Álvaro Begué wrote:
On Sun, Feb 24, 2008 at 4:22 PM, Raymond Wold wrote:
It would be a lot more straight forward if there was just /one/ protocol
to implement. But I guess it's true what they say about standards being
nice.
In case anyone didn't understand this, I think Don is talking about an
old joke said: "Standards are great, that's why everyone has one."
Well it was me, not Don, and the joke I was thinking of was "Standards
are nice, there's so many to choose from!", but it conveys the same
sentiment anyway.
I like the current scheme where a little program talks GTP to the engine
and then something else (I don't care what) to the server. It would be
better if the little client were written in Perl (there used to be a
Perl version but I don't know if it's up to date) so I wouldn't have to
install another interpreter (TCL).
What language is preferred is dependent on the programmer, and sometimes
there's no choice at all, without switching API/framework/platform
entirely. Besides, open source or not, running others' code isn't always
wanted.
It would be so much simpler if everyone used GTP directly on the server.
I am aware that GTP lacks authentication features, but it should be
simple enough to agree on a command and say that that's now the standard.
KGSGTP has some of the same problems, but in my case that's more easily
solved since the interface code is written in Java and can be hacked
around with and be run in the same JVM as my engine.
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
