Re: Level 1 Working! (was: Some compiler issues and a request for debugging/design help )

Fantastic! :-)


On Thu, Jun 20, 2013 at 7:38 AM, Paul R. Potts <[email protected]>wrote:

> With some helpful hints from you all, the Arctic Slide implementation is
> working (for level 1, at least).
>
> Most notable changes:
>
> - Using 2-D array without "limited" works OK so I'm using that
> - Relaxing some types here and there
> - Using singletons on board, but dispatching only on types
> - <edge> is a type to indicate interacting with the edge of the board;
> making it a type reduces the number of special cases and checks
> - <pos> is allowed to be #f in some methods that handle <edge>, using
> <pos-or-false>
> - Tile class hierarchy is considerably changed to try to find a design
> with the least confusing overlap between most-specific/least-specific
> dispatch cases.
>
> I'm not sure all possible tile interactions are perfectly covered yet but
> it's a big improvement. The problem has become interesting again (that is,
> what's the optimal class design, how is dispatch determined from most
> specific to least, etc.) rather than tedious. I'll be blogging about it
> some more ASAP for anyone interested.
>
> Enclosing the latest if you want to play with it.
>
> Paul
>
>
> _______________________________________________
> hackers mailing list
> [email protected]
> https://lists.opendylan.org/mailman/listinfo/hackers
>
>

_______________________________________________
hackers mailing list
[email protected]
https://lists.opendylan.org/mailman/listinfo/hackers