RE: computer-go: Interfacing 2 engines for automated play

["Fant, Chris" <chris.fant@xxxxxxxxxx>]

Thanks for the info.  Don't worry, nothing will discourage me at this point.
It's pretty much all I think about.   I'm in the process of reading through this
list's archives, starting w/1998.  I have two more interfacing questions for the
group:  

1.  What interface (go modem protocol, local IGS server, local NNTS server,
OpenGo, or CGoban) will have the lowest overhead and be able allow the fastest
play?

2.  Do the good Go engines usually make these interfaces available to the
public?  Because it seems that if you made a living from writing Go software for
public consumption or competition, you may not want that kind of insight into
you engine accessible.  In other words, if someone did make a learning Go
program good enough, it could be trained specifically to take out a difficult
opponent.