[X3D-Public] O3D - Google's 3D JavaScript API

Len Bullard cbullard at hiwaay.net
Thu Apr 23 06:00:46 PDT 2009


That’s a crock front to back, Ben. 

1. Google is well-aware of X3D and VRML.   They declined to participate.
2. No one has made any case for the shininess of the teapot and you don’t
seem to be answering the technical question, what is better about the O3D
scene graph.
3. If you can’t answer two, there is no discussion.

You are confused.   A call for clarity is no a call for war.   You were
recruited into a secret project to launch a Google product before it is
ready for launch.   It is understandable that this is exciting for you, but
understand that this is the X3D Public list for an international standard.
If you don't wish to call out the technical merits of the work you've
participated in, so be it, but X3D is an ISO standard developed under open
policies with regards to participation and intellectual property.  Google
and IBM are not members of the consortium and do not participate in open
standards efforts in general under such agreements.

You are beginning from the positions that a) something is wrong with X3D b)
something is right about the Google API and you offer no technical arguments
to support said claims.

The business of creating standards is serious and requires both scrutiny and
open policies with participation agreed in advance.   Otherwise, it devolves
into the kinds of shenanigans the 3D web community has been watching from
IBM, Second Life, Forterra, Google and others in the last few years of
failed initiatives.   

X3D is working now.   So I ask again, what about the O3D scene graph is
superior or preferable.  Vague statements about its size or perceived market
value are not answers to these questions.  So, what you are saying so far is
a crock.  That isn't hostile.  It's the case as you presented it.

len

-----Original Message-----
From: x3d-public-bounces at web3d.org [mailto:x3d-public-bounces at web3d.org] On
Behalf Of Ben Francis
Sent: Thursday, April 23, 2009 4:25 AM
To: x3d-public at web3d.org
Subject: Re: [X3D-Public] O3D - Google's 3D JavaScript API

Hi all,
Thanks for the responses.

On Wed, Apr 22, 2009 at 9:18 AM, GLG <info at 3dnetproductions.com> wrote:
I am wondering if the fact that such companies as
Google and IBM (last year) are not getting it is symptomatic
of something WE are not doing?

A lack of communication? Let's ask them.
 
If they simply choose to ignore Web3D then what
are the reasons? Technology, Control, or Money?

The two companies you mention, Google and IBM, are both in my view big
supporters of open standards - so I wouldn't think it's about control.
Perhaps either they think that for some reason Web3D's technology isn't
appropriate for their target market, or the standards are too costly for
them to implement? I wouldn't like to say which.

On Wed, Apr 22, 2009 at 11:08 AM, Joe D
Williams <joedwil at earthlink.net> wrote:
Well, what do you think, Ben?

I think that they have a very shiny teapot. I think that Google started work
on this long before Khronos and Mozilla announced their intention to create
a new standard. I think Google has got around to creating a release of their
project as a conversation starter, to say "Hi, look at we've done, what do
you think? How can we improve it?". I think Google is serious about
contributing to standards to make the web better, but I think that their
secretive nature as a company can sometimes result in a hostile response
from established, open communities.

I think the fact that there are so many different imperative approaches to
3D graphics on the web popping up is symptomatic of something, but I'm not
sure what. My gut feeling is that eventually a single approach (imperative
or declarative) will eventually become dominant, but the fact that SVG and
Canvas both exist in web browsers today suggests that maybe that isn't the
case. Maybe there really is a need for two different levels of abstraction
for two different types of content creator.

How would you compare this with X3D in terms of authoring and delivery? 

O3D and X3D are two different approaches to the same technological problem,
for two different markets. O3D is suited to games developers who want a low
level API they can use to implement their own gaming engine in a web
browser, X3D is suited to web authors who want to be able to rapidly create
3D web applications or 3D visualisations of information. Beyond that, I
don't think that it would be fair to draw many other comparisons, as O3D is
in such an early stage and perhaps a precursor to something bigger from
Khronos/Mozilla/Google et al.

Anything in particular besides the fact that it was already done at an
abstraction level miles above what they are offering? 

It's going on for a year since I sat in a videoconference room in London and
chatted to two members of the O3D team in Mountain View, so I can't remember
their exact responses. I wouldn't want to mis-represent them so in time I
hope they can respond to this themselves. My general impression was that
they perceived X3D to be a rather bloated standard, mixing 3D assets,
animation and behaviour into one big XML soup. They seemed to have some
sketchy ideas about the requirement for a persistence format for 3D assets
in future, though COLLADA seemed to fit the bill there. Mainly, it seemed
that they had a small team who thought they'd spotted a niche and had a
limited amount of developer time to come up with something nippy and useful.

Me, after seeing the 'spinning' cube, I don't think they will make it even
to the spinning tennis shoe. Not because I don't want them to be able to do
it, but because it will be too much to handle. 

Having implemented said spinning cube using OpenGL in C, and some other
slightly more complex scenes, I know where you're coming from. With a
background in web development X3D seemed to me to be much more what I
thought it should be like to create 3D content for the web. But that's kind
of the point - I am not a games developer, in fact I have very little
interest in games. I have an interest in 3D visualisations and applications
- I don't need to implement a gaming engine because X3D can do most of what
I need already.

On Thu, Apr 23, 2009 at 12:50 AM, Len Bullard <cbullard at hiwaay.net> wrote:
Or maybe the web is bigger than they are.  What say you folks?  We've been
on this list a long time together.  Do you still believe communities are
stronger than companies?  Or more to the point, are communities of companies
stronger than any company?

This sounds like a call to arms! Be careful not to create an "us" and "them"
before finding out whether a division really exists. On the web communities
need companies and companies need communities. Why not talk to the companies
and ask them how the communities can better collaborate with them?

On Thu, Apr 23, 2009 at 12:59 AM, Len Bullard <cbullard at hiwaay.net> wrote:
The ball is in Google’s court.  
 
Is it? Haven't Google just taken their shot? Is the ball not in everybody
else's court?

On Thu, Apr 23, 2009 at 1:15 AM, Doug <Doug.duBoulay at gmail.com> wrote:
I just can't understand why Google/Mozilla/Opera/whatever don't just
build on the OpenVRML or FreeWRL libraries.
... 
Then the browsers would have instant X3D ability with basically
no effort on their part.

This is a question I would very much like to know the answer to. Again,
maybe we should ask them. I also think that the paper  that Johannes Behr et
al. are presenting at the Web3D Symposium will go some way towards solving
this.


I just don't understand their desire to aim at such a low level
OpenGL interface.

As I've already mentioned, perhaps it's that for some 3D content authors,
X3D just isn't sufficient, and they have a need to develop their own engine
on top of a lower level API. There's a tradeoff to be met between providing
functionality and introducing restrictions. Is X3D too high level for many
content authors? Is O3D too much of a thin layer on top of OpenGL? Perhaps
the market will decide. 

What's important to me is that Google have expressed an interest in 3D on
the web, backed it up with some code, and called for feedback from the
community. It's up to everyone to figure out how standards and technology
should ultimately fit together in this complex landscape.

Let's continue talking. To IBM, To Mozilla, Opera, Microsoft, Khronos and
Google if we can. I will invite Google to respond.

Best Wishes

Ben

-- 
Ben Francis
http://tola.me.uk




More information about the X3D-Public mailing list