Re: [AD] Two custom mouse cursor patches

[Evert Glebbeek <eglebbk@xxxxxxxxxx>]

> There are two options
> 1. Don't do any retains. System cursors are fine because they're built
> in. Custom cursors are managed 'externally' by the lifetime of the
> ALLEGRO_CURSOR. We will need to get rid of the release on line 674
> (destroy_display) At the moment I think there might be a problem if
> the same custom cursor is active in two windows when it is destroyed.

Testing with ex_mouse_cursor, I don't see any problems in practice.  
Doesn't mean there aren't any of course. :)
I thought I might actually have too many retains in there, which  
would leak memory if a cursor was actually discarded.

> 2. Do a 'proper' retain/release whenever display->cursor is changed.
> This is a bit more work but would solve the problem above. I've got a
> patch for this and I'll post it later (I can't get to it at the
> moment)

That does sound like the correct approach, yes. Since you already  
have a patch for this I'll leave this to you (I can test it of course).

Evert