Let's make Google adopt VRML

[Hiperia3D - Jordi R Cardona]

Please read:

http://news.hiperia3d.com/2009/04/another-standard-for-web-3d-googl...

I have submitted the suggestion to make O3D adopt VRML or X3D in classical vrml encoding. The option is getting votes.

PLEASE GO THERE AND VOTE! Let's make VRML and X3D be supported by O3D!

Just login in Google account and go to:

http://moderator.appspot.com/#16/e=41eb1

And vote the option:

"Create support for VRML or X3D in classical VRML syntax because it's an already existing standard for web3D and it's easy for everyone"

[Steve]

"The 14th Annual Web3D Symposium Celebrates Milestone of Innovation and Durability

The fourteenth annual International Web3D Symposium will showcase X3D, the only open, royalty-free and ISO-certified technology available today for interactive 3D graphics on the World Wide Web. "
This statement isn't entirely true. VRML is also open source and royalty-free, available for interactive 3D graphics on the World Wide Web.

I suppose it depends on what you want your 3D to do but I haven't seen much widespread use of X3D to date. I still see alot more vrml on the web than anything although some of the other 3D platforms are gaining a presence. I still don't get the X3D thing. If it is so phenominal then why does it still have to use VRML syntax in order to make it work? To me it seems like kind of self defeating if it cannot stand on it's own as a language. instead of 1 language, you carry 2 languages forward as your standard and they are not seprable if you wish to call X3D your chosen standard.

BTW most of Bitmanagement's demo's are bugged and inoperable so it is really hard to tell exactly WHAT they have to show us sometimes. I am not sure if they did this on purpose or not but most of the demos or examples I have tried to view had many errors and wouldn't even load. Would be nice to see some of these demos so that we can incorporate into our 3D content, which is being geared towards BM's product to begin with. One would think it is advantageous to them for more content being developed that showcases what BS Contact can do.

[Hiperia3D - Jordi R Cardona]

It's not two languages but two syntaxes. The XML syntax, that is what some programmers like, and is what is usually supported in many applications, and the VRML syntax, that is what EVERYBODY likes.

The XML syntax has advantages for external applications to read and write info because of the XML structure that makes interchange easy, like happens with COLLADA.

But for humans, XML looks like a tag soup. You have to write 20 times more and it is not readable at all.

X3D in classical vrml syntax is like VRML but with a different extension for the file, and a different header. The rest is the same, plus many other new nodes. So VRML can be seen as a subset of X3DV.

The applications for VRML have the same problem as with X3D: every business does what they like. This problem is common to many other languages.

[Steve]

"But for humans, XML looks like a tag soup. You have to write 20 times more and it is not readable at all."

This is part of what I mean. If it is so much more difficult to program with XML then that will leave a heck of alot of people out of the mix on the web. Only experienced programmers will be able to do anything worthwhile and the art form of vrml will be lost. There is not alot of 3D artists that want to sit in front of a text editor and create tag soup they hopefully someday will be able to call a world.

VRML has several EAI's that can communicate with it effectively. Java, JS, PHP and a few others. I know there is always going to be some issues but at least it leaves the door open for newcomers to express themselves with VRML without having to deal with XML besides. At least with vrml, most of the modelling programs will export .wrl effectively so people without vast programming knowledge can put their work on the web for others to share.

I understand what you are saying Jordi about X3d being a "syntax" but technically XML is a "language".
Xml was around before X3D was adoptyed as the new standard by the consortium and therefore was completely seperate from VRML at one time. To say that they are the same thing but just a different syntax doesn't really make any sense. The two are completely different actually. Basically by wrapping VRML inside an XML code structure you have an "IAI" (internal authoring interface) and the vrml is just running inside of an xml shell. This xml structure is what is preventing many people from getting on-board with this standard. If scientists prefer to use x3d for their models then that is fine. If there is a business strategy that can benefit from it, that is great too. But when you talk about an open source standard for the web, one might consider who the target group of programmers and authors will be. Most people that would use VRML are not as likely to adopt X3D just because of the complicated creation methods. I personally feel it would make more sense to adopt a strategy to employ XML as an EAI and leave VRML alone sans making improvements to it. If XML is a legit "syntax" then there shouldn't be any reason it shouldn't be able to access VRML nodes externally and leave the jumbled, two-syntax out of the equation. The nice thing about VRML is it's simplicity. Just about anyone can learn it's syntax and build creations that all of us can enjoy.

