Previous: Earlier Versions, Downloads, ... Contents
These notes are intended to help anyone who decides to write code to implement these features. They have no special status. They simply represent the views of the authors of the notes. Maybe you can improve on them, or provide detailed specifications of the features that you want!
The basic idea is that we can replay a whole game without having to press any buttons. This will be a bit like watching a time-lapse film of a plant growing!
Two separate ideas here. Have a delay between display of moves which is related to the (taxicab) distance between them -- maybe approximately the logarithm (1 = 0, 2 = 1, 4 = 2, 8 = 3, 16 = 4, 32 = 5, 36 = 5.1!). In addition to this, if the (taxicab) distance is more than (configurable?) 6, say, then do a quick flash at the new location, before the delay.
It would be very nice to have an optional feature which stopped/paused at every branch point, let the user choose (using normal gadgets) the next variation to follow, and then Resume Auto-replay. Similar (but weaker) considerations apply to being able (optionally) to stop at any of Memorized Position, Next Comment, and Next Annotation -- for these latter two, we would want to have first "Goto Next Comment", and "Goto Next Annotation".
This is not the same as Auto-Replay.
At present, if we have an SGF record containing
a TsumeGo problem, we can test ourselves thus:
Assuming that the problem is Black (say) to play ...
This makes our opponent (the problem setter) White.
In view mode, tap a point to guess, as often as we like -- if successful then we will have found one of the variations that the composer has supplied. It will be White's turn. We can simply tap/press forward to get White's response (assuming only one!, - if not then we will see a variation indicated).
Optional extra mode such that PilotGOne autoplays the responses of the "other" (how to tell PG about this?) colour.
Evident difficulty: If the composer offers more than one refutation for a line of the user, then how do we deal with this?
A tight specification is needed! :-)
Rodney Topor (2002/08/24) made some comments about using this
feature with problem sets:
After the initial setup, I think it is preferable to
have all moves [in problem sets] by both sides entered as
B[]/W[] -- not as AB[]/AW[]. It's easier for users to enter
problems this way and it provides more options with
auto-replay off.
Where there is more than one refutation for a line of the user, here is a suggestion. Play the first refutation automatically and set the current variation indicators at the top right and in the move description at bottom to indicate that there are one or more alternatives. The user can then either continue with the current refutation or press the next variation gadget at the top right (or hard-mapped button) to change to the next (or previous) refutation. This seems like it would work. But I may be missing something.
Previous: Earlier Versions, Downloads, ... Contents
HotSync and Palm OS are registered trademarks of Palm, Inc.