**** BEGIN LOGGING AT Sat Jun 08 02:59:58 2013 Jun 08 05:53:00 ka6sox: found them at the local place. they were wrong about not having them and Manny found it in another bin in another bag looking for a replacement JFET Jun 08 05:53:27 sweet Jun 08 05:53:34 in SD or SF? Jun 08 06:06:15 in SD Jun 08 06:06:26 San Diego Electronics Supply has them Jun 08 06:07:28 ka6sox: this is what I'm buildling... Jun 08 06:07:31 http://www.qsl.net/k/k0ff/Balanced%20loop%20Antenna%20Remote%20Preamp/N1LF%20Remote%20Loop/Balanced%20Loop%20Preamp.htm Jun 08 06:07:44 It's second schematic without a transformer Jun 08 06:07:58 * bhuey hates RF transformers Jun 08 08:20:00 morning Jun 08 08:38:57 morning Jun 08 09:08:18 morning Jun 08 09:19:00 morphis: I guess there's no problem supposing that in our QPA, ws_CreateWindow creates something inheriting BaseNativeWindow; that would be my way of calling the resize routines Jun 08 09:19:32 Tofe: so we add a resize method to BaseNativeWindow? Jun 08 09:19:45 no no, I simply call setBuffersDimensions Jun 08 09:20:07 (and I completed the implementation of the latter, of course) Jun 08 09:20:40 I could have also called the ANativeWindow API, but I didn't see the benefit Jun 08 09:20:56 hm, so we would call ws_CreateWindow directly from our QPA? Jun 08 09:21:17 Tofe: did you report that android-audiosystem_bzr.bb fails to find time.h in your build? Jun 08 09:21:28 JaMa: yes Jun 08 09:21:29 JaMa: yes, it was Tofe Jun 08 09:21:58 morphis: yes, is that wrong ? I thought our QPA was hybris specific ? Jun 08 09:21:59 good as I have the same now in 2 independent build Jun 08 09:22:00 s Jun 08 09:22:14 Tofe: it is for sure Jun 08 09:23:04 Tofe: do you also have "make: dpkg-architecture: Command not found Jun 08 09:23:12 JaMa: I didn't really solve it, by the way; just got lucky that at some point, bb lost the track of android-audiosystem not being compiled, and continued with the rest Jun 08 09:23:12 at the begining of log.do_compile? Jun 08 09:23:24 hm, the native window gets created when we create a surface with eglCreateWindowSurface Jun 08 09:23:28 JaMa: ah, no, I didn't have that. Jun 08 09:24:08 morphis: ah, you mean I got the thing upside-down Jun 08 09:24:15 it's using it in build /trunk/lib/build-tree.sh:architecture=`dpkg-architecture -qDEB_HOST_ARCH` Jun 08 09:24:48 Tofe: I am just thinking aloud if we may find another way without exposing the native window classes from libhybris to our QPA Jun 08 09:25:13 JaMa: didn't I patched that line out? Jun 08 09:25:24 I thought so Jun 08 09:25:24 morphis: ah ok yes, so using eglCreateWindowSurface instead of ws_CreateWindow Jun 08 09:25:31 nono Jun 08 09:25:45 we can use first ws_CreateWindow and then eglCreateWindowSurface Jun 08 09:25:49 or just eglCreateWindowSurface Jun 08 09:28:21 Tofe: but ok, lets go the ws_CreateWindow route Jun 08 09:28:42 as a first step, mostly for the resize thing, it will make things easier Jun 08 09:29:27 Tofe: we have to modfy libhybris for this Jun 08 09:29:36 as ws_CreateWindow isn't exposed as public API yet Jun 08 09:30:16 well, we can call eglCreateWindowSurface and do the cast anyway, if we want Jun 08 09:30:47 you mean with nativeWindow set to NULL? Jun 08 09:31:06 ah no wait, that was stupid Jun 08 09:35:06 morphis: nope, there is only 1 patch dont-use-specific-sysroot.patch Jun 08 09:35:14 hn Jun 08 09:35:25 then we really should drop it as it makes not sense for us Jun 08 09:36:19 but if I remember well I found a way to avoid go through that part of the build system Jun 08 09:37:32 hm, build-tree.sh should be never called Jun 08 09:37:48 it's also in 2 other files Jun 08 09:37:59 ah yes Jun 08 09:38:03 build-component.sh project_make Jun 08 09:38:03 project_make Jun 08 09:38:16 build-component.sh is to be ignored Jun 08 09:38:47 JaMa: but we can drop both lines from project_make Jun 08 09:38:49 I guess you're building on machine with dpkg, right? Jun 08 09:38:59 yes, ubuntu 13.04 Jun 08 09:39:14 that can explain why only me and Tofe see it Jun 08 09:39:26 ok I'll check and push it Jun 08 09:39:46 JaMa: can you try this http://pastebin.com/qGReiDui Jun 08 09:42:18 # we don't depend on dpkg-native Jun 08 09:42:19 sed -i 's:/\(^.*shell dpkg-architecture.*$\):#\1:g' ${S}/project_make Jun 08 09:42:24 is what I'm trying :) Jun 08 09:42:37 ups extra / Jun 08 09:42:58 ok :) Jun 08 09:47:29 error message about dpkg is gone, time.h issue still there Jun 08 09:47:52 it doesn't point to sysroot at all: -I. -I../../include -I../../lib/include -I/include -I/include Jun 08 09:48:07 I guess some variable wasn't expanded in last 2 Jun 08 09:48:25 maybe Jun 08 09:49:02 -I$(mot_home)/include -I$(oss_home)/include Jun 08 09:49:52 I'll compare with borabora Jun 08 09:59:43 morphis: do we still need recipes-webos/connman-adapter/connman-adapter_git.bb? or was it completely replaced by webos-connman-adapter? Jun 08 10:00:17 we have the recipe, but I don't see any runtime dependency on it Jun 08 10:08:51 morphis: one more question :) Jun 08 10:08:53 recipes-webos/nyx-modules/nyx-modules.bb:WEBOS_TARGET_MACHINE_IMPL_qemuall = "emulator" Jun 08 10:09:09 I'll move this to webos.inc, next to conf/distro/include/webos.inc:WEBOS_TARGET_MACHINE_IMPL ?= "device" Jun 08 10:09:30 this way we have "device" set in WAM and qemuarm is trying to build qt-webos-plugin Jun 08 12:34:35 interesting environment and run scripts are the same in android-audiosystem even time.h is in the same spot and still one fails and other works Jun 08 12:34:52 and building dpkg-native doesn't help too :) Jun 08 14:14:00 JaMa: we can drop connman-adapter Jun 08 14:14:12 it's completely replaced by webos-connman-adapter Jun 08 14:14:46 JaMa: and the nyx-modules change is also fine for me **** ENDING LOGGING AT Sun Jun 09 02:59:59 2013