**** BEGIN LOGGING AT Mon Feb 09 02:59:58 2015 Feb 09 07:34:35 morning all Feb 09 08:33:57 good morning Feb 09 09:53:29 morning all Feb 09 10:01:28 hi bluelightning Feb 09 10:04:07 hi mckoan Feb 09 11:49:35 I have a strange usecase, on some machines (with a harddrive) i can consider having a enough space to install extra/optional packages for convenience, so i made a package group for those extra packages. Then i added this package group into MACHINE_ESSENTIAL_EXTRA_RDEPENDS and everything works as expected. But i don't know if it's a good idea or if there is an alternative way, because this variable implies that those packages are _essential_ Feb 09 11:51:46 MACHINE_EXTRA_DEPENDS looks better indeed Feb 09 13:10:04 I'm seeing some weird errors after updating master Feb 09 13:10:05 | arm-oe-linux-gnueabi-gcc: error: unrecognized command line option '--should-not-have-used-/usr/bin/xml2-config' Feb 09 13:10:21 I suspect cleaning shared state helps Feb 09 13:12:48 Crofton|work: no, it will not fix that Feb 09 13:12:59 bother Feb 09 13:13:09 fix your recipe to stop using xml2-config Feb 09 13:13:34 bother Feb 09 13:13:38 python-lxml Feb 09 13:13:43 http://www.yoctoproject.org/docs/current/ref-manual/ref-manual.html#migration-1.7-binary-configuration-scripts-disabled Feb 09 13:23:24 open embedded: you don't know when patches are taken but when they do, your recipes are broken Feb 09 13:25:02 this is not the first time someone has mentioned they don't know the status of their patches Feb 09 13:31:53 bluelightning, ok fixed Feb 09 15:24:05 Crofton|work: there are 5 recipes failing like this in last bitbake world Feb 09 15:25:58 Summary: 10 tasks failed: /home/jenkins/oe/world/shr-core/meta-openembedded/meta-oe/recipes-support/fltk/fltk_1.1.10.bb, do_compile /home/jenkins/oe/world/shr-core/meta-openembedded/meta-python/recipes-devtools/python/python-lxml_3.2.5.bb, do_compile /home/jenkins/oe/world/shr-core/meta-openembedded/meta-networking/recipes-daemons/ptpd/ptpd_2.3.1-rc2.bb, do_compile Feb 09 15:26:03 /home/jekins/oe/world/shr-core/metca-openembedded/meta-networking/recipes-support/wireshark/wireshark_1.12.3.bb, do_configure /home/jenkins/oe/world/shr-core/meta-openembedded/meta-oe/recipes-support/start-stop-daemon/start-stop-daemon_1.16.0.2.bb, do_compile /home/jenkins/oe/world/shr-core/openembedded-core/meta/recipes-bsp/u-boot/u-boot-mkimage_2014.07.bb, do_compile Feb 09 15:26:06 I have one fic, will send soon Feb 09 15:26:08 /home/jenkins/oeworld/shr-core/meta-openembedded/meta-oe/recipes-aextended/unixodbc/unixodbc_2.3.2.bb, do_configure /home/jenkins/oe/world/shr-core/openembedded-core/meta/recipes-devtools/btrfs-tools/btrfs-tools_git.bb, do_compile /home/jenkins/oe/world/shr-core/meta-openembedded/meta-networking/recipes-connectivity/snort/snort_2.9.6.0.bb, do_configure /home/jenkins/oe/world/shr-core/meta-openembedded/meta-oe/re Feb 09 15:26:14 cipes-graphics/gimp/gimp_2.8.10.bb, do_compile Feb 09 15:41:56 I have python-lxml, and the recipe could use a version update Feb 09 15:42:04 I think I iwll send the build fix first though Feb 09 16:11:00 RP: can you think of a way to get some python run before base.bbclass's anonfunc, to adjust PACKAGECONFIG, other than hooking at RecipePreFinalise and doing a manual updateKeys/update_data? Feb 09 16:11:23 someday we need named event handler dependency handling Feb 09 16:15:07 kergoth: You could do something nasty with := I guess Feb 09 16:15:38 oh, some sort of def'd functionc alled in a := but with side effects outside that variable? i've done that before, it's not pleasant Feb 09 16:16:00 kergoth: right, its nasty Feb 09 16:16:08 k Feb 09 16:16:33 kergoth: although now I think about it, I think the datastore in := is thrown away? Feb 09 16:16:52 RP: since base is inherited globally now anyway, maybe it could be migrated into INHERIT, then you could inject classes before base, eventually Feb 09 16:17:06 hmm, good question, if not it really should be Feb 09 16:17:36 kergoth: I think I tried this recently and concluded we'd prevented that particular horror :) Feb 09 16:17:45 ah, good Feb 09 16:18:28 if we move base to inherit, we could ditch the hardcoding of it in bitbake. might not be a bad thing Feb 09 16:18:33 * kergoth shrugs Feb 09 16:18:58 kergoth: its never struck me as our biggest issue, but yes Feb 09 16:20:16 * kergoth oh, i could just make this code a .inc instead of a bbclass. since its required immediately, any anonymous python it runs would be before base gets parsed Feb 09 16:20:19 * kergoth rolls eyes Feb 09 16:47:30 hmm, why is wic hanging Feb 09 17:01:11 leftover cruft in the way Feb 09 17:45:02 * challinan just hacked around python-lxml xml2-config build issue - ugly Feb 09 18:01:12 challinan, see my patch on the list :) Feb 09 18:02:20 and ignore trondeau, make a PyQwt recipe :) Feb 09 18:02:38 heh it will definitely be a major learning curve! ;) Feb 09 18:02:54 don't worry about it Feb 09 18:03:00 there are other issues long term Feb 09 18:03:17 i really want PyQwt, because there is much content available Feb 09 18:03:38 beyond gnuradio? Feb 09 18:03:47 PyQt is sane. Now Feb 09 18:03:48 Crofton|work I had several internet outages today with firmware upgrade to router and call to comcast :( Feb 09 18:03:50 missed your patch Feb 09 18:04:09 ouch Feb 09 18:04:12 no, it's gnuradio context that I'm referring to Feb 09 18:04:24 I do not think we use to much from PyQwt Feb 09 18:04:38 all the good qui blocks are qwt, no? Feb 09 18:04:40 PyQt is another matter, but that should be OK in master now Feb 09 18:04:47 mostly PyQt Feb 09 18:05:24 * challinan is building master, have worked around several issues, including the lib64 issue on several packages. I'll report when I have something working Feb 09 18:06:37 what lib64 issue? Feb 09 18:06:57 several packages don't obey libdir when installing Feb 09 18:07:03 oh, right, that one Feb 09 18:07:20 in fact, their logic for determining which libdir to select is not cross-friendly ;) Feb 09 18:09:01 vim fails with this message: | /bin/sh: ./sjiscorr: No such file or directory Feb 09 18:09:20 and subsequent runs succeed **** BEGIN LOGGING AT Tue Feb 10 01:42:16 2015 **** ENDING LOGGING AT Tue Feb 10 02:59:58 2015