[Freeciv-Dev] Re: movement

[Raimar Falke <rf13@xxxxxxxxxxxxxxxxx>]

On Wed, Aug 14, 2002 at 05:07:58PM +0200, Thomas Strub wrote:
> On Wed, Aug 14, 2002 at 10:57:25AM -0400, Brandon Craig Rhodes wrote:
> > "Anthony J. Stuckey" <astuckey@xxxxxxxx> writes:
> > 
> > >   IMHO, we want to always fail the rand().
> > 
> > My suggestion was precisely that the rand() should always fail, but
> > instead of throwing away the remaining points commits them toward
> > completing the movement at the beginning of the next turn.
> 
> Beginning of the next round? So a the rand() would succeed every time .. 
> 
> Moving with a horsemen:

Points are internally multiplied by 3. So a horsemen has 6 points.

> rrpmrrrrr
> 123456789
> 
> r - road
> p - plains
> m - mountain

points:   113911111
position: 123456789

> Starting a 1:
> End of the first round the horsemen would be on 3 at the begin of
> the next round on 4 (Mountain) with 2 movepoints (or 2 1/3 - 3 ?) 
> and could move to 8.

My interpretation of Brandon's idea:
 End of first turn: at 3 with 1 point left
 Start of second turn: at 4 with 0 points left
 Start of third turn: at 4 with 6 points left
 End of third turn: at 9 with 1 point left
 total of 17 points spend

My interpretation of Per's idea:
 End of first turn: at 4 with 0 point left
 Start of second turn: at 4 with 6 points left
 End of second turn: at 9 with 1 point left
 total of 11 points spend

If we cap the move costs of all tiles at the move rate of the unit (6)
the unit has to spend 1+1+3+6+5*1 = 16 points. This is IMHO a fair
solution.

> With throwing away the movepoints the horsemen would be on 4
> (Mountain) Think thats a better solution. I don't like doubleturns
> and that would be a serverdoubleturn.

doubleturns?

        Raimar

-- 
 email: rf13@xxxxxxxxxxxxxxxxx
 checking for the vaidity of the Maxwell laws on this machine... ok
 checking if e=mc^2... ok
 checking if we can safely swap on /dev/fd0... yes
    -- kvirc 2.0.0's configure