[Freeciv] Re: siege

[Peter Schaefer <peter.schaefer@xxxxxxxxx>]

Well, that algorithm is too easy for my taste, a good algorithm should
somehow require factoring a 2048-digit two-factor number to find out
whether the tile is besieged

On 8/9/05, Sam Steingold <sds@xxxxxxx> wrote:
> > * Sam Steingold <fqf-zKKw517/mfD@xxxxxxxxxxxxxxxx> [2005-08-07 13:33:01=
 -0400]:
> >
> > Placing an enemy city under siege should be easier:
> > a city should not be able to work a square that is reachable by an enem=
y
> > unit in 1 turn.
> > this means that entering the area around the city so as to be able to
> > displace another player's city work should require a declaration of war=
.
> > alternatively, displacing worked squares requires declaration of war.
> > (squares covered by friendly units can be worked)
>=20
> for each tile and player, define "tile availability" to be a number from
> 0 to 1, and use it to determine the actual benefit derived from working
> the tile by multiplying it by the nominal benefit (food, trade &c) and
> rounding appropriately.
>=20
> tile availability is defined as follows:
>=20
> 1. if there is an enemy[*] unit on the tile, tile availability is 0.
>=20
> 2. if there is no enemy units on the tile, tile availability is defined
>    to be the probability that the player will win an all-out war for the
>    tile.
>    specifically,
> ....