**** BEGIN LOGGING AT Wed Sep 24 03:00:00 2014 Sep 24 07:24:04 morning all Sep 24 15:22:25 morning Sep 24 18:05:09 this is a horribly vague qt question ... Sep 24 18:05:19 I have a piece of sw using qt Sep 24 18:05:39 that has some svg icons that appear to be embedded i the executable Sep 24 18:05:53 but when I run the app, the icons are blank Sep 24 18:06:03 anyone have clues on what might be going wrong Sep 24 18:07:13 Crofton: a horribly vague guess: maybe your qt installation is missing the plugin to handle that image type. Sep 24 18:08:27 how would i identify such a plugin :) Sep 24 18:08:52 Good question. :-) Sep 24 18:09:24 I'm just guessing. I'm not even sure this is the actual problem. Sep 24 18:10:05 I suppose I need to look at avaiable qt packages Sep 24 18:12:12 qt4-plugin-imageformat-svg/latest Sep 24 18:12:20 suspiciously absent from rootfs :) Sep 24 18:16:08 mario-goulart, I owe you a beer Sep 24 18:16:14 thanks for leading me to water Sep 24 18:17:03 I'm glad you could find the solution, Crofton. :-) Sep 24 18:18:02 annoying we do not seem to auto detect plugins Sep 24 18:29:15 what is state of the art for touchscreen calibration? Sep 24 18:35:16 Can anybody assist with an ENABLE_BINARY_LOCALE_GENERATION question? I have it enabled in my local.conf but the eglibc-binary-localedata-* packages do not seem to exist in the directory with the eglibc-localedata-* packages. Sep 24 18:36:13 I cleaned both eglibc and eglibc-initial and redid them but still no files. Wondering if there is something else that could be preventing their creation. bitbake -e confirms it is set to 1. Sep 24 18:37:15 Crofton: we have used xinput-calibrator. It works fine for our needs. Don't know if it is the state of the art. The "x" part may be relevant, depending on your graphics backend. Sep 24 18:38:00 I am getting sucked in to gui stuff by marketing so they can make flashy demos Sep 24 18:38:12 and the screen is a usb fb thing Sep 24 18:38:32 no one in their right mind would do this on purpose :) Sep 24 18:38:45 :-) Sep 24 18:59:20 * nerdboy re-hacks his chromebook Sep 24 22:04:10 we are using ASSUME_PROVIDED for git? Sep 24 22:14:32 hmmm, so I have a package that wants to run git during build to embed the git hash in the binary Sep 24 22:14:48 but, it uses cmake and we told cmake not to look on the build machine for git Sep 24 22:14:56 which is good Sep 24 22:15:08 (binaries in general actually) Sep 24 22:15:23 so I am thinking of making the recipe depend on git-native Sep 24 22:15:35 but I wonder if this will drive the fetcher crazy? Sep 24 22:15:45 we require git before we even get to building anything, due to parse time fetch stuff Sep 24 22:15:54 afaik Sep 24 22:16:30 unless you have no git:// using recipes, anywhere, of course. otherwise it's needed on the host at parse time Sep 24 22:17:30 I know Sep 24 22:17:47 I am wondering if using git-native will confuse anything Sep 24 22:18:05 hopefully, bitbake keeps using the one from the build machine Sep 24 22:18:17 and things trying to find a git binary use the staged one Sep 24 22:18:18 that depends on where it's called Sep 24 22:18:31 we also need cpio in ASSUME_PROVIDED, quilt-native expects it Sep 24 22:18:34 parse time would use teh build machine, do_fetch would use the one in teh sysroot, depending on when the task is run and when it's built Sep 24 22:18:47 since the native sysroot is in the PATH for tasks Sep 24 22:18:50 afaik, anyway Sep 24 22:19:20 * Crofton does not want to comment on an environment where git hashes need to be embedded in binaries so people can figure out who is running what Sep 24 22:20:18 but as long as there's no huge behavioral difference between the two git versions, i doubt there'd be a problem in practice Sep 24 22:20:29 yeah, I am also thinking this :) Sep 24 22:20:41 still, since we need git on the host anyway, i'd just let cmake find it. or better yet, decide whether to based on ASSUME_PROVIDED Sep 24 22:21:06 the easy solution is stgae git in the native sysroot Sep 24 22:21:16 since I can do that without messing with cmake toolchain file Sep 24 22:22:03 otherwise I suspect I need to mess with the cmake.bbclass Sep 24 22:22:10 * Crofton goes back to cmake docs Sep 24 22:39:14 arg Sep 24 22:39:35 it ignores my efforts to put git in the sysoot by running bitbake git-native **** ENDING LOGGING AT Thu Sep 25 03:00:01 2014