Re: IGS client server protocol.

[rbrown@xxxxxxxxxxxxxxx]

Paul Leonard wrote:

> But less of this waste of bandwidth. I seek pointers to the
> specification for the IGS client server protocol so I can hang a program
> off cgoban to observe and analyse games.  I did a web search and browsed
> about a bit but with no luck.

There are no specs.  As far as I understand it, all the client writers had to
use trial-and-error to discover for themselves the behavior of the server.

If you telnet to port 6969 with a vanilla terminal, but "toggle client on"
you can see the "client mode" strings that igs sends when you type a command.

If there are codes that the client can send to the server _instead_ of typing
a command, they are the world's best-kept secret.  I suspect they don't exist.

I don't think there _is_ a true client-server protocol.  There's only an
interface to a command-line driven server.

My understanding of this could be way off-base; I'm not trying to misrepresent
the situation.  If some reader knows better, please correct my misapprehension.

You could try contacting Hui Huang, John Bate, Bill Schubert, or Rene Grothmann
to see if any of them know anything about an igs client/server protocol.  There
may be other client writers out there, too, but those were the four who first
came to mind.  Maybe they know the secret.  Shhhh!

Rich
-- 
Richard L. Brown              Office of Information Services
rbrown@xxxxxxxxxxxxxxxxx        University of Wisconsin System Administration
rlbrown6@xxxxxxxxxxxxxxxxx    780 Regent St., Rm. 246  /  Madison, WI  53715