RE: [Bug-gnubg] Game Record Marked Decision Problem

[Albert Silver]

This is a good idea. I was going to propose leaving the checker
decisions as are and only include this for the cube decisions, but then
I visualized the results and must agree your idea as stated is better.
I'd only propose a single space when stating an error to not overcrowd
the move and improve visibility. If no spaces are included it may look
like this:

55: 10/5(2) 8/3*(2)M!C

instead of 

55: 10/5(2) 8/3*(2) M! C

The question is whether this might not cause the game record window to
widen too much.

One thing: is there no way to create a square symbol as a character?
Does the Freetype (if that's what's being used) support this? If so
there is no need to translate the letters. One could simply place a
large circle such as an 'O' for checker play errors, and a square for
cube errors.

55: 10/5(2) 8/3*(2) O! !

I found this square above in the character map, but do not know if all
will see it correctly. Also I didn't find different sizes.

Albert

> Suppose we used
> 
> C! C c for cube very bad/bad/doubtful and
> M! M m for chequer play very bad/bad/doubtful
> 
> These are distinct and wouldn't occur in normal move
> annotations. Wether or not they are suitable for other languages is a
> different question. The choice of letters could be configurable. How
> it would work for languages such as Japanese is an issue I can't
> speculate on.
> 
> --
> Jim Segrave           address@hidden