To: rap@maths.soton.ac.uk
Hi Rubber Bucanner,
>I like the idea of webneurons. It does seem eminently sensible to use
>something with the behaviour which you describe in web pages etc. However,
>I do wonder whether it will be as versatile as, say, Java. To be so would
>require a significant increase in the number of commands in HTML.
Maybe, but we are looking to cover things like looping, by linking webneurons
rather than in a script, so this simplifies things immediately. Please see:-
http://brain.eu.org/weblet.htm
In fact we will try to keep the script inside a webneuron very small and use the way
webneurons are linked to control logic flow. The system is a complete programming
language so should be able to do anything Java can (I hope).
As you say though we will still have to add a lot more commands.
>My final thoughts, however, are positive. I like variety and choice, and I
>can see no harm in having two (or more) media for web pages. All that is
>required is a browser that can understand all the standards, no great
>problem. I think that neural networking is an important idea and so should be
>promoted, but object-oriented design and programming is important too. There
>is room for both.
The problem we have is do we write a new browser, try and get Netscape to change
theirs, or even do we write a new OS.
Thanks for your encouragement.
|