This is the mail archive of the xconq7@sources.redhat.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]
Other format: [Raw text]

Re: code in adjust_tooling_crossover?


>On Wed, 2002-07-10 at 11:39, Stan Shebs wrote:
>> But clearly nobody plays postmodern (the only game using tooling
>> crossover), because the formula is indeed wrong.  Both of the
>> tp_max tables should be functions of the unit tooling up, so
>> the args should be unit->type, u3 and unit->type, u2 respectively.
>
>Looking at postmodern, I can see another bug in the AI code.  It seems
>that the good_cell_to_colonize function does not consider whether or not
>the unit to be built will vanish or wreck on a tile, and so the AI will
>instruct engineers to try to build cities on ocean tiles!  And, of
>course, because cities cannot be built on ocean tiles, the engineers
>will go into reserve and try again the next turn, repeating the cycle
>until either the end of time or until someone destroys the engineers
>(whichever comes first).

This does not make sense. The AI does not instruct colonizers to build a
city in a certain cell. The good_cell_to_colonize code (which strictly
speaking is not part of the AI code) merely inspects a cell after the
colonizer arrives in it and checks if it is a good place to colonize. So it
is quite impossible for engineers to ever consider building something on an
ocean tile. Unless they can walk on water ...

The problem with units going into reserve mode each turn is, however, real
and happens in other games and with other types of units, too. A good
example is flattops.g where half the fleet on each side usually ends up
doing this. As a consequence, it is rather rare for an AI-against-AI battle
to be faught to conclusion in this game. This is a rather recent problem
(meaning it appeared in the last 2 years or so) since the AI always used to
fight flattops games to the bitter end before. I'm not sure what is causing
this. In some games, I have seen lots of units going into endless repair
sessions, but this does not seem to be an issue in flattops.g. It rather
seems to be a distance problem. Units that are too far from the action
somehow loose interest in the game, and end up doing nothing.

Hans


Hans Ronne

hronne@pp.sbbs.se



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