[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 'not an artist' article

On 05-Jul-2000 Steve Baker wrote:
> Chris Purnell wrote:

> I mean't the software structures that animators use with matrix weighting
> and such to animate characters.  There would be 'bones' in the characters
> hair, clothing, weapons, etc.

typical IK modeling? :)

>> > It's tempting (for example) to think of 'symmetry' genes - but does that
>> > overly-restrict or pre-judge the nature of the things you build?
>> I was thinking of enforcing left-right symmetry so we only have to
>> generate half a character.
> Well, I'd like to be able to generate characters like the hunchback
> of Notre Dame from the Disney cartoon.
> He is pretty asymmetrical - geometrically...but symmetrical topologically
> (if you see what I mean).
> I was thinking about maybe having one gene specify the order of
> symmetry (8 for an octopus, 5 for a starfish) - a gene to specify
> mirror symmetry - and a gene to control the amount of variation
> between two limits for the otherwise nominally symmetrical object.

a starfish may have 5 unique orientations that look exactly the same, but an
octopus does not, an octopus is symmetric, it just happens to have 4 appendages
on each side... iirc, starfish possess no internal organs, no centralized
functionality, were it do be damaged in a way that split a small part off, the
small part would develop into a complete starfish, etc. From a biological point
of view, they're far simpler than insect, mollusks, and simplest vertibrates.
(and while we may be more or less symmetrical externally, very few creatures are
fully symmetric, our internal organs are all over the place :)

> Dunno though - the more I think about this, the harder it gets.
> I think it will be necessary to start simple and add complexity
> once some experience has been had with such a system.

naturally, this isn't a common-place technology, there has to be some testing,
experimenting, etc :) if I had the time, I'd be coding RIGHT NOW :)

> If we had one that worked, it would be really simple to do
> Pokemon-like games with 150 cartoon characters.  That would
> be inconcievable for an opensource project without such a tool.
> (BTW: My son tells me that the next Pokemon cartridge for the
> GameBoy will allow you to cross-breed your Pokemon to get entirely
> new creatures.  That says they must already be thinking along these
> lines).

mmm, wonderful, teaching our children to discriminate and all about selective
breeding :) But are they doing gene-based morphing, or simple image morphing?
Following the 'morph fad', any monkey can select a few control points on a
couple pictures and render a composite... :) The exciting part about evolving
creatures with a genetic algorithm is you aren't limited to appearance, you
aren't limited to two gene donors, you aren't limited to small numbers of
generations... My mind is seeing 1000's of candidates, 1000's of generations,
and a set of 'base' creatures that can be stepped thru the algorithm to create
several unique creatures of the same 'species'... The only limites are the
machines available and the time to properly define the problem. If your game
takes place in a swamp, you build the swampy environment into the problem
definition, and you gets a set of creatures adapted to the swamp... 

        -Erik <erik@smluc.org> [http://math.smsu.edu/~br0ke]

The opinions expressed by me are not necessarily opinions. In all
probability, they are random rambling, and to be ignored. Failure to ignore
may result in severe boredom or confusion. Shake well before opening. Keep

To unsubscribe, e-mail: linuxgames-unsubscribe@sunsite.auc.dk
For additional commands, e-mail: linuxgames-help@sunsite.auc.dk