Protocol-Free Communication
Protocol-free communication means communication that has no high-level content protocol. A hardware protocol, a low-level content protocol and a communication-language are required.
This is very similar to human communication: To communicate we are dependent on a hardware protocol like air or a piece of paper and a low-level content protocol, which are in most cases spoken or written words. We do not need to agree on a rigid protocol specifying what words in what order are acceptable in order to be understood. Not saying "good morning" first when you order a coffee might be rude, but it will not stop the waiter from understanding you.
Samarai brains communicate by exchanging messages that are Havel-expressions packed as srf-sources, serialized into a binary stream that is transferred using TCP/IP. Incoming messages are processed depending on content and configuration. The sender has the full power of Havel's expressiveness to formulate its request which can contain conditions, contextual formulations and much more, if security allows it even dynamic formulas.
Havel's modularity, semantic hierarchy and expressiveness allow messages to be relatively generic, very much like natural language. For example, if you call a company to ask how many items of a specific article they have on stock you do not need to know how that company is internally managing its stock nor do you need to find out prior to your call how exactly you need to formulate your request in order to be understood. To a certain degree, Havel allows the same. You can express a generic message containing a request for the stock quantity of an item. The recipient should able to process your request and send you a generic reply which in turn you should be able to understand.
With the first Havel generation, these communications will be relatively simple and reduced to a finite set of functionality especially for two-way communication strands. However, Havel has, so we believe, the potential to enable complex two-way communications that will enable two computers to chat with each other. Call, request, negotiate and order, all on their own. Later language generations should become far more flexible. It needs to be noted that these features are not limited to order-workflows but that they are generic features that can be applied to a very wide field of domains.
This post is part of a comprehensive introduction into SRDM, graph languages and Havel.
Continue reading: Collaborative Computers
Comments