idea that OE metadata has entered a brief period of instability ;) Mar 11 19:40:24 03Tom Rini  07master * r4cd4e55933 10openembedded.git/recipes/libiconv/ (libiconv.inc libiconv_1.12.bb libiconv_1.13.1.bb): (log message trimmed) Mar 11 19:40:24 libiconv: Clean up Mar 11 19:40:24 - Since we have a big comment about not being able to use gettext.bbclass Mar 11 19:40:24 drop it. Mar 11 19:40:24 - Use $GNU_MIRROR Mar 11 19:40:24 - Switch to INC_PR, move bits from the two that used the inc file to Mar 11 19:40:24 the incfile itself. Mar 11 19:40:36 * Tartarus is currently out of brown paper bags but appreciates the effort :) Mar 11 19:43:21 Tartarus: thanks! Mar 11 19:43:32 let's see if it builds something Mar 11 19:43:41 console-image stopped dep-looping Mar 11 19:43:51 and i -c patch'd libiconv since I dropped the i swear not needed S= Mar 11 19:44:45 can't parse that last line :) Mar 11 20:50:25 ka6sox-surfin: surfing? Mar 11 20:50:39 hawaii Mar 11 21:09:42 Nice, huh? Mar 11 21:14:23 how to define environmental variables in our recipes Mar 11 21:16:27 export MOO Mar 11 21:18:26 woglinde : so it is same as setting the environment in a shell script Mar 11 21:18:41 yes Mar 11 21:18:45 thnaks Mar 11 21:18:45 mostly Mar 11 21:18:52 sreddy whats your context? Mar 11 21:19:03 within the recipe Mar 11 21:23:36 woglinde: yes it is within recipe Mar 11 21:23:42 and for compilation Mar 11 21:24:15 sreddy ah you can could set it in do_compile only too Mar 11 21:24:45 hhmm that would be more helpful Mar 11 21:25:12 sreddy grep after do_compile Mar 11 21:25:28 oe_runmake FOO=moo Mar 11 21:26:52 got it Mar 11 21:29:30 or is it autotools? Mar 11 21:29:40 can you pastebin the recipe? Mar 11 21:29:58 i have yet to create it Mar 11 21:31:02 i can it is defined in some of the recipes using export yet to find how to define in a do_compile () Mar 11 22:20:00 ericben_: around ? Mar 11 22:20:23 khem: hi Mar 11 22:20:30 ericben_: Can you try to send a patch to ml and see if it appears in pw now ? Mar 11 22:20:42 ok I resend the last one Mar 11 22:20:55 lets wait for 5 mins now :) Mar 11 22:21:50 http://patches.openembedded.org/ -> The requested URL / was not found on this server. Mar 11 22:32:07 ericben_: uff fixed Mar 11 22:32:17 but it might have missed your patch again Mar 11 22:33:07 khem: ok I send an other one Mar 11 22:33:10 ericben_: send it once more. hmm may be I can take one of your mbox Mar 11 22:33:18 and try it locally Mar 11 22:34:53 hm.. iirc OE or Yocto/Poky have bbclass which pastebin's faillog Mar 11 22:35:09 seems we should resurrect/import it for now Mar 11 22:35:26 before we got 'normal' stats collector Mar 11 22:35:30 Jay7: or simply we could apply the patch I sent and resurect oestat ! Mar 11 22:35:38 ericben_: or this Mar 11 22:35:59 ericben_: tinderbox is too intrusive Mar 11 22:36:04 someone eating my cpu Mar 11 22:36:27 hm.. it's Xorg Mar 11 22:36:28 khem: I sent a patch which only use tinderbox in case of failure Mar 11 22:36:42 ericben_: yes I remember Mar 11 22:36:53 which is better than no feedback at all Mar 11 22:37:15 ericben_: I will ask ka6sox-away when he is not away Mar 11 22:37:21 if he can deploy it Mar 11 22:37:48 khem: the patch is in patchwork Mar 11 22:37:52 yippie http://patches.openembedded.org/patch/1303/ Mar 11 22:38:10 so now all my utf-8 friends will be happier Mar 11 22:38:19 yes :) Mar 11 22:38:47 what is strange is that is translates the énard to ENARD Mar 11 22:39:01 but that's not a problem while the patch makes it into patchwork Mar 11 22:46:20 ericben_: only problem with that patch - it should be applied everywhere Mar 11 22:46:29 or we should change server-side Mar 11 22:46:41 well.. we can pass 'version' field Mar 11 22:47:01 it should be easy to add version check on server side Mar 11 22:48:42 Jay7: well if we don't start by pushing it ... Mar 11 22:48:43 :) Mar 11 22:49:55 Crofton|work, howdy Mar 11 22:50:05 hi kgilmer Mar 11 22:50:12 hi woglinde :) Mar 11 22:50:39 i hear from stefan things are already proceeding very well, good to hear. Mar 11 22:51:23 ;) Mar 11 22:52:14 ericben_: another way is change url Mar 11 22:52:18 that may be easier Mar 11 22:52:53 Jay7: yes patch + URL, good idea Mar 11 22:53:16 so that we start with a new setup of oestat with an empty database Mar 11 22:54:08 yep, right Mar 11 22:54:12 ka6sox-away: ping Mar 11 23:05:05 hi all, anyone building against oe-core already? Mar 11 23:05:49 I am not Mar 11 23:06:46 I think we need community effort to work on the next release which will be oe-core based Mar 11 23:08:32 I'm doing daily poky-image-minimal builds with DISTRO=poky with oe-core. haven't tried angstrom in a while, will probably do so again shortly Mar 11 23:10:53 kergoth: have you layered new machines on top? Mar 11 23:11:26 kergoth: I think we should aim for a recent set of recipes from 2011.3 to be ported over to oe-core, as well as wider machine support. Mar 11 23:11:42 likewise: I've tried Mar 11 23:11:54 to build oe-core, I mean Mar 11 23:11:58 me too, I am running into issues building for powerpc Mar 11 23:12:14 but stuck on qemu requiring libGL Mar 11 23:12:48 I'm still not decided what to do with this better - use patch to remove dependency or install it into my LXC container Mar 11 23:14:49 we have a lot of difference between gcc 4.5.0 in oe and yocto meta Mar 11 23:16:00 there are a lot of improvements in OE tha tneed to be pulled over Mar 11 23:16:14 lots of class stuff, lib/oe stuff, and the recipes Mar 11 23:16:20 should probably start a wiki page on it Mar 11 23:20:12 03Eric Bénard  07org.openembedded.dev * rde2f03aa02 10openembedded.git/recipes/busybox/busybox_1.18.3.bb: Mar 11 23:20:12 busybox-1.18.3: add latest fix Mar 11 23:20:12 fix cksum Mar 11 23:20:12 Signed-off-by: Eric Bénard Mar 11 23:20:24 haha Mar 11 23:20:30 busybox again Mar 11 23:20:49 more stable than ever :-D Mar 11 23:21:27 hm Mar 11 23:21:30 stable? Mar 11 23:21:39 didnt notice unstable stuff with busybox Mar 11 23:21:58 well apparently cksum was failing now it's working fine :) Mar 11 23:23:21 hm Mar 11 23:23:25 ah never needed it Mar 11 23:24:17 never needed it also. I'm just following busybox mailing list and pushing fixes when I see them :) Mar 11 23:25:19 ericben_: much appreciated Mar 11 23:25:49 an actual recipe maintainer Mar 11 23:31:46 03Koen Kooi  07org.openembedded.dev * re0e2107311 10openembedded.git/conf/distro/include/angstrom-codec-engine-latest-preferred-versions.inc: Mar 11 23:31:46 angstrom-codec-engine-latest-preferred-versions: bump versions for >2.6.36 support Mar 11 23:31:46 Signed-off-by: Koen Kooi Mar 12 00:06:11 well, we should fix CMDLINE_append = " ${CMDLINE_DEBUG} " in linux.inc Mar 12 00:06:26 s/fix/remove ;) Mar 12 01:05:58 likewise: kergoth yes there is intention of what needs to ported from oe into oe-core and gcc is on my list of things Mar 12 01:06:06 but starting a page is probably best thing Mar 12 01:06:24 actually it was an outcome of TSC meeting to start a page Mar 12 01:07:20 kergoth: mind starting such a page ? Mar 12 01:07:23 please Mar 12 01:08:35 erm, don't think i have a wiki account, will check Mar 12 01:08:36 hehe Mar 12 01:08:49 hmm I can create one for you Mar 12 01:09:18 just give me your email that you want notification sent to Mar 12 01:09:20 i might, not sure Mar 12 01:09:22 will check Mar 12 01:09:25 ok cool Mar 12 01:09:27 don't do much with our wiki Mar 12 01:15:24 OK, I just finished a build for e500v2, using oe-core, meta-openembedded and my custom layer with the p2020rdb machine. meta-openembedded contains the OE master gcc 4.5.0 (that layer is koen's work). Mar 12 01:15:56 there is quite a lot of diff in classes, site handling etc. Mar 12 01:18:21 and the Eclipse IDE Yocto tool did not pickup my new target. Mar 12 01:23:03 * Jay7 -> sleep Mar 12 01:26:09 * likewise <- sleep Mar 12 01:26:34 nite all Mar 12 01:48:40 khem: I've started builder with your gcc patch Mar 12 01:48:51 jfyi Mar 12 02:34:24 Jay7: thx Mar 12 02:35:11 Jay7: I am interested in all combos but mainly it fixes x86 regression so if you can build MACHINE=i586-generic that will help **** ENDING LOGGING AT Sat Mar 12 02:59:57 2011