3D Formats from Commercial Software vs. VRML & Java3D?

RickMuller asks: "I'm interested in molecular graphics, a field entirely dominated by OpenGL. I would love to see something more web-friendly come along. I've worked with VRML and Java3D, and have been very puzzled why these technologies didn't become ubiquitous. Now there are new 3D efforts by Adobe and Macromedia (the Adobe Atmosphere download is available as a beta-download and is way cool!). The press is heralding this as the 3D web revolution. Why should these technologies succeed when VRML or Java3D failed?"

demand

[agentk]

It's a matter of demand. People don't want to make 3D web spaces yet. HTML is still being developed with intensity, web-based business is growing (with pains, natch).

VRML and other web-based 3D technologies are newer than HTML, which is still pretty darned new. VRML is harder to code than HTML, you have to think in 3D and numbers, so the number of web 3d amateurs is much smaller than web 2d amateurs. Shoot-em-up 3D games make money, that's whats big now, but soon enough, the major enabling technologies for really exciting internet 3D communication will be pretty widespread: cheap powerhouse video cards (driven by the game industry) and broadband.

To become widespread, a internet 3D technology will need:

1. to be general, and useful AND exciting to many
2. fairly easy to build, either through simple languages (like HTML) or free, intuitive dev tools.
3. powerful enough to approach levels of (visual) quality now achieved regularly by desktop gaming industry.

While we're at it, I say we go for open standards, extensibility, and pioneer some interesting aspects of multimedia and HCI.

agent k

i love vrml

[drfrog]

anyone who has actually done some
research into what mpeg4 will do when it reaches FULL maturity
realizes that vrml is inside it at the core system level

why?
because it will be there for streaming 3d
and avatars etc

anyone who says vrml is not robust enough has
not used it to its full potential

that being said im the first to admit that the main problem with vrml is conformance and
interoperability {check the archives at web3d.org} . too many plug ins and viewers
but that just pulls it up to par with html xml and
most other web languages
:O

the others are fairly cool
{adobe and macromedia}

macromedia has expressed interest in
having some sort of vrml to shockwave3d i/o and interoperability

java 3d does have the ability to
produce worlds based off of vrml
but its so slow and i dont want to compile a
new applet for every different browser and version!

the other more corporate offerings are just that

btw vrml is an iso standard and the next level:
X3D will be a great step in its continued growth

also check out blendo a vrml derivative to be induced onto the playstation2, its done by sony
so... http://www.blendomedia.com
it has the abiity to use live video as well

and as far as being too lazy to produce quality
vrml well ....
this took me eight hours:
http://www.hyperbomb.com/ayb/

use cortona plugin {by far the best for windows}
for proper enjoyment {course macromedia and adobes are only avail on windows too}
http://www.parallelgraphics.com

i have exp helping vrml on linux too via the openvrml {rpms at www.hyperbomb.com/openvrml/} and freewrl projects both located at sourceforge.net

but these projects go slower than the commercially developed ones... i wonder why?

oh and here is another short example of a sphere in vrml
-------
#VRML V2.0 utf8
Sphere {}
-----------

phwew that was tough

be advised ive been using vrml since 1995
its easy out put from databases too
why?
cuz its plain old text not some bloated binary format!

toodles

They're all dif't...

[ameoba]

VRML is a markup language, like HTML for 3-D spaces. OpenGL is a library that programs use to do graphics (primarily 3D) with. I could see a VRML program written in GL, but other than that, I don't see how they relate much.

As for Java3D, unfortunately, the Java2 spec has yet to see much market penetration, since IE & Netscape still ship w/ braindead JVMs.

Easier

[gtada]

Two things come to mind:

1. Most of the artists I know are too lazy to learn how to write VRML code and even less motivated to learn Java.

2. The Macromedia and Adobe efforts have support from companies like Discreet and Alias|Wavefront.

It's about ease of use. To be fair, most 3d packages can export to VRML, but usually they're very basic (no animation, no materials, etc.).

Also of note, OpenGL implementations are available for Java. Two come to mind: GL4Java and Magician. I'm not too sure how popular they are.

Why VRML failed?

[artymiak]

... have been very puzzled why these technologies didn't become ubiquitous ...

anybody with a couple of brain cells can cobble together a "Web page" using Notepad, whereas it takes a lot of effort and money to create anything interesting in VRML. Also, the bandwidth wasn't there when VRML was born.

LinuxApprentice.com [linuxapprentice.com]

Adobe Atmosphere

[artymiak]

... (the Adobe Atmosphere download is available as a beta-download and is way cool!) ...

So is ViewPoint's VMP [viewpoint.com]

AnyOldComp.Com [artymiak.com]

My take on the situation...

