**** BEGIN LOGGING AT Fri Jan 04 02:59:58 2013 Jan 04 04:12:03 kergoth... Jan 04 04:13:10 does abuse of OE qualify me for the OE booth at scale? Jan 04 04:23:38 nerdboy: no publicity is bad :) Jan 04 04:27:08 well, it might actually be interesting for people to hear how we abuse oe/git while trying to be "agile"... Jan 04 04:29:16 at your wotk ? Jan 04 04:29:22 make that oe/git/jenkins/bash i suppose Jan 04 04:29:28 yup Jan 04 04:30:22 nerdboy, you need to come and help @ the booth! Jan 04 04:30:47 put me on the list... Jan 04 04:31:44 i have to hang out at the gentoo booth too... Jan 04 04:32:48 but pretty much all the action for the last year falls under the heading "fun with (arago) open-embedded" Jan 04 08:44:17 morning all Jan 04 09:03:54 khem: still awake? Jan 04 09:07:32 awake.. jsut Jan 04 09:57:21 is it a bad idea to use a custom image recipe that appends the core-image-minimal and add the package-management feature ? Jan 04 09:58:13 afournier: if that's what you need then it is good way Jan 04 10:05:46 ok Jan 04 12:48:45 RP_, are you still on vacation? Jan 04 13:33:07 does a bitbake-devel web archive exist anywhere? The gmane archive stops at the time the list was moved from berlios. Jan 04 13:34:25 http://lists.linuxtogo.org/pipermail/bitbake-devel/ Jan 04 13:50:34 JaMa: sorry, I should have qualified that with "searchable", like a gmane based archive Jan 04 13:52:37 wmat: Google specifying site:? Jan 04 13:56:37 broonie: thx, that worked Jan 04 15:20:57 kergoth, I think I have a clue! Jan 04 15:21:10 so the fftw recipe uses the SRC_URI from a .inc file Jan 04 15:21:19 but the checksums are in the .bb files Jan 04 15:21:40 and there are three .bb files depedns ing what data types you are building for Jan 04 15:22:34 so if yuo upgrade one recipe, so you have checksum entries for both tarballs it "works" Jan 04 15:22:37 evil Jan 04 15:23:04 crofotn still figuring with neon fot fftw? Jan 04 15:23:32 yeah, and trying to update the version Jan 04 15:23:45 it feels like we really need to make fftw machine specific though Jan 04 15:24:10 which kind of sucks Jan 04 15:24:58 first I need to update to 3.3.3 and get the checksum sorted out Jan 04 15:34:57 Hello. Jan 04 15:35:26 I've probably found „bug” in pulseaudio recipe in openembedded-core repo. Jan 04 15:35:53 Where can I report it and what is possible time of pushing changes? Jan 04 15:36:04 openembedded-core@lists.openembedded.org Jan 04 15:36:13 or the yocto bugzilla Jan 04 15:36:14 Beucase I do not know if it is necessary to overlay this receipe, beuase I need this fixed ASAP. Jan 04 15:36:20 no times are ever guaranteed. Jan 04 15:36:32 then fix it in your layer for now Jan 04 15:37:25 https://bugzilla.yoctoproject.org/ Jan 04 15:37:27 right? Jan 04 15:37:37 that's what i said, yes Jan 04 15:37:41 otavio, that should work Jan 04 15:38:29 kergoth, you see what I said about my fftw issue? Jan 04 15:38:43 basically, the checksum are duped three times Jan 04 15:40:01 How can I generate patch for this recipe correctly? Jan 04 15:40:30 git diff? Jan 04 15:40:33 git format-patch Jan 04 15:40:40 something like that Jan 04 15:42:10 Crofton|work: not surprising, that's quite common. the .inc SRC_URI uses ${PV}, so there's no point in redefining it, but the checksum differs as the SRC_URI changes based on th ePV failure Jan 04 15:42:12 arrrg Jan 04 15:42:15 s/ePV/PV/ Jan 04 15:42:17 s/failure/value/ Jan 04 15:42:23 still can't get failure though Jan 04 15:42:48 I moved all the bb's to th enew version Jan 04 15:42:58 I'll keep poking Jan 04 15:43:01 weird Jan 04 15:43:06 seriously Jan 04 15:44:45 9~http://wklej.org/hash/68bd395e87d/ Jan 04 15:44:47 Attach such file? Jan 04 15:45:05 can you repost the url without the leading crap? Jan 04 15:45:11 yeah, I am being really lazy Jan 04 15:45:11 http://wklej.org/hash/68bd395e87d/ Jan 04 15:45:15 Of course, sorry. Jan 04 15:45:37 sure, it makes it clear what you are doing Jan 04 15:49:14 finally Jan 04 15:49:19 a checksum failure Jan 04 15:49:40 I had to beat bitbake with a club (rm the tarball from downloads) Jan 04 15:49:53 and remove all the checkum lines from the recipes to get a failure Jan 04 15:52:01 yep, looks like having more than one checksum entry, even when they are all wrong cause s the fetcher to theink to tarball is OK Jan 04 15:52:04 #fail Jan 04 15:58:13 in the bitbake source, is conf/bitbake.conf a sample? Jan 04 16:03:48 meta/conf/bitbake.conf is the actual conf file that's parsed right? Jan 04 16:04:52 yes, it's a sample, a minimal config intended for anyone wanting to use bitbake without oe/poky Jan 04 16:05:20 thx Jan 04 17:06:00 Can I only overlay .inc file? Jan 04 17:06:54 you can't bbappend to anything but recipes, if that's your question. not .inc, not .conf, not .bbclass Jan 04 20:12:11 we remove PR now, since we move to PR server Jan 04 20:12:18 well, PR =ro is removed Jan 04 20:13:46 work in progress, dont think i like the task deps format: https://gist.github.com/4455568 Jan 04 20:14:17 might just make it dump it in metadata format Jan 04 20:29:16 OK, so how to I make EXTRA_OECONF depend on a tune features? Jan 04 20:29:31 probably base_contains Jan 04 20:29:44 same way you'd check distro features Jan 04 22:08:03 khem: ping Jan 04 22:54:44 hi Jan 04 22:59:20 herro **** ENDING LOGGING AT Sat Jan 05 02:59:59 2013