I wrote: > Shouldn't we call the full activate_task(...) instead of > __activate_task(...) in the !sync case? > > > The attached patch fixes seems to fix it for me. It is against V0.7.48-05. This patch should work the same and is a bit nicer. Michal