Hi Tom,
I'm working on an even tighter integration where PD objects can be directly managed by object-oriented C or Python code, without the detour of message generation. This can then be used for dynamic patching or just another patcher system.
very interesting.
can you elaborate on this a bit more?
the project is still in the concept stage, but it will be a topic for the PD convention in Graz. In short words, it will be a shared library with a C-style API that acts as an interface to the PD system. It's built on top of the traditional API and i don't think that changes to PD are needed. It will manage construction, destruction, connection of objects and storage of associated additional data. Just as with dyn~ the objects will live in a sandbox, they will not appear in the tcl/tk patcher system.
i'm trying to do something similar with mole and pf, but i'm stuck at the point where i need to catch the objects's output and map it to a function return value..
In dyn~ this is done with "proxy" objects that are connected to the objects of interest.
best greetings, Thomas