That being said, I am not completely against furthering research into X3D, however I feel that it really is not going in the direction that the consortium thought that it would. I understand that it is too late to revert to the former but this decision really has divided the 3D community and several other new options are dividing further. Apparently some of the VRML people are not the only ones who would rather not use X3D. That is why we have such diverse 3D structures on the web today. This is also why companies that make browsers are still making them compatible with both because many still choose to use VRML over x3d.

It would be great if the 3D community could be united in there search for a better structure. I still feel VRML could have had vast improvements to it if that had been the choice of the consortium. The differences between vrml1 and 2 was immense and rather impressive. Who knows what vrml3 might have brought to the web ;)

[Doctor]

Steve said:

Who knows what vrml3 might have brought to the web ;)

It is my understanding that X3D (in the x3dv syntax) is essentially VRML 3.0, just with a different name.

Personally I do not see the problem with two different syntaxes for X3D, as long as each syntax supports all the features that the other one does. This just makes X3D even more appealing, since each person can adopt whichever syntax suits their needs best. (i.e. If adding 3D web content to existing applications that already generate XML based files is the need, then the X3D/XML syntax is probably preferrable. If opening the 3D content file in a text editor and hand coding is the need, then the x3dv syntax is probably preferrable.)

In my opinion, there are more important considerations when choosing a 3D web format, such as:
-- Is the language extensible and are there any restrictions or difficulties with using the extensibility features?
-- Does the 3D content syntax offer more compact files than other 3D formats?
-- Or (if not smaller files) does the language offer ease-of-adoption by conforming to already popular standards (e.g. XML)?
-- Does the language also provide for connectivity with other programming languages through a fully featured external authoring interface?
-- Does the language easily map onto 3D scenegraph heirarchies that are common to CAD systems, gaming, scientific, medical and other 3D applications? (i.e. If the scenegraph heirarchy of a 3D application is not shape primitives as children of transforms and other group nodes, and sensors and lights are not handled as children of group nodes, then probably neither X3D nor VRML are suitable as a file format for this 3D application.) In other words, if a simple traversal of the scenegraph of your software (with simple output at each graph-node) is sufficient to generate an X3D compliant file, then X3D is probably in nearly one-to-one mapping with the internal scenegraph of your software and would be suitable as a file format for your system.
-- and there are probably other considerations.

The fact that X3D has two syntaxes seems to me to not be a valid criticism of the X3D specification, and in fact is one of its strengths. In my opinion, that is :)

Note: It isn't my intention to infuriate nor annoy anyone, just to fuel more discussion with my two cents worth of opinion. :)

[Steve]

Doctor,
The point I was trying make was...

If a newcomer to 3D wanted to create some content for the web, how easy would it be for them to learn X3D? to try and learn xml syntax and VRML syntax both is probably a bit much for many newcomers. Without an infusion of new blood occasionally, most Communities(colonies) are not viable. If X3D is going to be widely used in science and medical capacities then that doesn't do much for the web either. I guess my overall point is that it isn't really a friendly overall language for the web in general. It has more advantages for those who are programmers rather than artistic developers.

as far as X3D being the same as vrml 3.0, that isn't really true although I have heard it called that. VRML 3.0 denotes that a new version or updated version of VRML was made. That isn't what happened. The consortium used the exsisting form of VRML and just put a new wrapper on it. A wrapper than many people don't really want to use. And then if they used the exsisting form of VRML then they also are limited by all the shortcomings of vrml2.0 since that is the base syntax. X3D cannot function without the old VRML 2.0 syntax. That alone sort of makes a statement as to the validity of X3D as it's own entity. When you have to rely on the old tech in order for the new tech to function then perhaps they need to re-evaluate this standard. :P

And btw no worries about infuriating or annoying. This is a discussion board. thats what we are doing is duscussing. ;)

[Doctor]

Steve said:

Doctor,
The point I was trying make was...

If a newcomer to 3D wanted to create some content for the web, how easy would it be for them to learn X3D? to try and learn xml syntax and VRML syntax both is probably a bit much for many newcomers. Without an infusion of new blood occasionally, most Communities(colonies) are not viable.

But a newcomer to 3D that wanted to create content would most likely use a 3D authoring tool, not code directly into a text file as advanced VRML designers do --- so the format of the X3D file is not as important as the features that X3D supports (i.e. NURBS, shaders, physics extensions, etc.). And if an intermediate content creator wanted to start hand coding X3D then they need only learn the x3dv format (i.e. the VRML97 style of code). There would be no reason for them to learn the xml form of X3D.

