As of about five minutes ago, I just completed an iteration that involved re-implementing the Agent Device's disengage operation. Hopefully, these things are relatively stable and I can finally approach the Agent Collaboration's pulse message.
It is worthy of note that the engage and disengage operations differ in rather strange and curious ways. The engage operation is reciprocal: there are somewhat symmetrical programmatic structures in place for engaging Agent Devices; one for the initiator, and the other for the responder. On the other hand, the disengage operation is recursive; the same method that is called on the responder in reply to the initiator's request is called again on the initiator in reply to the responder's actions. Of course, the recursion stops at the initiator on the disengage call-back. Just how the engage/disengage operations are implemented is rather complex (it could be more complex than it needs to be), but I have tried to be as straightforward as I could. The code might be more complex than it needs to be because I travelled down a few blind alleys, and may not have completely removed work-around code when I could revisit it to put in the intended code after covering incidental issues. It might be the simple case that I was in a reciprocal mood when completing the implementation of engage operation and in a recursive mood for the disengage operation.
Maybe one day, if the implementation becomes open source in some way, a million or so brains will pack down the footprint.
Looking good...
No comments:
Post a Comment