Ocamm´s Principle

Turing cracked the German Enigma code using reaction diffusion, hence Cracking Code Zebra.

This is Sara. I am writing in part by candlelight so that I can see the mountains in their glory. My table has its banquet leaves in and is imposing, covered by a sumptous green turtle Indian table cloth, my momento from , recent Australian travel was turtle themed, I realise. I worked in a former turtle sanctuary and war zone, the Alchemy participants became turtles. Less lycanthropy than context.

But on to another animal: the zebra.

Daniel believes that she must be Open Source: She must Stand alone, download browser onto the client side and affinity plug-ins, leave hooks and in the future decide how to add stuff Netomat, Napster, downloadable browser 50 millions people using it. We need to download cause its about speed of access. Then you would have r and d engines and patterns generation on the server.

We (Marcelo and Daniel) all think that perhaps Code Zebra should roll around continents as an executive motor home, preferably an Air Stream, with ocelot markings etched into the metallic, borg patterned lining and plush carpet all in animal and other cellular authomy patterns. That is where we lock up the artists and the scientists, that is where we stream from, that is the source of reaction.

Now here comes our computing biology lesson. It is important to memorize this because Marcelo will test all of us by creating patterns for the site.

Daniel suggests that we build an open source stand alone, downloadable browser, download onto the client side and then download plug-ins, like affinity plug-ins, leave hooks and in the future decide how to add stuff. Netomat, Napster, downloadable browser 50 millions people using it. I contacted Netomat with the idea after this suggestion, to be continued. What is needed includes Visual representation, text interface, iconic representation of participants. Daniel thinks we need to download cause its about speed of access. Then you would have engines and patterns generation on the server.

Ocamm´s principle: follow the proof (and path) of least resistance and its likely to be right. The Rule is the Simplest System. In the instance of explaining spots, Clonal Mosaic Processes seem to be the rule of the day. But there is hope for r and d, if only as a metaphor.

Basic Idea, with somewhat Portuguese grammar structure thanks to Marcelo: Fur will get colour. For some reason evolved so that pigment is not distributed evenly. Evolution process‹computer science not solving this bit, but the biologists are solving when they are interested, we know that 32 genes are involved in patterning, from mice, l991 survey article, listed all the genes‹the 32 that effect patterns in mammals. Some control distribution of pigment cells in the body. In giraffe, some pigment cells did not make it.Different genes control different parts of the process. But what is the process? Isolate one gene for particular transmissions. Genes that control more than one feature are:
Plyotropic
Control splitting rates of cells. Brown cells split fifteen times faster than background cells. Explain the pattern and the thickness (why it is thicker where there is pattern, you can feel it on the spot).

Idea of reaction/diffusion is that there are morphogens (generators of form), Turing is to thank for this, that Code Cracker. These would be on the surface. Marcelo heard the mention that this was recently proved for embryo, but this would be new, what is the ³body² would be the ³chemical substances², the latter most often used for reaction/diffusion testing. Both agree that see translation of pattern underneath the surface. One theory proposes a chemical system, the other a cellular arrangement. (Marcelo and not r and d). Marcelo started with reaction diffusion, but moved away because of Ocamm´s principle to
Voronoi or
Clonal mosaic patterns
Or
Cellular autonomata: cell not related to real biological cell, rules to evolve behavior, abstraction of rules from local Two D sea of cells, if neighbours have properties, set of rules well defined, but have patterns that emerge unexpectedly. Not clear that will emerge. Global pattern comes from individual cells and their local rules, the problem is that you do not know what you are going to get. ­ at a certain threshold effect each other.

If all patterns that R and D can do and all patterns that CM can do, what is the relationship between these. i.e. They can both do spots, measure the power of generating patterns.

True Voronoi, geometric construction, are seen in the area inside the cell, the smaller the areas, the more the polygon is Voronoi, used the measure to try to tell how close they were to the source pattern, results were compared. Vertices of true voronoi has to be less than one percent. First drew 2d, realized it was onto something that was 3d. Discussion, further than visual, subjective measurement, wanted it to look qualitative ­ if filled with basic structure get similar pattern with leapord and giraffe. Pattern on the surface. Like a fractal, since fractals have self similarity. The voronoi cells get together to create an overall pattern. On the surface. Cells centers. At some point in the development of the giraffe, there was a real voronoi, the animal grows at different rates, the older the giraffe, the longer the neck the further you are from the true voronoi. Theory is that you measure the geometric construction from the original pattern, provides a number.

Contacts with biologists‹would suggest that mathematical biology, come up with mathematical models to explain phenomena, some biologists do not have training, so do not grasp the intuition, complaint of reaction/diffusion. Clonal Mosaic Patterns. The voronoi‹giraffes, How to translate point representation into 2D representation. The second was, how to prove that patterns are correct as well as who the visual. Looking for a metric to prove that

Fur Fetched Facts:
Genetic background: pigment cells migrate into the follicle under the skin and modify it. MSH mealonocyte stimulating hormone. Stimulates production. From yellow to black. Know which gene controls production of hormone, Acouti gene. Threshold, either produces yellow or brown, black. If enough of hormone produces.

