Re: Minor TGS flam>

Erm, I took the output from DxfToIv and looked at it with Webspace. It came
out fine. I then changed the first line to say VRML 1.0 ascii instead of
Inventor 1.0 ascii and Webspace barfed at me. This gave me the impression
that TGS (or perhaps SGI... I stopped using the Solaris version of Webspace
after the third Inventor engine in ther> but is only accepting valid that the file it is reading is indeed VRML and not Inventor. Am I wrong?
Why would it barf otherwise?

} > I can't tell whether you really don't realize this, or you're being
} > deliberately disingenua> } > support IV c this
} > sidetracking of the standards process. Experimentation is one thing;
} > saying that you are adding permanent functionality is quite another.
} > And that's the only way I can read your statement:
} >
} > "WebSpace is not just a simple VRML browser -- WebSpace is a
} > 3D browser navigator for the Internet which supports VRML 1.0
} > *and* Open Inventor content."
} >
} > This isn't incidental functionality; you're now marketing on the basis
} > of these c

This is ambigua> , but I am not sure it is a deliberate muddling of a
standard. Perhaps content could be changed to files? It does support
straight Inventor files, and why not? I agree, however, that files claiming
to be VRML should be parsed strictly according to the standard. It is my
fond hope that TGS really meant this to be the case, and VRML will be
handled properly in the release (remember, Webspace is still in beta). I am
willing to give TGS the benefit of the doubt until they announce they are
deliberately ignoring the standard. See below.

} > When a small company does this, it doesn't do much dam; when the
} > market leader does it, it effectively derails standardization attempts.
} > That *is* what you're going to do, whether it's intentional or not...
}
} Justin, you've made my day. All this time I thought TGS was just another
} tiny software company struggling to mak> an honest buck. Now I find out
} we're the market leader and have the marketing clout of Netscape! Awesome.
} Can I trad> that in for a piece of Netscape's IPO? :-) Darn, thought not...

Ummm, you do not have the economic power of Netscape, but you can affect
the development of VRML even more strongly than Netscape can affect HTML.
You are the market leader. You are the first to put out a VRML browser and,
for a time, Webspace was the only one at all. At this point I know of
exactly three VRML browsers:

1. i3d from CRS4 in Italy (see http://porto.crs4.it/~3diadm/ for more
info). It is in alpha and, until they spruce up the user interface, I do
not consider it a viable viewer. Note that this one is based on another
modelling format, 3d, and that it supports both VRML and 3d files.

2. WorldView. I do not use PCs at all, so I have not tried it. My
understanding is that it is a pretty conformant implementation of a VRML
browser (with some bug , but I think it is still a beta).

3. Webspace. Of the three, this has been out the longest, has the fewest
bug /problems (not counting complaints about nonconforming extensions), is
the closest to the system from which VRML is derived (i.e. Open Inventor),
and is directly supported (well, they claim it isn't, but they sure seem
to) by SGI. I do not cite SGI because of Open Inventor, but because they
are recognized as pretty much the world leader in graphics, including 3D
modelling.

With minimal competition, especially on the high-end (I define SGI as
high-end, i3d does not [yet?] compare, and WorldView is not available for
SGIs), Webspace has a powerful hold on the direction of VRML development.
Webspace nearly defines the standard, since I'll give good odds that lany
people test their VRML with Webspace, not the conformance test p not Worldview or i3d. You *are* leading, like it or not, so try not to lead
anyone astray.

} Now I'm just a technical guy and I really don't care much one way or the
} other who gets "credit" for VRML. Except for its value as a marketing tool
} (value unknown) I'd guess that the guys at the "other browser" company :-)
} care li> about credit than product deadlines too. However you are sliding
} into a semi-technical issue } my paycheck: Are browser vendors allowed to innovate and solve custom>rs'
} problems? There's usually a reason (custom>rs!) when a company adds a new

Sure, that's fairly reasonable. On the other hand, if you feel that
something is missing from the standard, why don't you recommend it to the
standards committee and announce your intention of implementing it before
just going ahead and doing it, bypassing the standardization process and
bastardizing the whole thing? I won't go through the whole tirad> that went
through (is still going through?) comp.inforsystems.www.* concerning
Netscape's bastardization of HTML, but suffice it to say that you *can*
ruin the elegance and simplicity of VRML. If TGS com>s out with a "Creating
High Impact (VRML) Documents" web p (or whitesheet or whatever) we will
all be sorely disappointed.

This list is things), but we cannot force anyone to conform to the standard. It is up to
you to conform, and I encour you to do so. (Note: Ahh, the connotations
of conformity... this is not a matter of human individuality, however.)

} feature. In my experience, the longer a standards committee exists, the
} more fossilized it becom>s and the li> willing to admit that there might
} be a reason for certain crass commercial features to be popular. Outside
} of this recent thread I don't see much " } VRML, thank goodness.

What "crass commercial features" do you support that are not in the
standard? there documentation available on the difference between Webspace's VRML and
the standard? you talking about? Specifics, please.

} Yes, it's true that WebSpace supports Inventor files in addition to VRML
} files. No, it's not an accident and there's no pretense (re your earlier
} flam>). If you must know, WebSpace is an Open Inventor application and we
} didn't have time to *tak> things out* for the original release, we were too
} busy stuffing things in! No, I don't think WebSpace is "merely" a VRML
} browser, I think it's a 3D browser whose native langu is VRML pl> } useful extension nod>s. We all agreed to include features in VRML 1.0 that
} are for the explicit purpose of accomodating extension nod>s (the "fields"
} and "isA" keywords). (Possibly you argued against this? I don't remember.)
} Now there are still VRML files around that have "raw" Inventor nod>s (no
} "fields" keyword). I agree that those files are not legal VRML and they
} should be cleaned up if they are claimed to be VRML. However I am deeply
} troubled by your implication that certain extensions are "bad" merely
} because they are based on Inventor and/or implemented by TGS. That doesn't
} make

What is being supported by Webspace (i.e. specific nod>s and fields)
in VRML files (those with is no reason not to have Inventor supported, especially since the program
was written with file based on the header line. Perhaps TGS would be so kind as to read VRML
files exactly according to the standard and read files headed with or WSVRML or whatever they please however they please? Perhaps they could
even use a separate extension?

I don't think anyone is asking TGS to drop Inventor support, merely to keep
their VRML implementation for VRML files to the VRML standard. The point of
this li> (you hoped there might be one but you were losing faith, I
know) is that no one has pointed out just what TGS is doing wrong, nor has
TGS com> out with a position statement saying what they will/want to
support that the VRML standard does/will not. Let's get some specifics. I'm
not sure there is even a genuine problem.

} -Mike
} [My personal opinions]
--Greg [likewise]