Re: Scripts vs APIs

Kevin Goldsmith (
Fri, 9 Sep 1994 22:21:16 -0700 (PDT)

> Re: expression for authors:
> The point is that I'd like the prefered tool be some drag-n-drop,
> menu-driven interface. Something my wife could use. Open an
> existing file that describes a room with a table, pick "sphere"
> from menu, drag it over to the table, and drop it. Viola.
I agree that our goal with a VRML creation package should be ease
of use. Although I don't think that a creation package should
necessarily be necessary, and I have yet to see a 3D design or animation
package that I would describe as "easy to use"

> Gavin Bell reminds me that an editor for postscript that can open
> up and succesfully edit a pre-existing postscript file is essentially
> non-existant. Its just too complicated. Ditto for C. COBOL does
> have some primitive tools to do this, but they cost $100K and up.
> The COBOL market is bigger and has more money.
well, then. Lets not make something as complicated as Postscript.

> May I also remind that virtually nobody writes postsript by hand.
> Ditto for AutoCAD DXF files, which are actually a descriptive language.
> In the word-processing world, HTML and TeX are about the only two
> things people write by hand, rather than with some word-processor.
> And HTML is succumbing fast ... while TeX is proceedural.
However, I have built quite a ton of objects in Inventor by
hand. This is a credit to its simple file format.

> So do you want to use VRML for complicated VR simulations? Or do you
> want it for simple 3D on the net, viewable on lowly 386/486 with
> interactive speeds?
VRML should be scalable. I think it should work on a 386/486 if
we ever expect it to be widely accepted. Sure, someone might want to put
a really compliated scene on their home page, but then very few people
will be able to enjoy it.