Biological basis of CM, how fast the cell splits is also an issue in its research. Cat. Big cats, if touch spot, higher, so for their model need larger splitting rate for that area, layers of cells.

Patterns: black and white patterns represent yet different mechanisms. Not reaction diffusion, but in fact, for cows it is another gene, not a pattern. Holsteins are ³failures², they would be all white except for gene failure in the area.

What about zebras: they might ACTUALLY be reaction/diffusion. All regular, not like cow. What a relief!

Affinity group discovery: r and d U. of Washington: forest fires were random yet became a closed system, fire fighters: after trees, no trees. Fires burn out their pattern, abutted onto another whole fire. Burn out mostly within boundary of other fire.

Dirichlet: adds triangulation to the voronoi, like a duality principle, means that if gets one, can great the other. If know the voronoi, then computer can find the domains.

There are other rules, duality principle, Bolian principle, exchange n or r, if know identity true, duality holds for two, makes proof easier. Prove for other half. If you know one stripe, you know the other. Philosophical interpretation: If believe in another dimension, then it is provable, or it´s the other side, provable here, can prove there.

How to get things named after you and is it still possible: According to Marcelo, most algorithms are named after the inventors, in the early 1970s there was nothing, so everything was considered new. e.g. Sutherland #34. Cause Sutherland discovered lots of small stuff way back then.

Marcelo reckons that its too late to get a software with your name on it, though we CAN patent kitties with your name spelled on in fur patterns for sure. ³Lions make leopards tame.² Shakespeare Yeah, but not change its spots.

What is a Voronoi? 3D tetrahedron space, inside is a space, closest to any point in that space, so extensible to other dimensions. Test of truth: three points of contact. Each vortex is a meeting point of three edges.

Calling all giraffes: You need twenty or thirty giraffes at different ages to test the growth process. Need photographs of the giraffe, then you could measure age. You could tell the age of your giraffe through one spot. If leapord, the deviation would also be clear, but whole body grows at same rate, less deviation.

Across species animals are adapting to environments?

Do leapords feed on giraffes? Did the prey adapt to the hunter? Or did the hunter develop spots like its prey because it feeds on its babies?

Will we humans even have cow patterns in the future?

What would happen if tested different models on the system (i.e. Voronoi, r and d, cellular autonima) i.e. r and d: wide range of possible patterns, heavily researched area weak control of results, Ocamm´s razor, some patterns hard to produce Subjective: what is a basic pattern (e.g. giraffe where must add cascade). Complexity, think of five chemicals interacting. Control mechanisms for five chemicals would be a mess to explain. At least should test for simplicity.

Ocamm´s Razor counselling service: develop a meeting point to find your animal soul mate at the Ocamm´s Razor Bar. This would be the place of least resistance. Marcelo´s idea. Very nice (the dating service).

What is next research: Marcelo can see the clonal mosaic patterns and dynamic patterns developing in this area. How do you change perameters as system is changing. e.g. giraffe patterns, takes two minutes for giraffe to generate from start to finish (you can create your own giraffe), but since there is a notion of time, relates directly to chat environment.

Okay, so here is an exciting bit: Marcelo and I set up a series of rules, days (over time), plus different centres and sources, we evolved a pattern: the map (jungle) that it generated was gorgeous, with large floral centres (the more active ranters) that connected through a web. A Louise Bourgeous of the internet. Now he is changing the peramters.

How it works: the colour represents the type, if brown need to stay connected to other browns, forced interfacing into the pattern.

Marcelo: splitting rates of cells, (how often you split) and parameters Mutation probablity (when split mutate to third type, daughter cell a different type) With cellular automita, set of rules, figure out types of rules to apply to system? Marcelo could help with rule generation

Xmorphia: system that gets variations from spots to stripes, can use equations as shareware, Marcelo´s code is available to play with. CM patterns would create dynamic patterning, How would we asign qualities? You start without qualities, with nothing, then asign possible qualities, ask system for parameters and use as input into a neural network. Need to train neural network, there are biases written into the network. You assign qualities to perameters. Make a systematic exploration of perameters, tell the network parameters and criteria then extrapolate to get good patterns.

What else might we learn: Morphing of CM patterns‹how from ocelot to tiger pattern. Two starting points from beginning to the end. Faster patterns. (Major but three minutes versus r and d which takes ten minutes) Explore the range of possible patterns (e.g. ocelot by accident), to think about multidimensional space. Reaction diffusion: cascade clonal mosaic pattern. Mutation probability If mapped data relationships from chat to reaction diffusion and then to c.m., would test model. Pattern generation engine, looking at the power of generating, capacity of rd and cm to overlap, each are infinite, answer is not through running forever and seeing which gets more panels. Need analytic. How different biological systems map: cellular automita versus CM Models: methods, e.g. ten perameteres, fix nine, vary and fix, then fix other nine, ten dimensional space, where all the patterns live. Have an idea of what perameters exist for each methaphor (ocelot, giraffe). Computer science: once have the model, figure out what kinds of phenomena can explain through that pattern. Games: name that pattern. What does this remind you of? Methaphor or structure? Pattern generation in nature as general issue.

Marcelo suggests: Look at xmorphia (if plug in numbers know what do)‹see the site. http://www.cartia.com/