Hi Luiz, > Von: Luiz Augusto von Dentz [mailto:luiz.dentz@xxxxxxxxx] > > I was about to commit this but then I realize the code under test/example- > advertising does not actually uses ObjectManager and still works, so even > though GetManagedObjects does return an error the code seems to cope > with it just fine so Im not sure what exactly is your bug? I reverted all changes to advertising code to verify and agree: as you claimed --> it works. I'm really not able to reproduce what went wrong anymore. After playing a bit with the changes I recently made, my guess is the error was due to some other bug in my code. When reporting as a bug I observed "failed to parse advertising" errors on the dbus and parsing problems for "Type" within bluetoothd's debug output when registering advertising using the objectmanager's path. On the other hand when advertising using the LEAdvertisement1 path I got another bunch of errors. Those boiled down to something "no connection"-ish from my different dbus-proxies (including the working GATTManager1) that occurred only when I made the call to RegisterAdvertisement. My guess is some flawed data/code within my test-advertisement and I most likely cleaned it somewhere in the process of patching the "evil" part of the advertising.c Responsible for the observed dbus errors. As my problems were gone after patching the blueZ part, I obviously got on the wrong track. I'm sorry for the confusion. I think you should commit the patch anyway, even if it's now "just" a minor improvement towards cleaner code instead of a bugfix. Thanks, Markus ��.n��������+%������w��{.n�����{��� �^n�r������&��z�ޗ�zf���h���~����������_��+v���)ߣ�
