Things that would be good. 

Hi Maciej,

My programming is not up to much and would have to be in BASIC so I don't
think I could contribute much to the development of Framsticks. But I do
have some idea's for things you could implement.

I suggested before that having a selection weight for keeping the centre of
gravity as high as possible would facilitate the evolution of better
walkers. As implemented this does not seem to be the case. It seems to be
much easier in evolutionary terms to stick your feet up in the air! I
should have seen that coming. So my new suggestion is this, it would be
good if after having built a creature you could then specify which stick or
sticks will contribute to the creatures fitness if kept up off the ground.
There could be several parameters for each stick specified i.e. Preferred
Height, Height Precision, Preferred Angle, Angle Precision, Preferred
Direction and Direction Precision. Also I feel that the inclusion of joint
constraints would be very useful, i.e. an evolvable parameter that specifies
the angle through which a join can bend or twist and even some sort of
buffer so at the end of the swing or twist the movement can come to an
abrupt or gentle stop. Maybe in the case of the gentle stop the buffer
could store energy like a spring and then return it later on.

I am sure that you have already thought of a million things that you would
like to implement and don't have the time. Anyway thanks for Framsticks.

Cheers,

Ander Taylor

I agree that affecting taller animals would improve ambulatory performance,
but I think there's a much simpler way to encourage this instead of lots of
manual formulas which try to select promising attributes-

Why not place the "food" on "trees"? This means that animals *must* reach
upwards to eat. These trees should always be breeding increasingly taller
versions. Seems to me this would automatically lead to a sort of
"Height-Race" similar to the evolution of the giraffe...

Would anyone be interested in working on this with me?

Jack Pryne
jianju@hotmail.com

"Ander Taylor" wrote in message
news:8n2jgt$1ah$1@net.frams.poznan.pl...
> Hi Maciej,
>
> My programming is not up to much and would have to be in BASIC so I don't
> think I could contribute much to the development of Framsticks. But I do
> have some idea's for things you could implement.
>
> I suggested before that having a selection weight for keeping the centre
of
> gravity as high as possible would facilitate the evolution of better
> walkers. As implemented this does not seem to be the case. It seems to
be
> much easier in evolutionary terms to stick your feet up in the air! I
> should have seen that coming. So my new suggestion is this, it would be
> good if after having built a creature you could then specify which stick
or
> sticks will contribute to the creatures fitness if kept up off the ground.
> There could be several parameters for each stick specified i.e. Preferred
> Height, Height Precision, Preferred Angle, Angle Precision, Preferred
> Direction and Direction Precision. Also I feel that the inclusion of
joint
> constraints would be very useful, i.e. an evolvable parameter that
specifies
> the angle through which a join can bend or twist and even some sort of
> buffer so at the end of the swing or twist the movement can come to an
> abrupt or gentle stop. Maybe in the case of the gentle stop the buffer
> could store energy like a spring and then return it later on.
>
> I am sure that you have already thought of a million things that you would
> like to implement and don't have the time. Anyway thanks for Framsticks.
>
> Cheers,
>
> Ander Taylor
>
>
>
>
>
>

If you want tall walkers, you need selective pressures that won't serve just
as well to give rise to jumpers. One strategy available in the current
incarnation of Framsticks is to define fitness as a combination of
horizontal and vertical distance, but this alone is insufficient to favor
tall walkers over high jumpers, I think. Similarly, fruited trees may just
as well favor jumpers.

Cheers,

Pete

--
___________________________________________
P E T E M A N D I K
Assistant Professor and
Coordinator, Cognitive Science Laboratory
Department of Philosophy
William Paterson University of New Jersey
265 Atrium Building
300 Pompton Road
Wayne, NJ 07470
(973)-720-2173
mandikp@wpunj.edu
http://www.wpunj.edu/cohss/philosophy/faculty/mandik

Would walking not evolve within Framsticks in a rough terrain simply because
it is easier to step over terrain features rather than crawl over them, and
as you prgress to being able to 'step' over larger terrain features you will
be more succesfull that crawlers?

I understand that if you think about it in terms of frams in a flat land,
you need to look at some 'incentive' to evolve 'taller' (Ie, pressure for
vertical velocity/position), but with a rough terrain with features whose
scale and the distribution of the scales of the features, is carefully
chosen, no incentive is needed other than that is better to be able to move
without having to crawl up and down every bump. An experiment who's terrain
became rougher over time would be a good environment, maybe?

In simulations I am running at the moment I am sometimes seeing frams evolve
that are like large 'V's with a protrusion of some sort on both ends of the
'V' (a stick or two at right angles). They move over a landscape by rotating
their bodies seemingly in alternate halves (Or trying to...) and their
'legs' help the cover ground by stepping over features of the terrain as a
consequence.

It is hard to keep the conditions just right to keep these Frams evolving
however - usually they are usurped by the small fast 'worms' who can cope
with being dropped into a larger number of locations without being 'broken'
in the process. (IE; they can begin to move in almost any terrain, whereas
the larger Frams have a greater need to be placed somewhere they can 'cope'
with - but if they get this nice start location they can do much better than
the worms)

I think the 'protrusins' on the large V like frams evolve because they
afford the V some kind of weak dragging movement if it does not end up in
terrain it favours. This helps keep the V in the GenePool.

Matt.

"Pete Mandik" wrote in message
news:a7lcmv$9nh$1@cancer.cs.put.poznan.pl...
> If you want tall walkers, you need selective pressures that won't serve
just
> as well to give rise to jumpers. One strategy available in the current
> incarnation of Framsticks is to define fitness as a combination of
> horizontal and vertical distance, but this alone is insufficient to favor
> tall walkers over high jumpers, I think. Similarly, fruited trees may
just
> as well favor jumpers.
>
> Cheers,
>
> Pete
>
> --
> ___________________________________________
> P E T E M A N D I K
> Assistant Professor and
> Coordinator, Cognitive Science Laboratory
> Department of Philosophy
> William Paterson University of New Jersey
> 265 Atrium Building
> 300 Pompton Road
> Wayne, NJ 07470
> (973)-720-2173
> mandikp@wpunj.edu
> http://www.wpunj.edu/cohss/philosophy/faculty/mandik
>
>
>

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.391 / Virus Database: 222 - Release Date: 19/09/2002