Re: [ba-ohs-talk] Node Sequencing [Was: **** Instant Outlining!!!***]
At 05:54 PM 4/23/2002 -0700, you wrote:
>Here, it seems to me, you are asking the question: should a cell be aware
>of which dimensions contain it?
>At the moment, I'm agnostic on that. When I started playing with this
>idea, I started with a class Cell, and another class Network that contains
>a hashtable the keys for which are dimension names and the objects
>associated with dimensions are vectors of cells. In that case, a Cell is
>just a Cell. It knows from nothing who its neighbors are.
>OTOH, there's a biological metaphor at work being ignored in that design,
>one in which cell-cell interactions come into play. In such a case, you
>might want to think about cell behaviors when contained within some
>particular context. Maybe those behaviors are learned. Maybe they are
>defined "genetically" and can be mutated over time.
>It strikes me that such a project can start with a trivial implementation
>-- the hashtable/Cell model, and grow with experimentation into some
>complex organismic entity. (01)
I went back and looked at my code.
The engine consists of two hashtables, one for named nodes and one for
named dimensions which are vectors of nodeNames. Important point, based on
Ted Nelsons comments that the actual data resides one and only one place
such that when that data is updated, all dimensions are instantly "aware"
of the change. (02)
All of which seems to be consistent with Nelson's Xanalogical Structure, in
which all data is just one long (really long) character string and the
actual action occurs in a "virtual document" that resides above that
string, with pointers into the string. Versioning simply means changing
the pointers by pushing them down on a stack with new pointers to the fresh
information. Pretty much the same thing can go on in ZigZag, where all
the data itself is just that long character string, and the values in the
cells are actually pointers that are contained in stacks (version
control!). Stack depth tells you what version you are at with that pointer
(here, I am thinking it really is a pointer (maybe an offset) into an
information resource (e.g. string). (03)