[X3D-Public] my musings on O3D and JavaScript on the web

[GLG]

Hello John, 

Thanks for your post. I'd like to continue on some of that
and see where that takes us. First you said...


>... O3D takes COLLADA and generates JavaScript.  


Sort of... using the sample COLLADA Converter provided.
Google's site also indicates that "The O3D's Developer's
Guide provides information on how to create a file converter
and loader for any 3D content."

http://code.google.com/apis/o3d/docs/techoverview.html


There are nice figures on that page showing where everything
fits together with O3D.

The Developer's Guide also shows how to import scene files
from various other sources such as 3ds Max, Maya, and
SketchUp. 

http://code.google.com/apis/o3d/docs/devguideintro.html

They add... "To be clear, you don't need to use this format
(COLLADA), or even use the sample reader code to read your
assets. In the end, this format is simply converted to calls
to the O3D API, which you could generate on your own from
any input you choose. If you want to read your own format,
you will need to write your own import code in JavaScript
that calls the O3D API calls to construct a scene graph with
your data."

Conceivably, if we wanted an X3D importer then we could
write one. It seems that, Google is positioning O3D to
possibly become a 3D plugin that can run everywhere (so long
as it's in a Web browser) and support any 3D content. They
also want the community to write the converters/importers. 

That strategy would effectively pull the carpet from under
the feet of browser/plugin vendors and potentially content
development application vendors, while insuring O3D's place
on most if not all platforms.  GRANDIOSE!!! A real spark of
marketing genius. IF IT WORKED!!!!! 

For the sake of brevity, even assuming that it would work
satisfactorily (to the level of current plugins) in say 5,
10, maybe 15 years, who would then really gain anything but
Google? Is that their idea of a standard? The O3D standard
which they own (or at the very least own the one and only
O3D browser plugin)? Plus, does that mean we can now only
view 3D in Web browsers?

This is all marketing. 'Let's sell it to the public. If we
get traction and they buy it, then we'll make it, while
killing (or hurting) nearly everyone else in the business.'
Like I've said in one of my original posts. 'We are the
almighty Google. We are new at doing 3D but we can do it
better, the marketing part of it that is, because we have
the dollars. Programming it, actually building it, will come
later. And the best part is, you will do most of the work.' 

Fantastic! I love O3D! It is probably no coincidence they
chose to do a COLLADA converter first. Go for the easy
target. Web3D might not be so gullible.




JavaScript (remember Ajax folks)
>appears to
>be the low level language of the future, not XML or HTML.
>JavaScript
>is rapidly replacing C, C++, and Java as the most portable
>procedural
>language, and it's getting more secure and fast to boot.



But Javascript only works in container apps anyway such as
Web browsers which are written in C++/Java.



>(see Caja,
>V8)  Now that JavaScript has 3D in all major HTML browsers,


Not yet. And I wouldn't hold my breath waiting for the day
when this performs the way current X3D plugins do. This is a
very long shot. I don't believe the hype so readily. Just
look at how long it took to get the current plugins to where
there are now. Easier said than done.



>this sets
>the stage for declarative tools to be built on top of it.
>



Yes let's reinvent the wheel. :|




>Be smart, generate O3D from X3D/VRML.



Ok, let's do that and leave behind the plugin vendors that
have devoted a significant portion of their careers to X3D.
Sounds like a good way to keep the community going. :| 




>
>I said this a long time ago, perhaps some will recall.  You
>could use
>aspect oriented programming in Xj3D to generate O3D..  Just
>log all
>the JOGL calls as O3D and you are done!  Your application
>will be
>converted to O3D!  And you can share your aspects with
>others in the
>Web3D consortium so that everyone can jump on the O3D
>bandwagon.



Dito




>Sigh, I should probably sit down with an animator,
>mechanical engineer
>or electrical engineer and learn their tools.  When will
>their tools
>be available on a web browser that runs on a Macintosh?



Hum! I doubt they'd really want their tool to run in a Web
browsers. But if they did, they'd better want it bad because
that's the only place they'd run.






>Will these
>tools be written in X3D/VRML or O3D?  Does anyone want to
>write 3DS
>Max, Maya, AutoCAD, Pro/E or Flash's editor and make it
>available on
>the web?  



Any app can be made available on the Web given time if the
vendors are so inclined. There is no need for a Web browser
or O3D here, even if we assumed vendors would be willing to
go open source to begin with.





>Where is the X3D/VRML editor written in X3D/VRML?


I'm sorry what? Where is the text editor written in text, or
the HTML editor written in HTML? :| 

X3D/VRML editors can be written in many languages, but being
able to write one in a low level O3D Javascript API is
likely a long way off. 





>Where is
>the behavioral 3D editing program designed to be run in a
>web
>browser?  Is there a performance issue that V8 can solve?
>I am a java
>programmer.  When will NetBeans, Eclipse, IntelliJ,
>JBuilder, and
>JDeveloper run in a web browser?


Why does everything need to run in a Web browser?




>
>Open source and fast JavaScript is here.  Will your company
>be ready
>for it?
>


No it isn't, at least not in any significant manner. What is
here is a proof of concept, the sales pitch.  




>
>If you still think XML is the answer, read this:
>http://www.charlespetzold.com/etc/CSAML.html
>


I see the sarcasm in that.  XML is a lot harder to
read/write than the simple code it represents which could be
written much faster manually. But when you are ready for
machines or software that can program themselves on the fly
you'll need it. When you are ready for massively dynamic 3D
stored in databases, you'll need X3D. The shared apps you'd
like to see run in Web browsers, are much more likely to use
XML data than the Javascript mess that is being proposed.


End of rant. I wish I could agree with you more but these
are my opinions. Nothing personal.

Lauren