[Freeciv-Dev] (PR#6931) Returning GR_FAILED in execute_orders doesn't cancel unit orders

["Jason Short" <jdorje@xxxxxxxxxxxxxxxxxxxxx>]

<URL: http://rt.freeciv.org/Ticket/Display.html?id=6931 >

> [i-freeciv-lists@xxxxxxxxxxxxx - Sun Jan 25 07:59:16 2004]:
> 
> On Sat, Jan 24, 2004 at 10:56:35PM -0800, Jason Short wrote:
> > 
> > <URL: http://rt.freeciv.org/Ticket/Display.html?id=6931 >
> > 
> > > [ali - Sun Nov 23 16:25:51 2003]:
> > > 
> > > Bug: maybe_cancel_goto_due_to_enemy() should idle the unit,
> > > like in maybe_cancel_patrol_due_to_enemy().
> > > 
> > > Otherwise, unit retains the goto state when a non-attackable unit
> > > unexpectedly exists on a waypoint (only when using waypoint).
> > 
> > The goto (now orders) returns GR_FAILED.  If this happens then the goto
> > should be canceled; the unit will be left in idle.  It looks like the
> > goto/orders function expects the caller to do this, but of course the
> > caller doesn't.
> > 
> > Execute_orders should just return a bool indicating whether the unit
> > survives.  Any other handling should be done within the function. 
> > Perhaps a wrapper function would be helpful (to handle the different
> > return values).  If not then the GR_*** enumeration can just be dropped.
> 
> I'm for the wrapper. Else the functions grow too much.

I think a wrapper is overkill.  But we may have a sub-function or
sub-macro to do the free + notify line.

jason