A text VRML renderer might actually be quite useful for sight-impaired users.
I had a very nice email exchange with Christopher Serflek about making VRML
more accessible, and think it is do-able.
Ideally, interacting with a text VRML browser would be like playing an
adventure game:
"You are looking at a small, messy cubicle, crowded with bookshelves and
whiteboards full of scribblings.
The room has the following links:
1- Gavin's Home Page
2- The Open Inventor Home Page
3- Gavin's playroom
The room has the following viewpoints:
1- Top view"
We should add support for this kind of thing into VRML 1.1; I think all it
takes is a Description associated with objects, viewpoints (which should be
formalized more), and WWWAnchors (which we already have). Convincing VMRL
authors to fill in the description fields will be the hard part...