If X3D is going to be widely used in science and medical capacities then that doesn't do much for the web either. I guess my overall point is that it isn't really a friendly overall language for the web in general. It has more advantages for those who are programmers rather than artistic developers.

And, once again, I would claim that the artistic community would just continue to work with tools that they usually work with (i.e. 3D authoring applications like 3DS Max, Maya, Blender, Vivaty Studio, etc.), so the syntax of the content file itself would not be very important if 3D authoring tools just supported the X3D standard. I have never heard of a 3D content artist that codes directly in any proprietary format like that of 3DS Max, trueSpace, Maya, etc. --- and I do not think they would need to code directly in X3D XML syntax either. On the other hand, the x3dv syntax option of the X3D standard is still available for the truely advanced designer or developer (such as in some science and medical projects that are developing specialized 3D software), which is more than can be said of most other 3D content formats (most of which are binary encoded, not text).

as far as X3D being the same as vrml 3.0, that isn't really true although I have heard it called that. VRML 3.0 denotes that a new version or updated version of VRML was made. That isn't what happened. The consortium used the exsisting form of VRML and just put a new wrapper on it. A wrapper than many people don't really want to use. And then if they used the exsisting form of VRML then they also are limited by all the shortcomings of vrml2.0 since that is the base syntax. X3D cannot function without the old VRML 2.0 syntax. That alone sort of makes a statement as to the validity of X3D as it's own entity. When you have to rely on the old tech in order for the new tech to function then perhaps they need to re-evaluate this standard. :P

That is not my interpretation of the statements made by the Web3D Consortium in (for example) the following links:

What is X3D?
Why should I migrate from VRML97 to X3D? - The Top 10 Reasons
Can I play VRML97 files in my X3D browser?
What functional changes are there between VRML97 and X3D?
VRML is extensible. Why is X3D more extensible than VRML's EXTERNPROTO mechanism?
Is X3D really just VRML-Lite?
Do I really need to use Maya/3DSMax/AutoCAD/etc to write X3D?
Why bother with Extensible Markup Language (XML)?
How many three letter acronyms will I need to understand in order to use X3D?
Overview of X3D
X3D Plugfest
X3D Node Reference
X3D Specification ISO/IEC 19776-2.2: Classic VRML Encoding

Compare the contents of the last link above to the equivalent specification for VRML97 (i.e. VRML 2.0) at the following link:

VRML97 Specification ISO/IEC 14772-1: Part 1, Functional Specification

The list of supported nodes alone should convince anyone that X3D is not just an XML wrapper around VRML97.
Nor is X3D just inheriting the drawbacks of its predecessor VRML97, as should be clear from the FAQ links of the Web3D Consortium (some provided above) and the X3D specification (for anyone masochistic enough to study through these documents --- like me, lol).

Just another two-cents --- Canadian currency. :)

[Hiperia3D - Jordi R Cardona]

But for example, look at a collada file: http://www.collada.org/mediawiki/index.php/Skinning

It is also a tag soup and COLLADA is taking some advantage over X3D.

And X3D has lots of new nodes, it is not just the old VRML with another name. The problem may be that there has been little emphasis on the vrml syntax, and lack of some features for gaming and realtime 3d like collision detection and the identification as a same thing of avatar and viewpoint.

[Steve]

"And, once again, I would claim that the artistic community would just continue to work with tools that they usually work with (i.e. 3D authoring applications like 3DS Max, Maya, Blender, Vivaty Studio, etc.), so the syntax of the content file itself would not be very important if 3D authoring tools just supported the X3D standard."

So then, Does 3DS Max and/or Maya support x3dv? because last time i checked they did not. which means the artistic designer would have to be able to code x3dv directly or they would not be able to display their work. I do know however that both of those programs export to vrml 2.0.

You do make a compelling argument Doctor but it still kind of sidesteps my point about potentially new designers and their ability to export and display their work directly from their design program. If the 3D design programs do not support or export x3dv then it really isnt attainable for a new designer. With VRML a creator can start out building things and exporting them directly then work themselves into the coding later if they choose. With x3d it is almost necesary for them to have a working knowledge of vrml AND x3d before they even can hope to display their work on the web, except for screenshots or video.

[Steve]

Hiperia3D - Jordi R Cardona said:

But for example, look at a collada file: http://www.collada.org/mediawiki/index.php/Skinning

It is also a tag soup and COLLADA is taking some advantage over X3D.

