|
<![CDATA[
To: Frank Schuurmans <frank@bio.vu.nl>
Hi Frank,
MAJOR CHANGE OF PLAN
--------------------
I did say the original posting was only a rought first shot.
Having thought about things for a while, and also referring to my notes for
the assembly language (hub) implementation, I now think that each INPUT
to a webneuron should have its own separate HTML page with links to
caller and target, and no code. Call this a WEBAXON.
The name of the data can be the webaxon page title. This can be used to
reference the data when writing the webneuron script.
Let's assume a webneuron contains a table with headers, 3 columns, and
data area below - as for our website. Then I would expect a list of inputs
as follows:- (3 input example)
<a href="132.htm">animal<|a>
<a href="452.htm">vegetable<|a>
<a href="771.htm">mineral<|a>
Let's put them at the bottom of column 1, perhaps with a suitable header or
icon to the left. The names "animal" etc must be unique otherwise the
interpreter will get confused.
To view the webaxon inputs click on the respective link.
The code or <PROG> script could be in column 2, and could contain lines
like:-
IF Animal = mydog.gif
<FIRE href="http://myrobot.eu.org/go_walkies.htm">
ENDIF
The current data in the animal webaxon will almost certainly not be called
mydog.gif. The interpreter will directly compare the two images pixel for
pixel and decide if there is a close enough match. There is no reason I can
think of, that the interpreter couldn't use weblets for this analysis.
It looks like the beginnings of some very flexible image/pattern recognition
possibilities here, especially if the gifs are small. Larger images could be
broken down (or built from) many small gifs and operated on, similar to or
integrated with, operations on strings and numbers.
Communication with the robot could be by a PPP radio link, and it could
have its own http server on board. This is not really a joke! You could also
make real progress in robotics using webneurons. I worked in robotics for
two years.
For a library webneuron we might have:-
<a href="132.htm">creature<|a> and,
IF creature = mydog.gif etc
When the library webneuron is plugged into an application then "creature"
can be converted automatically (throughout the script) to the actual input
name, e.g. "animal" in the above case.
If you wanted the inputs on another machine so you would need:-
<a href="http://..../132.htm>animal<|a> instead
One reason for webaxons is because an INPUT could be anything,
graphics, Email, string, numeric etc, and if there are lots of INPUTS then
the poor webneuron is going to get very large and confused. We may as
well simplify the programming task by making more links and pages,
which is one of the principles of webneurons anyway. So the webaxon is a
data storage location and a link page between the caller and target.
This fits in with the brain because it is the axon that holds the signal data
values transmitted between brain neurons.
The webneuron now has less to worry about, plus we can see instantly
which caller is associated with each INPUT, whereas before there was no
connection. I believe this is information that could prove useful.
We have now neatly sectioned input data and code, into separate pages,
but kept them close together.
We will also need LOCAL data storage, or workspace, for the webneuron
and expanding on the above principle we should again use a separate page.
This may replace the <MEM..> extension mentioned above. Let's call it a
WEBSTORE.
There will be a list of webstores just in a webneuron just like the list of
webaxons, except let's put them at the bottom of column two
The FIRE command must now change too. It simplifies to just:-
(example)
<FIRE href="target.htm"> or <FIRE href="http://...../target.htm">
There should be a link back from the target to all firing callers.
There will be a list of outputs (previously specified by PARAM). Let's put
them at the bottom of column 3.
Four outputs example:-
<a href="132.htm">animal<|a>
<a href="632.htm">name<|a>
<a href="110.htm">age<|a>
<a href="188.htm">weight<|a>
Note that input1 is going to the ("walkies") webaxon mentioned above but
the other three inputs go elsewhere, perhaps as part of a database. This
also shows how all aspects of computing can now be merged seamlessley.
and INPUTS are now set separately e.g.
animal = catpic.gif
or in the library
<a href="132.htm">Species<|a>
<a href="632.htm">title<|a>
<a href="110.htm">How old<|a>
<a href="188.htm">weight<|a>
and
Species = catpic.gif
(when plugged in the names update as for the input example above)
So now we can just set inputs without actually firing the target. I feel this
will be an extremely useful freedom.
I hope all this is not too confusing,
]]>
|
John's Emails
webneurons