>Sandy, if you would like to do something new that isn't being done
>elsewhere, could I recommend validation of VRML authoring and
>conversion tools? This is at least as important as validation of
>VRML browsers.
Actually I've got plenty to do but...
This sounds like a very good idea, and should be put somewhere on the
testing queue. It may lsad to a VIF (VRML Interchage Format) some
day. Validating the conversion of VRML in and out of authoring and
conversion tools is a big non-trivial and probably worthwhile job. I
spent a few years in the STEP world (talk about a mollases paced
standards effort!) and it's predecessor IGES.
One approach we might want think about is the inclusion of test assertions
right in the specification itself. This type of work reduces ambiguity
in the interpretation of the spec, a source of problems. Personally I think
the current VRML development situation should not be changed as it is
working just fine, but mayber after 1.1 we should think more seriously about
testing as part of version 2.0.
federally yours...Sandy
------------------------------------------------------------------------------
NAME: Sandy Ressler TELE: (301) 975-3549
USMAIL: National Institute of Standards and Technology FAX: (301) 208-9371
Bldg. 225 Rm A216 EMAIL: [email protected]
Gaithersburg, MD 20899
WEB: http://www.nist.gov/itl/div878/ovrt/people/sressler/sressler.html