[Retros] 50-moves rule and mate

Rol, Guus G.A.Rol at umcutrecht.nl
Mon Jan 15 09:11:02 EST 2007


Hi Andrew,

I'm all for the regular maintenance of the rule/convention system, in
the same way that I support mandatory expiration dates for most laws.
Keeps everyone on his toes. The pitfall is that the handling gets even
sloppier than it is today: "let's give it a try now, we'll have another
go at it next year". I believe it is quite feasible to build robust
composition laws that will stand the test of time as long as they come
from reflection as much as incident management. After all, Euclidian
geometry worked fine for over 2000 years. From my experience as a
computer programmer I learned to never design a program with the precise
specs required but always a bit oversized, a bit more flexible, a bit
more extendable. Today this design approach is known as "scalability".
In chess law a higher degree of scalability can be achieved by making
rules more generic, and by improving the concepts behind them. The
continuous evolution of chess composition guarantees that, even with
well-defined axioms, there will always be plenty of work for a permanent
committee on the subject. As an example, such a club could - for the
sake of orthogonality - have entensive discussions on a 50-moves rule in
Circe. What would happen to the idea of the (ir)reversibility of moves?
The joy of anticipating such meetings must be unimaginable!

I agree that defining retrograde and retrograde classes is important, if
only to determine which rule set applies to a composition. I currently
have no definite ideas on this subject besides the notion that there is
an aspect of "fact-finding research" in retro-analysis in contrast to
"application technology" in other chess variants. Many compositions are
of crossover types.

Like Tom, I am not much in favor of reappraising the validity of older
problems. It is however unavoidable that they will get reclassified as
time moves on. How would the reader know today that a bishop moves 2
squares indian style unless it were written somewhere near the
stipulation?

Guus Rol.



> -----Oorspronkelijk bericht-----

> Van: retros-bounces at janko.at [mailto:retros-bounces at janko.at]

> Namens andrew buchanan

> Verzonden: vrijdag 12 januari 2007 7:53

> Aan: 'The Retrograde Analysis Mailing List'

> Onderwerp: Re: [Retros] 50-moves rule and mate

>

>

> Guus, Tom,

>

> I am inspired by your dialogue. Creativity, civilization &

> sanity are all built on a healthy balance between Law, Chaos & Humour.

>

> If I was able to change *one* thing in the Codex, I would add

> an article to support Change. At the moment, the Codex is two

> versions out of date in terms of the Chess Laws that it

> addresses, and this reflects the fundamental fudge going on.

> The document is more or less frozen. It's not keeping pace

> with the (very slow & minor) changes in the chess world, nor

> is the Codex itself moving forward from what was a decent

> first draft to something which is robust enough to build

> proper tricky problems on.

>

> A fundamental sticking point I suppose is "Yes, but if we

> change it then what happens to a few old problems we are

> attached to which are thereby rendered unsound (fairy, joke...)."

>

> But we *have* to be able to move on.

>

> I suggest:

>

> - The Conventions be periodically updated, at least as often

> as new versions of the Laws appear (only every 5 years, so

> this is no big deal).

> - The Conventions be released no later than a year after the

> new version of the Laws.

> - By default, the new version of the Convention will apply to

> all compositions.

> - Common exceptional classes (e.g. retrograde problems

> handling of 50 move

> rule) must be specified in the Codex. There are some other

> classes I can think of too. But need careful definitions.

> E.g. what is a retrograde problem?

> - Anything else is simply understood to require an old

> version of the conventions, but if republished anywhere would

> probably be given an annotation.

>

> Comments welcome,

> Andrew.

>

>




More information about the Retros mailing list