Re: Do we need a common visual metaphor?

Rick Duffy (duffy@everest.den.mmc.com)
Sat, 11 Jun 1994 08:27:58 -0700


>> ... (common visual metophors etc)
>> I only question whether my conception of navigation spaces is completely
>> off-base or absolutely necessary....
>
>
> No doubt commonly understood elements in *any* medium are the
>esscence of it's ability to be part of a human language. Whether we *need*
>to *impose* one on anybody is really to get a little out of our domain.
>Suggest might be the more appropriate ideal. Trash cans in *reality* have
>come to look a certain way. An umbrella stand that looks like a trash can
>is likely to be confusing and get filled with trash proving somewhat
>unuseful or inconvenient. Most people will then likely get something that
>suggests it's purpose. Even so we are dealing with a new set of industrial
>design problems and there is likely to be a multiplicity of forms and
>possible design solutions for just about every purpose. What is needed is a
>stable lower level set of abilities that many many competing minds
>can make use of. ...

I agree. I think perhaps what we need is a standard taxonomy for things
in cyberspace (probably an object-oriented type of hierarchy). The
rendered representation for each object (or subtree) can then be defined
by the user as suits their needs (or primed by their navigator application),
but underlying it all everyone will be navigating the same basic conceptual
space.

If this taxonomy were defined as part of the VRML protocol (and it would
of course be extensible by the user), all cyberstuff would be classified
according to its ontology, allowing more powerful and useful tools for
navigation.

>They can all design their own way to do things and trade
>them with others because of the standard. The inclination of people is to
>use what they understand and can discuss with others -- naturally leading to
>a benevolent evolution of design forms and metaphor just as it has in
>the real world.
>
> It is up to us as early on designers to suggest some potential
>starting places, and avoid imposing restrictions on what is possible.
>
> PeterK

I think the breakout of services as defined by the Object Management Group
for the CORBA spec might be something useful to look at too. I think this
structure might be useful for defining the modes of interaction within a
VRML document.

Rick Duffy, Martin Marietta Astronatics
duffy@den.mmc.com or rduffy@nyx.cs.du.edu