RE: [www-vrml] Re: [x3d-public] Migrating ROL to X3D: Nyet.

["Len Bullard" <cbullard@hiwaay.net>]

For sure, but it also means we have to be careful about telling people how
wonderful X3D will be.  It depends.   If starting from jump, it is better to
jump.  If starting from cache, it may be better to stay with the date that
got you to the party.

1) Yes, X3D is more functional.  No, you may not be using those features
yet.  What is most important is to know where the sweet spots are locally
among ease vs scripting vs capability.  Some nodes that are declared in X3D
can be achieved with scripts in VRML97.  Some, say multi-texturing, can't.
Some nodes available in browser extensions (say particles) are only
reasonably achievable using the nodes because the scripting would be quite
difficult. Do you really want that fire or that fountain if it costs you x
percentage of surfers or are you building to the edge, putting in browser
sniffer logic, etc?  HTMLers have lived with these tradeoffs for years and
they show no sign of converging.  If anything, that got worse.

2) Browsers are improved.  They may have improved the most on VRML97.   You
still have to decide which rendering model you are happiest with.  That's a
taste thing.  You likely are going to pick one and stick with it regardless
of the language version you use simply because this is a 'graphics' language
and rendering is the strong pole of the rendering/behavior polarity most of
the time.

3) Some things that are more 'standard' may be more inefficient.  Thyme's
examples for the transform functions seem relevant here.  OTOH, without
H-anim, you don't have flexible fingers and that is worth having most of the
time if you are relying on a lot of human gestures.

4) Editors are more capable.  In some cases, they may not be as easy to use.

Ease of use, extensibility and feature-richness are three poles in a polar
coordinate of forces.  The tool you pick depends on where you want to put
the most energy and how much you have to spend.   For the price and
features, Flux Studio is simply amazing.

An update is always a lot of work.   Choose according to schedules and
energy.   I was giving it a shake test to determine roughly how much work
was ahead of me if I converted now.  It's too much at this point because of
my own learning curve.

But if this seems harsh, consider what happens to gamers who rely on custom
engines.  Consider what happens if anyone tries to take SL models to another
virtual world.  So Irving is right that standards will improve the
metaverse.  He doesn't seem to know how close these are to having and we
aren't always as honest as we need to be about how far they are from working
perfectly the first time every time for any content applied on every system
claiming conformance.   

What is certainly true is that having IBM or other companies putting
resources into the standards where they are the real standards will be a big
help as long as they are there to cooperate and not to co-opt.   We've
survived this long by sticking to the goal and hanging on to our values but
continuing to build content.  If that means we still often build for one
browser, that's just the price for getting kick-ass content and that's fine
as long as we stick to a browser built by a company that is working with the
other standards-based product vendors to narrow the gaps.

We're fine.  Just not nearly done.  It's a long trek to Jerusalem across the
Sinai from Cairo if you have the whole family with you, dogs and sheep to
feed but no horses, a boiling cloud for light, ground gradoo to eat from who
knows where, a leader who wanders off for days and comes back with more laws
than last time, and you aren't even allowed to keep one golden calf for
parties.

len

-----Original Message-----
From: owner-www-vrml@web3d.org [mailto:owner-www-vrml@web3d.org] On Behalf
Of Joe D Williams 
 
> project
> is proven correct.

So ok, you knew that would happen and so did i, if you tried to
get perhaps too much VRML97->X3D automation into the update.
We both know that once there is running code, most of the
time the biggest disaster possible is to put it through some
different authoring tool to try to get to the standards-track.

Like the written spec's update from VRML97 to X3D, we need
to at least start out gently and keep close track of what is
happening at each point in the transformation.

Best Regards,
Joe





-------------------------------------------------------------------------
for list subscription/unsubscription,
go to http://www.web3d.org/cgi-bin/public_list_signup/lwgate/listsavail.html