[Freeciv-Dev] Re: AI strategy

[Petr Baudis <pasky@xxxxxxxxxxx>]

> > Very nice too. However in order to be able to use CMA in AI, we need
> > CMA in CVS and probably in agents/, not client/agents/. Then, it would
> > be cool to get AI to use CMA. However I'm still not sure how easy it
> > will be, I will have to look at the CMA more carefully.
> 
> IMHO it isn't easy to use it in the server. It wasn't designed for
> this. It was design for the normal client and civbot (which is still a
> client to client/agents is ok).
True, I didn't think about that :-). Still polluted with current concept.

> > > > But we have to do following:
> > > > 
> > > > 1, Remove all AI hooks from server code. Way still too much of them.
> > > > Settlers code is an example, but many others randomly around all the
> > > > code.
> > > 
> > > This isn't an requirement. At least for the time till the civbot can
> > > compete with the server AI, the server AI has to remain in the cvs.
> > True, however the civbot won't be able to cheat. Otherwise we have no mean
> > to distinguish player and civbot, so players can cheat too, and that's bad.
> 
> A non-cheating AI is ok. It may also be possible to set the "transfer the
> whole map to player xyz" flag through a server command.
It's questionable how well it will perform :-). We shouldn't remove the
server-side AI until AI-AI diplomacy will be at least partially done, as civbot
will otherwise be even more easy for expierenced players, IMHO.

The flag is nice, only unlimited rates are remaining issue where AI cheats
IIRC. However it should be able to survive w/o them. Anyway, should be civbot
coded from scratch (feel even anyone here like doing that? please stand up :)
or copied^Winspired by current AI?

Another issue, what about observing AI? Tuning behaviour and debugging will be
a lot harder w/o that.

> > Anyway, if server has to do the spawning, shouldn't he also do spawning of
> > civclient on demand, etc? I would stay consistent with the behaviour :-).
> 
> Yes and no. Yes this would be nice. No, since the server currently doesn't
> know any gui. But the user will want to select "Load game", "New game" trough
> a gui. And than the server has to have a gui.
Hmm, well, I would rather stay with client as GUI for game control.  Too much
complexity would be added otherwise - GUI frontends for server etc, ble.
However, when I think about it, server should be still able to spawn civbot,
either after use of aitoggle command or when civil war will rip someone's
empire. Too bad, but I see no elegant solution for this. So it looks civclient
will spawn civserver and civserver will spawn civbot.

-- 

                                Petr "Pasky" Baudis

UN*X programmer, UN*X administrator, hobbies = IPv6, IRC, FreeCiv hacking
.
  "A common mistake that people make, when trying to design
   something completely foolproof is to underestimate the
   ingenuity of complete fools."
     -- Douglas Adams in Mostly Harmless
.
Public PGP key, geekcode and stuff: http://pasky.ji.cz/~pasky/