**** BEGIN LOGGING AT Tue Jan 17 03:00:02 2017 Jan 17 09:11:24 morning Jan 17 10:13:18 Crofton|work: koen angstrom site is very broken at moment Jan 17 10:18:59 I pinged the infra guys Jan 17 10:23:26 as is Angstrom build :-( Jan 17 10:32:06 ERROR: Unable to parse /home/graeme/Linaro/Work/angstrom/bitbake/lib/bb/data_smart.py Jan 17 10:32:08 Traceback (most recent call last): Jan 17 10:32:10 File "/home/graeme/Linaro/Work/angstrom/bitbake/lib/bb/data_smart.py", line 426, in DataSmart.expandWithRefs(s="linux${@['','-gnueabi'][bb.data.getVar('TARGET_ARCH',d,1) in ['arm', 'armeb']]}", varname='TARGET_OS'): Jan 17 10:32:12 except Exception as exc: Jan 17 10:32:14 > raise ExpansionError(varname, s, exc) from exc Jan 17 10:32:16 Jan 17 10:32:18 bb.data_smart.ExpansionError: Failure expanding variable TARGET_OS, expression was linux${@['','-gnueabi'][bb.data.getVar('TARGET_ARCH',d,1) in ['arm', 'armeb']]} which triggered exception AttributeError: module 'bb.data' has no attribute 'getVar' Jan 17 10:36:52 XorA: change that bb.data.getVar('TARGET_ARCH',d,1) to d.getVar('TARGET_ARCH') Jan 17 10:37:46 joshuagl: it was more a poke to Angstrom guys that default build is broken, but thanks Jan 17 15:13:50 XorA: i expect a poke of the angstrom guys would have better results in #angstrom ;) Jan 17 15:13:51 * kergoth yawns Jan 17 15:22:34 kergoth: except no-one hangs there anymore Jan 17 21:04:26 hi, does anyone here know TI processors AM35XX, and u-boot SPL code for them? I have a problem with x-loader. It prints 'C' letter and waits few secons before loading u-boot. How to change this x-loader(SPL) delay? Jan 18 00:02:08 Riddle me this, Batman: if you are writing a bbappend and doing an _*pend on a task, is there any way to specify that you want the function you are writing to be python or shell or must you match what the underlying recipe is using? Jan 18 00:03:38 NU-Slacker: it must match Jan 18 00:03:54 NU-Slacker: since you are just appending/prepending to the value Jan 18 00:04:09 NU-Slacker: however, if you want some more flexibility you can instead use a prefunc/postfunc Jan 18 00:04:26 which can be either shell or python regardless of what the main function is Jan 18 00:17:29 Is there any dis/advantage to the prefunc vs _append? Jan 18 00:20:14 prefunc/postfunc is cleaner and more logical Jan 18 00:26:08 Understood. So two major disadvantages. Jan 18 00:32:49 Looks like that functionality is not too widely used in the wild apart from a couple bbclasses. Now I dont feel as bad for not finding it. Jan 18 00:34:04 yeah most people use appends as they're simpler, sort of Jan 18 00:34:17 if you've more than "delete a file that was installed" to do, a postfunc is neater Jan 18 00:34:23 * rburton1 -> bed Jan 18 00:35:26 I thought _append was simpler too until I started getting parse errors all over the place. Jan 18 00:38:16 NU-Slacker: I presume that's because of python vs. shell or python with incompatible indentation ? Jan 18 00:41:58 bluelightning: Yup. All of those in turn. **** ENDING LOGGING AT Wed Jan 18 03:00:00 2017