SEE also paper.tex!

I need to have a simple enumeration in Rsf. That will simplify any quick 
elementwise implementation and it will allow single default for higher 
dimensions. On the other hand, maybe to wasteful. 


THIS NEEDS TO BE CLARIFIED CONSISTENTLY!!! 
Java arrays have the fastest changing iterator in the last position 
(unlike FORTRAN). I consider changing the entire SEP convention to list 
the fastest axis first (e.g. in RsfSpace's getNSamples()). 
%
Simple general rule: Consistent in code, documentation, toString() etc.
fastest is last index. since axis are array with last axis having the 
highest index, axis[n], and the first axis the index 0, axis[0], we 
naturally should associate the slowest axis with 0 and the fastest 
with whatever is the highest index: 
call fastest axis with highest number a[i0][i1]
call slowest axis with lowest  number
In constructors etc always order from slowest first to fastest last! 
%
IMPORTANT: 
There is a change in the allocatedata method in all rsf/vector methods!
to make this consistent!
%
IMPORTANT: 
My current looping habit is bad!! Outer loop should be i0 which is the 
new name for the slowest loop. Here is what it should look like:
for (int i0 = 0; i0<n0; i0++) 
for (int i1 = 0; i1<n1; i1++) 
	array[i0][i1] = 1;
%
The Rsf toString() function is misleading: I suggest to print fastest
axis (highest index number) horizontally. 

Maybe the Rsf toString() method should write the dataset out. That would 
make it very standard on how to write a dataset. I am looking forward to 
serialization. 