**** BEGIN LOGGING AT Mon Jan 12 02:59:59 2015 Jan 12 06:52:35 so if a source has platform & mach files Jan 12 06:52:44 is that enough for a BSP? Jan 12 08:06:22 regarding the cmake in poky, it should be preconfigured to install compiled binary into the sysroot image instead of the host's root directory right? Jan 12 08:12:27 good morning Jan 12 08:42:57 regarding the cmake in poky, it should be preconfigured to install compiled binary into the sysroot image instead of the host's root directory right? Jan 12 09:42:57 morning all Jan 12 12:02:08 may I know what oe_runmake stands for? Jan 12 12:07:07 chankit1: hi, it just runs make with the options we typically supply to it (e.g. the value of EXTRA_OEMAKE) Jan 12 12:07:37 chankit1: you can see the definition of it in meta/classes/base.bbclass Jan 12 12:22:55 bluelightning: thanks. So if we call cmake with installation prefix PATH in a recipe, does poky install to PATH in host or in the final image? Jan 12 12:23:17 bluelightning: also, when did yocto change from eglibc to glibc? Jan 12 12:26:24 chankit1: http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=54a3375a1084f02ebd95886739e951228e945c57 Jan 12 12:31:03 maxin1: thanks Jan 12 14:06:58 I'm having an issue when building -- /home/erisley/poky/build-atmel/tmp/sysroots/x86_64-linux/usr/lib/libglib-2.0.so: file not recognized: File format not recognized collect2: error: ld returned 1 exit status Jan 12 14:16:23 Rizzo: what does this say: file /home/erisley/poky/build-atmel/tmp/sysroots/x86_64-linux/usr/lib/libglib-2.0.so Jan 12 14:50:42 Hi, not sure if this is the right place to ask for question related to freescale boards but since there is yocto support for imx6, may be someone had faced similar issue before. I'm listening to multicast packet on 5353 port to see what do i get out of multicase request Jan 12 14:51:22 on the device, i'm running avahi and on the host, it has msdn responder kind of stuff which sends multicvast packets every few seconds Jan 12 14:52:44 well, after few minutes, i don't see any packet on port 5353....even with all filter disable. so i'm trying to figure out why there are no packets on my device even thouhg the packets are received by other devices withing the network Jan 12 15:00:50 q Jan 12 16:11:58 * paulg wonders what the heck BeautifulSoup is and why the fetcher needs it. Jan 12 16:12:48 ...since it breaks the self hosting build appliance, with an unsatisfied dependency on htmlentitydefs Jan 12 16:15:14 iirc it's an html parsing library for python. Jan 12 16:15:32 as you'll see in the first couple links of googling for it :) Jan 12 16:15:33 http://www.crummy.com/software/BeautifulSoup/ Jan 12 16:17:49 paulg: I thought we fixed the build appliance to have that in it, but I can't see that we did :( Jan 12 16:19:25 I'm checking with reverts of 25e3e57c55129 and 6bb241a278135 to see if that gets me building again... Jan 12 16:20:15 paulg: I think perhaps we need to replace the list of python modules in meta/recipes-core/packagegroups/packagegroup-self-hosted.bb with just python-modules Jan 12 16:21:52 probably not a bad idea. Jan 12 16:22:00 it already looks kind of long. Jan 12 16:22:18 and I can't imagine it saves us much time or space to be so specific. Jan 12 16:22:23 that's what we did in buildtools-tarball to solve the same issue Jan 12 16:22:39 I'll abandon my revert plan and give that a whirl... Jan 12 16:29:01 of course now I'll have to rebuild the builder to know if the builder can build the builder with that in place... Jan 12 16:29:13 I'll report back in a couple hrs. :) Jan 12 16:30:06 bluelightning, thanks for the heads up on 0add8abc12b850e3 "buildtools-tarball: package all of Python" Jan 12 16:30:33 paulg: np Jan 12 16:42:51 bluelightning: oddly, I got completely different results than you running against mako. what version were you operating against, 1.0.0? Jan 12 16:42:58 https://gist.github.com/kergoth/dafa05572fa6506edf88 Jan 12 16:43:57 * kergoth pokes around its source tree Jan 12 16:44:57 kergoth: recipetool create -o python-mako_0.9.1.bb https://pypi.python.org/packages/source/M/Mako/Mako-0.9.1.tar.gz Jan 12 16:45:11 ah, i was testing 1.0.0, i'll try 0.9.1 instead Jan 12 16:45:31 I only used 0.9.1 since that is the version we currently have in OE-Core Jan 12 16:45:40 ah, for copmarison. makes sense Jan 12 16:46:33 okay, that's weird. p ythondeps -p shows lookup & template, yet it didn't remove them from the detected module deps? Jan 12 16:46:34 * kergoth digs Jan 12 16:49:50 aha, got it. it wasn't handling find_packages(), so it failed to pick anything up for packages=, yet it still successfully picked up scripts. it ended up just scanning Mako-0.9.1/scripts/mako-render :) oops Jan 12 16:50:20 there's code specifically to check for this so we don't end up checking partial paths, but it must not be functioning Jan 12 16:50:21 * kergoth digs Jan 12 17:13:56 why do we still have a python-distribute recipe? Jan 12 17:14:01 distribute is part of setuptools 0.7+ Jan 12 17:14:08 * kergoth mutters Jan 12 18:36:37 bluelightning_: k, after the latest fixes: https://gist.github.com/kergoth/0a641e90cdfd37fd2100 Jan 12 18:37:52 prefixes install_requires with python-, moves metadata fields above SRC_URI for oe/yocto conventions, replaces existing LICENSE line, fixed the scanning bug, for now excludes sys.builtin_module_names from the python running recipetool Jan 12 18:38:30 bah, prefixing python- broke the removal of those items from the unmapped. hehe Jan 12 18:38:32 * kergoth fixes Jan 12 21:25:32 hmmm Jan 12 21:25:41 not sure i like how i'm handling extras_require, but at least it works for now Jan 12 21:34:38 grr, this is taking way long than i was hoping Jan 12 22:17:59 Okay, just need to fix package_dir package name handling and I can re-submit **** ENDING LOGGING AT Tue Jan 13 02:59:59 2015