Trent Piepho:
> On Sat, 8 Dec 2001, Kurt Siegl wrote:
> > > feature, and let unix clients ignore the server setting just like 
> > > the windows clients.  Since most players have continuous 
> mouse, it 
> > > can't be that borg of a feature.
> > 
> > We will have to fix the bug in the windows client and use 
> the server 
> > settings.
> So I take it that you are for revoking the keys used by 
> probably 80% of the current netrek players?

Naw, it's been fixed in my NetrekNT branch since 11-Jan-99.

I found what I did.  It was in mswindow.c.  Basically the problem was
that the features packet was setting a variable motion_mouse_enablable,
but it wasn't being validated in the mouse handling routines.

Here's the code fragment from my fix:

#if defined(MOTION_MOUSE) || defined(XTRA_MESSAGE_UI)
// SRS placed this here to fix the not being able to disable continuous
mouse functions
         if (!motion_mouse ||
             (!motion_mouse_enablable && !motion_mouse_steering))
           return (0);

         if  ((win->type != WIN_GRAPH) || (hwnd != LastPressHwnd))  
            return(0);  // only allow in graphics windows and must be
the window last clicked in
                        // (the latter avoids a bug where Windows sends
a mousemove message
                        //  to a window exposed when the one on top is

To be honest, I think it works but I really hate the way I fixed it as
it's not readable code.  The feature packets is another section of code
I wanted to rewrite in a clear and consistent manner.