[pbmserv-dev] M50 notation

["Randall Bart" <barticus@att.net>]

>                         3     
>                     2       7     
>                 1       6      12     
>                     5  ___ 11      17     
>                 4     / X \___ 16          
>                    10 \___/ X \    21      
>                 9     / X \___/20      26  
>             8      15 \___/ X \___ 25      
>                14  ___/ X \___/ X \    32  
>            13     / X \___/ X \___/31      
>                19 \___/24 \___/30      36  
>            18      23      29      35      
>                22      28      34              
>                    27      33                  
>                                  
> 
> No ambiguity.  No arbitrary start location.

This is no better than Cameron's scheme.  The problem isn't whether a computer can understand it; the problem is human use.  David doesn't want to play a game if he can't study games.  The current notation is not helpful in discussing the game.  If a threat is created on turn 5 then resolved on turn 10, I would like to discuss it's influence over moves 6 thru 9.  How do I refer to this place on the board, if it has a different number every turn?

So what David and I would like to see is a notation where the numbers don't change.  For all of these boardless hex tile games I offer what I will call M50 notation.

1)  The plane is divided into a hex grid, with each cell known by a letter and number as in Hex.
2)  The location of the first tile is M50.  Letter and number go up and down from there.
3)  If the board grows past A, the next row is ZZ, and continue YY, XX, etc.
4)  If the board grows past Z, the next row is AA, and continue BB, CC, etc.
5)  If both (3) and (4) occur in the same game, that's okay.  That would be at least 28 hexes wide.  How often does a game get 28 hexes wide?
6)  If it matters, 0 wraps to 99 and therefore 99 wraps to 0.  The standard tile set is 48 tiles, so you really can't get there.  You can't wrap a double alphabet either.  But just in case someone is playing Palago or Dragons with hundreds of tiles and manages to create a huge board, it actually turns out you are playing on a 52x100 torus.  (I want to see that game.)

> On Mon, Sep 21, 2009 at 04:37:53PM +1000, John Farrell wrote:

> > (b) equivalent board positions have identical descriptions.

> > To achieve (b) you need some algorithm to canonicalise whole board
> > descriptions.

Just canonicalize the first move as B2.  You will miss some duplicated positions, but not many.  I think the rules should just require this as the first move.  There's no need to have four different variations of that move, and the other 14 starts have been solved.