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

[Jim Segrave]

On Thu 20 Mar 2003 (05:51 -0500), address@hidden wrote:

I'm forwarding this to the mailing list for other people's opinions

> Hi Jim


> At the moment I'm only really concerned that move to next marked decision 
> doesn't bring up both the checker play tab followed by the cube decision tab 
> when there are both errors.

No, it does the opposite. If the cube handling is doubtful or worse,
then the cube page is active, otherwise it's the chequer play.

> I know that they are marked with !? etc and can be seen there but I'm 
> scanning through many many matches and it's bad on my eyes to look for the 
> markers.

That's not easy to fix - for a given turn, both the cube action (or
lack thereof) and the chequer play are kept in a single record in the
list of moves. So the code to move to the next marked entry always
starts with the next move in the list and scans forward for an entry
with either a cube or chequer play error. 

To achieve what you're looking for we'd need to:

Modify the logic to make a chequer play error on the same move as a
cube error the 'next' marked move.

Modify the GTKSetAnnotation function to have a flag which could
(sometimes) force the chequer play tab to be active even when there is
also a cube error.

Both are doable, the question is if changing the way in which errors
are flagged is sufficient.

One of these days, I want to add some options to the game record
window which will allow the next/previous marked to be restricted to
only cube or chequer and only player 0 / player 1, which might also be
more convenient.

-- 
Jim Segrave           address@hidden