refactoring of thread management
Currently we instanciate threads everywhere in the daemon. This makes capability dropping, proper thread shutdown and cleanup and module implementation tricky and unclear.
We want to provide a generic asynchron execution helper by introducing a "callback_job". This callback job gets a callback and is executed by the thread pool. This allows us to do all blocking activities in the daemon with these jobs.
Further we want to merge the job_queue/thread_pool into a single object: the processor. We do the same for the event_queue, it will be integrated into the scheduler for simplicity.