[cmowire]

The situation for 3D is messy. Just leave it at that.

VRML tried to be the end-all, be-all 3D file format. Which it isn't, BTW, as advances in computer graphics have made VRML not very useful for the kinds of things I want. It is just too general purpose.

The new Web 3D formats are trying to capture a discrete market. People want to see a product in 3D before they purchase it. Of all of them, I rate Macromedia as the most likely to grab the market, given that the Flash plugin is one of the few plugins you can count on having. But they have a little better likelyhood of some success, but only if they latch onto an existing technology, IMHO.

The reason why mollecular modeling is all OpenGL based is pretty simple. OpenGL is a pretty nice library for getting 3D geometry to the graphics card for rendering. It works under any decent OS (Solaris, HP/UX, AIX, WinNT, Win2k, IRIX, etc) that you would want to do that sort of thing on. Programmers have mollecular modeling code that is many years old, probably first designed to work on some sort of early SGI machine, that they have just been linearly porting over to newer platforms without rewriting it.

To make a VRML-based platform, they would have to rewrite things. Sure they might provide an output format for VRML so that you could put it up in a web-based query format. But anything more would require you to rewrite things massively, which isn't a good idea, especially when you can still get p1mp OpenGL cards from Sun, HP, and a few others.

Old news...

[jinushaun]

Wild Tangent [wildtangent.com] is already doing what Adobe and Macromedia are just now planning. The only difference is authoring it.

Whereas Wild Tangent must be programmed, Adobe and Macromedia's will probably be a cross between Lightwave and Flash.

People should checkout Wild Tangent [wildtangent.com]. It let's you create retail quality games using javascript! How crazy is that?!

Jinushaun

3D must be easier and more flexible

[flatland]

why didn't (doesn't) VRML take off - because its not easy enough to work with (how many people do you know who can model in 3D?) - the content that is created is not exciting enough (people are bored with the results) same with Java3D and the hundreds of other formats - Wildtangent suffers the same problems - its damn hard to create something compelling + once you do it begs the question of why I want an internet format to create a single player game How many times have i seen 3D stock displayers and 3D weather maps on sites like Excite - boooorrriiiinnnggg I have been trying my hand at a 3D format for the internet these last three years - its called 3DML (an XML format). I think we have a good approach to making the creation easier - use building blocks - have symbols to represent each block - an "#" is a wall, a "l" is a light, etc. Now anyone with two brain cells can make a level map in a text editor. The behavior part is a little trickier but we are working on it... Now for the excitement part of the equation - we are not there yet - we made it easy to make levels that you can walk around but what exactly do you DO in these spaces that makes it worthwhile - i think the game industry has the lead in this - you create damn good interaction - so that is what we are working on - knock out eye candy and amazing interactions so you want to move around and engage the content... In my opinion VRML will NEVER get there...its a file format, not an interaction builder - and certainly not easy to work with unless you are both a good programmer and a good modeler.

atmosphere vs vrml

[malkuth23]

I worked with VRML many years ago, it is a very old format. No doubt it is cool in some ways, but it is bulky, demanding, and still unsupported. I recently tried out adobe atmosphere http://adobe.com/products/atmosphere/sampleworlds. html and i have to say that it kicks butt. Totally customizable avatars, easy to build worlds, links to 2d web info, etc. It also has infinite possibilites because you can incorperate javascript into it. This will be the future of 3d web. Adobe has the power to move this product and make it an industry standard. I am glad they moved beyound VRML etc. by starting from scratch they made somthing much cooler. check it out.

Re:atmosphere vs vrml

[3dpro]

Adobe? Adobe is great, and Adobe might even win vs vrml alone, but let's not forget about macromedia. Atmosphere is a neat toy if all you want are 3d chat worlds, but it will not become an industry standard because 3d worlds with avatars are not going to be a killer app for 3d on the web, at least any time soon. How many people actually want to do this? The builder app is cool but it is a bloated plugin from a company whose only luck with plugins so far is PDF. Games and CBT are what is going to drive web 3D in the near term (and not those moronic 3D cartoons either). That's where the $ is. Flash and shockwave are the only plugins people will actually use for interactive content like that. They aren't perfect either but with 3D in shockwave now (not a beta), along with its huge industry support, why would a 3D developer go somewhere else? Only drawback is that it is proprietary but the benefit is it will actually work and be on peoples machines! This is what will make a difference. Some day hopefully they will open up the format like swf or support an open standard like x3d if that ever takes off, but having support on users machines and function consistently is what is needed to get the ball rolling. Now if only Adobe and Macromedia would get together (no not likely). Atmosphere should use Shockwave to display their worlds. Why have yet another plugin? Ain't gonna happen.