[Retros] way forward with 3Rep issue

Andrew Buchanan andrew at anselan.com
Wed Jun 4 13:49:12 EDT 2014


Dear friends,

Unfortunately this whole dialogue got going while I was on a long business
trip, and I've been unable to participate properly. There is now a long
backlog to catch up on. The emails are long and full of judgments, and
frankly I find some emails hard to understand.  I am sure that mine are
equally hard to grasp, but at least I have an advantage in knowing what I
was trying to say. :)

Rather than lengthy arguments for and against something vague, I would
prefer to read something short and precise (i.e. just a brief presentation
of proposed rules/conventions and algorithm for running them), so that
different participants can see what's exactly proposed. It would be great if
this can finally end up as a book, but in the mean time, I think we need
something much shorter so we can grasp exactly what is proposed. I will try
to find time to craft my own humble proposal at the weekend.

If the theory cannot be described in a couple of pages, then it's too
complicated. I do agree that we ought to have something which is open to
expansion to the fairy world, and any new thinking ought to bear that
scalability requirement in mind. But the first objective has to be to get
something workable for regular chess.

It is possible that there will prove to be two consistent theories. One of
them munges together the rules and conventions in what to me currently seems
horrible. The other tries to use conventions minimally and separately from
the rules, but Guus doesn't seem to like this. I think it is worth exploring
both, once and for all. If we find out that we only have one consistent
theory, then fine we have no issue. If both theories are consistent, then it
does become only an aesthetic question, and maybe we will then know enough
about what's on offer to be able to agree. Right now, I do not see that
either approach is necessarily more fruitful chessically.

Thanks,
Andrew



More information about the Retros mailing list