Trying to build target angstrom-qt5-x11-image on Colibri-imx6

Is the build target angstrom-qt5-x11-image for Angstrom Linux v2.7 supposed to build? I use the following command $bitbake -k angstrom-qt5-x11-image and I get too many errors to count. I have an application I want to build and run and it must use qt5.

Yes, of course, as explained in the following article on our developer website. Is that how you went about it? Otherwise do let us know about your exact build environment and post log files and or console output (textual and no shabby screen shots please).

Yes of course I followed the open embedded (core) article. it describes building the angstrom-lxde-image and not the angstrom-qt5-x11-image. I am able to build the angstrom-lxde-image v2.7 and v2.4.

The overall build procedure should really be no different at all.

Hi Marcel. I wouldn’t think so but wanted verification. I removed the sstate-cache and tmp-glibc directories before building but maybe I need to delete more directories from the angstrom-qt-x11-image build before building the angstrom-qt5-x11-image. I didn’t realize the angstrom-qt-x11-image target would build using qt4 instead of qt5.

Yes, in the embedded space Qt 4 is still quite commonly used.

If you do want us to help you please reveal what exact issues you are facing.

Marcel, I removed the complete repo tree and executed a new repo init, repo sync and bitbake -k angstrom-qt5-x11-image. It seems to be building properly now. I want to point out that the Open Embedded (core) web page has some errors in it. For example the repo init commands are wrong as the specify http instead https. In addition the manifest for branch LinuxImageV27 needs editing to change the URL accesses from http to git or in some cases https. The manifest should be updated in its repo. In addition I still have problems fetching the virtual/kernel and the virtual/bootloader through the recipes. I can however work around this issue by using the git clone command cloning the repos into …/downloads/git2.and rebuilding.

What’s wrong with using HTTP exactly?

repo sync fails because it cannot access the URLs with http. I receive 18/18 URL fetches with git and https and none with using http. Are you suggesting my corporate network could possibly cause the failure issue? I can retrieve the output of the repo sync using the downloaded default manifest.xml if you want.

Yes, very much so.