Re: Inter-robot Communications - Part II
Mordechai T. Abzug (mabzug1@gl.umbc.edu)
Fri, 29 Dec 1995 15:09:51 -0500 (EST)
"DE" == David Eagles spake thusly:
DE>
DE> Well, I never expected to receive such a favourable response about a =
DE> standard port/protocol for communication between robots. Although the =
DE> work being done by Harvest was mentioned several times, the great =
DE> majority of people who replied thought the Harvest system was too =
DE> complicated now, and I believe it also lacks some useful features (and =
DE> it's not on a standardised port yet).
DE>
DE> I'm going away for a couple of weeks, but I'll put some thought into it =
DE> during that time. Any comments, requests, ideas for any aspect would be =
DE> greatly appreciated (after all ,that's how the Internet was built). =
DE> When I return I'll setup a part of my WWW server dedicated to this =
DE> project (think I'll call it Project Asimov - seems appropriate for a =
DE> global robot communications system).
DE>
DE> The key features I have thought of so far as listed below, so you can =
DE> comment on these also (ie. tell m,e if I'm being too =
DE> stupid/ambitious/etc)
DE>
DE> 1. Dedicated port approved as an Internet standard port number. (What =
DE> does this require?)
DE> 2. Protocol (similar to FTP I think) which allows remote agents to =
DE> exchange URL's, perform searchs and get the results in a standard =
DE> format, database mirroring(?), etc. The idea behind this is that if =
DE> Robot A finds a URL handled by another remote Robot (such as by domain =
DE> name, keywords(?), etc), then it can inform the remote robot of it's =
DE> existance. Similarly, if a user wants to search for something which =
DE> happens to be handled by the remote server, a standard data format will =
DE> be returned which can them be presented in any format.
DE> 3. A method of correlating Robots with specialties (what the robot is =
DE> for). An approach similar to DNS may come in handy here - limited =
DE> functionality could be obtained by using a "hosts" type file (called =
DE> "robots" ?), while large scale, transparent functionality would probably =
DE> require a centralised site which would maintain a list of all know =
DE> robots and their specialties. Remote robots would download the list( or =
DE> search parts of it) as required. This could probably be another =
DE> protocol command on the port above.
[snip]
Before doing anything new on robot/agent communication, you wish to look into
some of the already in-place efforts, ie. KQML and KIF. Check out
<http://www.cs.umbc.edu/kse>. They don't do everything you want, but they do
provide a framework.
--
Mordechai T. Abzug
http://umbc.edu/~mabzug1 mabzug1@umbc.edu finger -l mabzug1@gl.umbc.edu
If you believe in telekinesis, raise my hand.