blaine wrote: ↑Sun May 20, 2018 3:28 pm

I like the idea of using a tree for the computer model because the issue with existing approaches in conventional physics is that they use a euclidean grid which is the "stage" upon which events are played out, whereas the RS deals with motions that are the stage. With a tree you are only simulating the existing motions and their possible transformations.

I was working with it last night... some interesting questions have come up:

**Different dimensions**: the papers indicate that all objects modeled in a tree have to have the same number of dimensions. This is NOT the case in the RS, however, since 1D "electric" and 2D "magnetic" exist in the same, 3D universe, at the same time. Larson addresses the problem by resorting to "random distribution" (such as the inter-regional ratio).

It is interesting because in order to "make it work," you basically have to "collapse a wave function." For example, take a 1D line on a 2D sheet of paper. How do you draw it? Using Larson's random distribution, you would have to draw ALL possible configurations, basically spinning the line into a circle. But in order to get a single line out of it, you must collapse that probability into a single solution by the observer adding a concept of "angle."

If I tell you to draw a line on a sheet of paper, it is "pot luck" if you draw it the way I intended you to draw it. BUT, if I tell you to "draw a line 30° to the horizontal axis, I will always get the line I was looking for. This takes the randomness out of it.

I think the same approach can be used for 1D and 2D objects in a 3D universe--the tree on lower-dimensional objects is just missing a dimension, so ALL possibilities exist--until something interacts with it, adding a value to that missing dimension, fixing the motion in place.

My conclusion from this is that "collapsing a wave function" is nothing more than "forcing an interaction" to supply the missing, dimensional component that converts a 1D/2D function into a static, 3D structure. It is not really "observer" based, because in order to observe it, we have to create some kind of physical interaction--and it is that INTERACTION that is collapsing the function.

**Yin-Yang Aspects**: these tree structures are based on linear, yang relationships (line, area, volume). In RS2, we also have angular, yin relationships (angle, solid angle, hyper angle). The yin relations are needed for "equivalent" space/time projections, but I have no idea on how to decompose the concept of "angles" into a tree structure. The key is probably in the use of counterspace, but as of yet, still alludes me.

**Space-Time Aspects**: conventional trees deal with areas (pixels) or volumes (voxels), so only the "space" aspect. To convert to speeds, time has to be included. It is more than just using s and t variables for each voxel to get a speed, because 3D time operates independently from 3D space--it is not a 1:1 correlation. Only the 1D projections of one sector influence the other sector, bring us back to the "different dimensions" problem--3D space is influenced by 1D temporal motion, and 3D time by 1D spatial motion, yet there is no way to fill in the missing data.

**Boundaries**: a concept introduced in the quadtree paper was that there are three pixel conditions for an object:

*external*,

*internal* and

*boundary*. The RS has boundaries, as well, such as the "unit space" and "unit speed" boundary, but they are not

*tangible* boundaries... there is no structure present. This is examined in detail in Larson's

*Liquid State* papers where he discusses the concept of "surface tension," concluding that such a concept does not exist. Either you are "inside" or "outside" of an object--there is nothing special about where the two meet, other than a change of state or material.

Now when addressing a concept like the

*gravitational limit*, a boundary, you now have to have some kind of "internal" structure to represent the region where gravity has effect, not just its limit. The same situation exists for electric and magnetic fields. Since you cannot include multiple "materials" in a single node of a tree, because each material has a different speed for a particular voxel, it seems to be necessary to use multiple tree structures to represent these features, like "layers." But this does not seem to be efficient, because you've moved off of the single tree structure representing all motion, to a number of trees representing layers of motion. This might be the case, but not certain.

Comments welcome.

Every dogma has its day...