Check out Symmetric Chess, our featured variant for March, 2024.


[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Earliest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

Earlier Reverse Order LaterLatest
Mutators. Article discussing the concept of Mutators.[All Comments] [Add Comment or Rating]
(zzo38) A. Black wrote on Tue, May 20, 2008 04:04 AM UTC:
I like this idea. However, I would make the mutators functions instead and use function notation, for example M(G) instead of G[M]. Limit is not a mutator it is a function that returns a mutator, so Limit(50) is a mutator. You could make a game also a function, which takes a move as its parameter and returns a game starting from the new position. You could also make something which is both a game and a mutator (so moves in that game would also be games). My idea is a strange idea. also, you could say other things as well: A mutator M is redundant with game G if G=M(G).

Doug Chatham wrote on Tue, May 20, 2008 10:44 AM UTC:
Do you have any example of an atomic mutator? It seems to me that all mutators are composite: for example, [Pass] could be factored as [Pass, if White][Pass, if not White].

Anonymous wrote on Fri, May 23, 2008 09:09 AM UTC:
zz38:

Yes, the function notation seems appropriate. This webpage, written some
years ago, was just a draft start for a more formal discussion on how to
make a mutator-oriented language for game specification. I didn't make
much progress in this sense. It seems, at least for me, a hard subject,
since there are so many different rules and ideas that games possess.

Doug:

Probably the only atomic mutator is [Void], ie, an empty canvas where you
start the composition of mutator to make games. Eventually, it would be
nice to have Macros, like [FIDE chess] or [Go] where the programmer could
easily build up new variants.

Rich Hutnik wrote on Fri, May 23, 2008 05:32 PM UTC:
I am of the believe the variant community should do more in this area in creation, rather than spin off every single mutator as a separate game.

📝João Neto wrote on Fri, May 23, 2008 06:14 PM UTC:
A mutator language must be typed. Not all mutators go with each other. Some mutators deal with turns (like progressive mutators), others deal with pieces (eg, apply [Real] to chess king], and so on... This should ease the job of the programmer as well as the language designer.

Rich Hutnik wrote on Fri, May 23, 2008 07:21 PM UTC:
There is likely a lot of work that needs to be done. I see it as a larger project that would be part of the future of chess variants actually, and chess itself.

Jianying Ji wrote on Fri, May 23, 2008 08:50 PM UTC:
Welcome back Joao!

On mutators, the way I look at it is that a game consists three things: A board along with its topology, A set of pieces along with their movements, and a set of mutators defining various aspects of the game not covered by previous two.

Rich Hutnik wrote on Sun, May 25, 2008 03:15 PM UTC:
May I offer up this page as a way to break out chess into its elements, to be able to see where you would classify mutators:
http://abstractgamers.org/wiki/definitions-of-abstracts

joaopedroneto wrote on Mon, May 26, 2008 10:28 AM UTC:
Hi Ji!

I'm not sure if we would like to see the board as a separate entity from
the game pieces. It must be thought over. A board can be seen as a set or
lattice of special pieces, that for most games are idle, neutral and
non-interactive, but in some games they interact. Some games allow players
to move cell boards, or cells can be removed (like in Zertz), or cells act
over pieces (like games with different terrains)...

📝João Neto wrote on Mon, May 26, 2008 10:29 AM UTC:
Hi Richard.

Thanks for that taxonomic work. The harder part is too make those concepts glue together. How to translate those ideas into types and how to compose those types into a game.

Rich Hutnik wrote on Mon, May 26, 2008 04:46 PM UTC:
João, thanks for the reply. My hope is there could be a 'Chess of Tomorrow' project taking off that could do all the classification work, and so on, and help plan this stuff out. I know IAGO is vested in such activity. Someone have asked why such granularity is needed. Well, my take is that such is important to be able to run tournaments, and help to be able to cause something as diverse as chess variants to get traction. Also, coming up with a system that has mutators, will enable people to propose elements, without having the usual 'hey, let's do ANOTHER GAME based around a single idea'.

George Duke wrote on Thu, Jun 25, 2009 04:58 PM UTC:
Taxonomy of Mutators from 2000 needs no piece-type naming. Contrast 'Man&Beast' of Gilman with 'Mutators' of Neto for nomenclature, one extending it to conceivable limits and the other without it, not even naming one piece here, but instead giving titles to dozens of general-purpose Mutators.

George Duke wrote on Fri, Jan 4, 2013 07:02 PM UTC:
A. Black comment here 20.May.2008 calls for function notation for Mutators, and Rich Hutnik calls for extending Mutators as system for play. Current Tandem-Pawn Chess is Modest CV because based on a Mutator widely applicable. As soon as the tandem-pawn Mutator is reduced to practice, it can apply to majority of pre-existing cvs, not only OrthoChess 64 squares. Neto develops or defines Mutators for CVs this reference article.

(zzo38) A. Black wrote on Fri, Jan 4, 2013 11:35 PM UTC:

Let's see if these mutators could be made with Haskell programming language, somehow; and/or just some general way in some mathematical category of CV mutators made up for this purpose (and you can see what are functors, monads, comonads, etc applying to).

In both cases, and in the mathematics in general, the NIL (identity) mutator should in fact be called the mutator, and it may form a monoid (if some mutators may make the game that some other mutators are not applicable to, though, then it may form a category but not a monoid).

If it is a category as above, and the objects describe features which are compatible with certain mutators, there might be an endofunctor to specify what applies to others too, and it might be a monad too (if it can be lifted into the set of additional features while keeping the same game, for example). In such case possibly even the games becoming mutators, being a morphism from a "null object" (meaning no features apply, so you have no game at all), to the object of their features.


14 comments displayed

Earlier Reverse Order LaterLatest

Permalink to the exact comments currently displayed.