[rfk-dev] code status

[Peter A. H. Peterson]

Quoting Benjamin Molitor:
> speaking of which, where do I get a current version of messages.h? The 
> one in sf.net CVS (which rfk.org points to) is rather... dated. I think 
> I remember some talk about moving to a repository on tastytronic.net, 
> but what are the connection parameters to pull a version from there?

Benjamin asked me this off-list, but it seems like everyone might want
to know.

1. I was working on an up to date POSIX tarball when the new codebase
came to light. It featured a messages.h with 731(!) NKI, most of which
came from Dave Griffith's IF ports as well as several more from myself
and Seth David Schoen. It's basically the Grand Unified NKI List.
Becuase that tarball wasn't released, and until we know what's going
on, I think I'll just put the messages.h out there:

http://robotfindskitten.org/aw.cgi?main=download/POSIX/

2. We discussed leaving Sourceforge and setting up darcs or something
here on tastytronic.net but with everything it didn't happen. I am all
for that unless people would *rather* stay on Sourceforge, in which
case we should update it. Discuss.

3. I was thinking about the whole "NKI in the binary" approach, and
there's a problem I hadn't considered: if no NKI are in the binary,
you could tailor your own setup to be any kind of rfk world you like.
All the NKI could be about pasta, or types of schist or something.
It's 100% open-ended (but closed beginninged -- there's still a robot
and a kitten). However, if NKI (even the first set, which I'm sure we
could extract from the rfk for DOS binary) are built into the binary,
your "robot finds kitten from amongst the pasta fields" adventure will
be interrupted by original NKI, which might be not part of your
vision. This lends me to imagine some kind of switch to skip the
original built in NKI, but then maybe that's starting to get
ridiculous. Discuss.

Peter

-- 
Peter A. H. Peterson, technician and musician.
 ---=[ http://tastytronic.net/~pedro/ ]=---