Re: Death of an advanced unit

[Lincoln Peters <peters2000 at mindspring dot com>]

On Fri, 2002-06-21 at 02:36, Hans Ronne wrote:
> >   Comments?
> 
> I agree with your analysis. The code in good_cell_colonize was just a first
> stab at a working algorithm.
> 
> As for space colonizing games, which is how this thread got started, I
> think an alternative approach is to do away with colonizers altogether and
> instead seed the map with independent planets that you can capture and then
> further develop. There are two avantages with this. First, it is more
> realistic. You don't build planets in places you choose, you find them or
> capture them (or so I imagine). Second, the attack and capture code works
> exactly like Lincoln suggested the colonizing code should work, i.e. it
> looks for those few cells that are worth taking and goes directly there.

I did take time to look at "galaxy" and "space", which both define
planets as units, and I wanted to avoid defining planets as units
because that would make it rather unrealistic when dealing with
uninhabited planets.  That's why I defined planets as terrain types and
civilizations as unit types.  The only time that a planet actually
changes (as opposed to the civilization on the planet) is if it's
terraformed.

As for conquest, if there is a civilization on a planet, it can be
captured or destroyed, just like in "galaxy" or "space".  The problem
with making that aspect realistic is that nobody knows how populated our
galaxy is, and so the values I put for civilizations in
independent-density are completely arbitrary.  I could set up variants
to make the galaxy more or less densely populated, though.

And, of course, if a planet is uninhabited, it is easy to build a new
civilization on it.  Just get some engineers to that planet and start
building.  Of course, as we've observed, Brownian motion doesn't work so
well in space.