This is the mail archive of the xconq7@sourceware.cygnus.com mailing list for the Xconq project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Game customization



   Date: Tue, 5 Oct 1999 14:29:14 +0100 (BST)
   From: Keir Novik <K.E.Novik@qmw.ac.uk>

   On Mon, 4 Oct 1999, Hans Ronne wrote:

   > We also discussed other forms of customization, some of which don't even
   > exist yet. Right now, I am working on giving the independent side a brain,
   > so that these units also can move around and take part in the game. An
   > important question is how smart the independent side should be. Should it
   > be able to build, do research and at least some strategic planning, or
   > should it just restrict itself to defending its own units when attacked? My
   > feeling is that this may vary from game to game, and also that different
   > players may have different preferences. I would therefore like to put some
   > customization in the setup dialog for the independent side.

   As I see it, independent sides are the same as the other sides except they
   are trying to stay out of the conflict.  Something like Switzerland in the
   Second World War, I think you'd agree.  Thus I think the standard AI code
   should be modified to include them.  Independent sides should definitely
   build units and place them defensively, but should not move units into
   other countries.  They should be capable of acting offensively to push out
   an invader in their country.

There's a little bit of confusion here I think - what you're talking
about is really a neutral side, as for instance Spain and Turkey in
ww2-eur-42.  The units are displayed with emblems, they're (usually)
grouped together geographically, the side shows up in the side list,
units can help protect each other, they can share view data, and so
forth.

Independent units are more like barbarians in Civilization or the
animals in africa-1850.  They are scattered randomly throughout the
world, have no notion of a territory, cannot communicate or coordinate
with each other, and may not even know of each other's existence.
This is reflected in GDL, where the game rules for independent units
can be quite different from those for units belonging to a side.

The key distinction for game design is the lack of coordination.  If
you let all the independent units band together, and say there's no
difference between the independent side and any other, then there's no
mechanism to have truly independent units.  So then you have to add
mechanism to be able to modify the AI to deliberately cripple itself
for a particular side, and worse, every future AI that might play the
independent side will have to be written to take this into account.
This is a big change in the AI, which right now is like the human
player in that it can try to do anything it can think of - if some
action is disallowed by game rules, the kernel says "sorry, no can do"
and the AI can try something else.  If AIs have to also enforce game
rules themselves, things are going to get more complicated and less
reliable internally, especially for networked games.

So while I'd like to see independent units be able to act on their
own, I think we have to be careful to ensure that the behavior is
handled in the kernel in a network-friendly way, rather than relying
on all the AIs to know and follow the rules for independent units.
I'd also like to see the mplayer extended to handle neutral sides, so
that it keeps its country uninvolved (but well-prepared) until
threatened.

								Stan

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]