Re: [Bug-gnubg] Clear analysis bug ?

[Christian Anthon]

Should have been !pmr->ml.amMoves, thx for pointing it out.

Christian.

On Sat, Aug 22, 2009 at 8:32 AM, Philippe Michel<address@hidden> wrote:
> On Fri, 21 Aug 2009, Christian Anthon wrote:
>
>> Should be fixed now.
>
> The changes caused other problems : now, when loading an already analyzed
> match, the checker play errors are not marked. Only the cube errors are.
>
> The "|| pmr->ml.amMoves" below looks incongruous. Removing it helps but I
> don't really understand the code there. Maybe you meant || pmr->ml.amMoves
> == NULL, or something other than amMoves ?
>
> --- gnubg.c.orig        2009-08-21 21:55:56.000000000 +0200
> +++ gnubg.c     2009-08-22 08:22:00.000000000 +0200
> @@ -2225,7 +2225,7 @@
>  {
>        move *move_i;
>        move *move_0;
> -       if (pmr->n.iMove >= pmr->ml.cMoves || pmr->ml.amMoves)
> +       if (pmr->n.iMove >= pmr->ml.cMoves)
>                return SKILL_NONE;
>        move_i = &pmr->ml.amMoves[pmr->n.iMove];
>        move_0 = &pmr->ml.amMoves[0];
>
>
>