Re: LANG: Binary formats.

Peter Kennard ([email protected])
Tue, 28 Feb 1995 07:05:33 -0800


Well, I know you know but Venue(tm) is an appropriate foundation (when it
gets a final linker) for a binary format. Basicly it offers:

- A platform independent integer format in one byte size increments
depending on number of bits required. (intrinsic compression)

- A platform independent floating point format in one byte increments
depending on the exponent bits and precision required. (intrinsic compression)

- A platform independent code and behavior description format deigned
for effective run time efficiency 0 to 20% loss over native code.

- A powerful binary interpreter for the algorythmic description of
"generated" geometries and rule based objects, as well as interactive
conditional processing with the server. An ideal basis for a 3-D "PostSculpt".

- A deliberately open ended design intended to be a seed for further
evolution.

It of course does not provide an architecture for a 3D database or other
"API" features other than those required to install another API, but,
support of such an architecture is it's main design purpose.

At 01:13 AM 2/28/95 -0800, Mark Pesce wrote:
>Folks -
>
>I know this subject has been brought up before, but I think this is a good
>time to begin discussion of a binary format for VRML. Whatever we decide
>upon will be several times more efficient than ASCII, and that will quickly
>become a very important issue.
>
>As there is no published standard for the binary format of OI, we have no
>easy solution; I would suggest that we solicit proposals from the
>membership, which can be evaluated at leisure at the VRML*Tech web site.
>I'll set up a page for this at http://www.eit.com/vmrl/binprop.html on
>Wednesday, and this can serve as a focal point for the discussion.
>
>I bring this issue to your attention because, as I'm beta-testing one of the
>soon-to-be-announced VRML sites, I've noticed that worlds are averaging 300
>- 400 Kbytes. Working in the XOF format (which is a possible candidate for
>a binary shape format), I've been able to get worlds of similar complexity
>in 80 - 150 Kbytes. That's the difference between one minute and five at
>14.4 Kbps, and the difference between success and failure to us.
>
>I look forward to your comments on this question.
>
>Mark Pesce
>---------------------------------------------------------------
>Mark Pesce
>General Partner,
>The Community Company * [email protected] * 415.621.6390
>[email protected] * http://www.hyperreal.com/~mpesce
>I AM * 31 ~ 418 ~ 2012 * That's AL
>
>
>