And X3D has lots of new nodes, it is not just the old VRML with another name. The problem may be that there has been little emphasis on the vrml syntax, and lack of some features for gaming and realtime 3d like collision detection and the identification as a same thing of avatar and viewpoint.

I agree. Although Collada seems to be a jumble of tags, it appears to have some pretty nice features and there are also 3D design softwares which export collada. Personally I think the consortium should have spent some time improving the vrml syntax itself if they intended to keep using it.

[Doctor]

Steve said:

"And, once again, I would claim that the artistic community would just continue to work with tools that they usually work with (i.e. 3D authoring applications like 3DS Max, Maya, Blender, Vivaty Studio, etc.), so the syntax of the content file itself would not be very important if 3D authoring tools just supported the X3D standard."

So then, Does 3DS Max and/or Maya support x3dv? because last time i checked they did not. which means the artistic designer would have to be able to code x3dv directly or they would not be able to display their work. I do know however that both of those programs export to vrml 2.0.

You do make a compelling argument Doctor but it still kind of sidesteps my point about potentially new designers and their ability to export and display their work directly from their design program. If the 3D design programs do not support or export x3dv then it really isnt attainable for a new designer. With VRML a creator can start out building things and exporting them directly then work themselves into the coding later if they choose. With x3d it is almost necesary for them to have a working knowledge of vrml AND x3d before they even can hope to display their work on the web, except for screenshots or video.

I thought this whole discussion thread was about convincing software development teams (in this case Google's O3D) that adopting X3D as the primary content format would be a good option compared with the alternative formats, and not a discussion about artists having to learn hand coding because their favorite authoring tool does not yet export x3d nor x3dv. I am forwarding the argument that X3D is a good choice (possibly the best choice) for development teams to adopt for their browsers and authoring tools, and as a side benefit I am offering that the x3dv format is just as easy as the wrl format (VRML) if an advanced creator wants to hand code (and it has more features and fewer drawbacks than VRML97).

There is no argument from me that 3DS Max, Maya and other popular authoring tools currenty do not export X3D, but I think that they should and that we should place our support behind the Web3D Consortium by asking the vendors of our authoring tools to add X3D support.

By the way, there are already tools to convert VRML97 into X3D (either x3d or x3dv format), as well as content authoring tools that support X3D. See, for instance the list of tools at the following link:

X3D resources

The more tools that adopt X3D, the more content we will start to see. The more X3D content we start to see, the more likely it will be that 3DS Max, Maya and others will adopt X3D export. Let's start somewhere and petition Google's O3D team to adopt X3D as their primary content format --- assuming that you believe X3D is feature rich enough, compared with collada and other formats.

I have a question for anyone reading this discussion: If you have reservations about X3D then what 3D Web content format would you propose would be better for Google's O3D project to adopt, and why?

[Steve]

"I thought this whole discussion thread was about convincing software development teams (in this case Google's O3D) that adopting X3D as the primary content format would be a good option compared with the alternative formats, and not a discussion about artists having to learn hand coding because their favorite authoring tool does not yet export x3d nor x3dv."

Actually check the title of this thread again. it reads "Let's make Google adopt VRML". This topic was started with both in mind but did not stipulate that this was about X3D alone. Now i understand that X3D is part of this discussion and I am not saying that it doesn't have a place. But Please understand there is a big division amongst the vrml x3d community. ALOT of people still favor VRML and it obvious with poles and overall popularity.

That being said I think this is a good place to leave this piece of the discussion.

"I have a question for anyone reading this discussion: If you have reservations about X3D then what 3D Web content format would you propose would be better for Google's O3D project to adopt, and why?"

Google sort of walks it's own walk. I personally don't believe they should be trying to support their own standard for the web but that is not within our control. The thing that IS in our control is how popular our own choices are for the web. To create compelling worlds and content that attracts others to the language and to initiate them to this style is our charge. If we do whatever we can do about this then maybe people will come back. The Biggest problem is that everyone has their own idea of what the standard should be and as long as thier are divisions this battle will rage. I would think that the most unified community will probably win in the end.

[Hiperia3D - Jordi R Cardona]

2 X3D exporters for 3D Studio Max:

http://www.edgewise-consulting.com/products/vrmlexp/

http://www.unrealroc.com/MaxExporter.html

And 1 X3D exporter for Maya:

http://rawkee.sourceforge.net/


The intention was to tell Google that standards for Web3D already exist, and that at least they should see that they exist before reinventing the wheel again, and going through the long process of standarization.