**** BEGIN LOGGING AT Mon Apr 28 02:59:59 2014 Apr 28 07:15:43 good morning Apr 28 07:18:14 mckoan, all, good morning Apr 28 07:27:08 anyone know of any recipes for enet? Apr 28 08:45:32 morning all Apr 28 08:49:08 hey bluelightning Apr 28 08:49:13 hi koen Apr 28 08:50:00 bluelightning: I just noticed that the meta-intel dora branch update triggered more linux-yocto LINUX_VERSION updates globally Apr 28 08:50:22 the good news is that meta-intel master does have all LINUX_VERSION vars properly overriden Apr 28 08:50:31 koen: is it a change in the right direction still though? Apr 28 08:50:40 the dora one, no Apr 28 08:50:51 argh :/ Apr 28 08:51:12 it's funny to see that some BSPs in meta-intel/dora do LINUX_VERSION_machine and others do not Apr 28 08:51:25 yeah it hasn't been consistent Apr 28 08:51:45 I think the problem is people only realised it was being done incorrectly recently Apr 28 08:51:48 I'll see if I have some time today to send a patch, but the 1.7 patches oe-core updates are wreaking havoc on $dayjob builds Apr 28 08:52:20 FWIW Darren agrees we need to fix dora as well Apr 28 08:52:47 koen: which OE-Core updates are causing you pain? Apr 28 08:53:23 the libgcc-initial addition, which I just fixed by adding a libgcc-initial recipe for the linaro-4.8 version Apr 28 08:53:37 but I bet I'm going to hit some seperate builddir issues soon Apr 28 09:01:11 well, if you want to track master you have to expect some pain Apr 28 09:06:31 I'm not complaining about the changes, just making clear that I might now have time to come up with a patch for meta-intal dora :) Apr 28 09:13:39 hi bluelightning, koen, all Apr 28 09:13:48 hey mckoan Apr 28 09:13:51 hi mckoan Apr 28 09:13:55 koen: fair enough Apr 28 09:55:56 can i clean/remove everything for a given machine ? Apr 28 09:57:01 afournier: no, there is no way to do that other than removing your entire TMPDIR (but for builds after that, items will be quickly restored from sstate cache if they have not changed) Apr 28 09:57:46 ok Apr 28 11:59:09 bluelightning, I'll begin to question my own sanity if I can't even get simple python if-then-else statements right.. Apr 28 12:09:04 kroon: it's not the behaviour of the "continue" that is the issue, it's the use of the "checksum" variable in two different places Apr 28 12:09:41 kroon: if you'd rather not revert we could patch it to use a different variable in the glob loop Apr 28 12:12:23 I just dont see the conflict.. Apr 28 12:17:50 There is no continue affecting the inner glob-loop, it affects the outer loop Apr 28 12:18:03 kroon: again, it's not about that Apr 28 12:18:34 follow through what happens when we expand a glob to a number of items e.g. *.patch -> 1.patch 2.patch 3.patch ... Apr 28 12:19:01 you'll get 3.patch added twice Apr 28 12:20:17 one time in the glob loop, and the second time ? Apr 28 12:20:38 the second time at the end Apr 28 12:21:15 even though that is now part of the "else" branch ? Apr 28 12:22:08 oh... I completely missed that you'd moved that Apr 28 12:22:39 sorry :( Apr 28 12:23:12 goddamit paul, i was beginning to seriously question my sanity :-) Apr 28 12:25:11 I've followed up on the thread Apr 28 12:26:09 kroon: thanks for fixing the bitbake-diffsigs issue as well, that was my fault too... Apr 28 12:28:47 damn, he's already reverted the patch Apr 28 12:28:53 I feel really stupid right now... Apr 28 12:29:54 I'll make sure it gets reinstated Apr 28 12:35:28 bluelightning, we could remove that initial "checksum = None" initializer aswell then.. Apr 28 12:36:29 kroon: yes we could Apr 28 12:37:45 kroon: it would be great if you could send a v2 that does that and all of this won't have just been me wasting everyone's time... ;) Apr 28 12:38:29 bluelightning, hehe, yes I'll do that Apr 28 12:39:46 it would probably be worth mentioning that the original reverted patch was fine in your commit message so we have something recorded in the history Apr 28 12:40:48 thanks Apr 28 14:45:17 goddamn yocto crap Apr 28 14:45:28 can't even get versioning policy right in the dev manual Apr 28 14:45:54 * koen decides it's time to call it a day Apr 28 14:55:34 damn it, I should have noticed that when I helped Scott write that section Apr 28 14:56:02 I'd appreciate you not labelling all of our stuff as crap though Apr 28 14:56:32 * bluelightning files a bug Apr 28 15:26:10 libtoolize: `COPYING.LIB' not found in `/home/admin/tmp/OE/tmp-eglibc/sysroots/vm_qemux86/usr/share/libtool/libltdl' Apr 28 15:27:32 i have this message when i try to build unixodbc, in the paste i made a recipe that append libtool-cross and add many files in ${D}${target_datadir}/libtool/ but i still feel it's not the correct way Apr 28 15:29:49 http://pastebin.com/26HXDU7d Apr 28 15:38:41 got it i missed a DEPENDS on libtool **** ENDING LOGGING AT Tue Apr 29 02:59:58 2014