**** BEGIN LOGGING AT Thu Feb 25 02:59:59 2016 Feb 25 09:17:32 hey! can anyone explain why all sourcecode for components are included in SDK when built with bitbake -c populate_sdk my-image-recipe ? Feb 25 09:34:10 mago_: it's included in the -dbg packages so that gdb can show source code Feb 25 14:48:53 joshuagl, you saw the comments about the core swupd should be in its own layer? Feb 25 14:49:30 Crofton|work: I did Feb 25 14:49:45 Hopefully there isn't a problem with the suggestion Feb 25 14:50:15 I don't think so, it does make sense Feb 25 14:50:20 great Feb 25 14:50:26 I just wanted to doble check Feb 25 14:50:49 I happen to be trying to figure out swupdate atm Feb 25 16:50:09 is there a way to create a mega ipk of all package-split ipks Feb 25 16:50:50 automatically? no Feb 25 16:51:16 if you're using do_split_packages then it returns a list of packages it created, to make it easy to make a metapackage Feb 25 16:51:50 rburton: I need to bundle the artifacts into tarball Feb 25 16:52:32 may be I can add tar package backend along with ipk or others just for this recipe ? Feb 25 16:52:58 tar packages will split too won't they? Feb 25 16:53:11 hmm Feb 25 16:53:14 yeah Feb 25 16:59:20 sounds like you'd need a new custom task Feb 25 16:59:40 yeah after populate_sysroot Feb 25 18:31:17 morning all Feb 25 19:24:42 Is someone succeed in building ARM Big Endian using Jethro? Feb 25 19:24:52 I am failing when building busybox Feb 25 20:09:27 Hello I am wondering if anyone can help me with writing a .bbappend file that modifies the install step for the target package but not the native package. Feb 25 20:10:37 Back in 2012 there was a patch to automake to prevent it from making *.pyc files and only make *.pyo files. This was apparently done because there was a patch to python3 to make optimization default. This was not done to python2 though. So now when I am trying to build python-dbus for python2 I do not get the *.pyo files. I am trying to figure the best way around this issue without including the *.py files on the target and letting Feb 25 20:10:37 them compile when used. Feb 25 20:19:19 Never mind I figured it out, it's do_install_append_class-target() Feb 25 20:52:02 CipherWizard: you don't want .pyo files really Feb 25 20:52:32 CipherWizard: a .pyo is a .pyc but with documentation and asserts removed. that doesn't really count as "optimised" unless you're that desperate for every single byte Feb 25 20:53:16 rburton: We are not wanting them because with python2 you still have to add the -O option all the time you are running the script. Feb 25 20:53:53 fwiw oe-core master removes the patch for python3 which makes it write .pyo by default Feb 25 20:57:27 ??? removes a patch? what patch? Feb 25 20:58:36 I know a patch is added to python3 to use optimization by default (-O). but there isn't a patch like that for python2 Feb 25 20:59:17 yes and in oe-core master i've removed the python3 patch Feb 25 20:59:18 as its stupid Feb 25 20:59:36 enforcing .pyo is literally a bad idea Feb 25 20:59:59 "my program uses docstrings and asserts but under yocto it doesn't work" Feb 25 21:00:54 if you really want .pyo files, then running a post-rootfs python -O will work Feb 25 21:01:21 I don't want the *.pyo files. I want the *.pyc files Feb 25 21:01:55 But there is a patch in automake that removes the *.pyc files Feb 25 21:02:40 well that's pretty dumb. Feb 25 21:02:41 * rburton removes Feb 25 21:02:44 It was placed in there because of the patch to Python3 that added the optimization by default Feb 25 21:03:05 immediate fix would be to bbappend that patch out of automake Feb 25 21:03:20 So should I just submit a patch that removes that patch from OE-Core? Feb 25 21:03:25 please do Feb 25 21:03:39 saves me doing it as i'm about to eat Feb 25 21:03:47 That is what I am attempting at the moment. (recompiling) Feb 25 21:03:49 if anything it should not build .pyo **** ENDING LOGGING AT Fri Feb 26 02:59:58 2016