Saturday, April 28, 2007

VRML and The MAC: Which Failed?

Elliotte Harold decides that VRML failed because it wasn't on the Mac. Jason Osgood claims it was because of XML.

Why is it on the web that people who don't do this kind of work, don't use a language, and aren't a part of the community that does get to decide that a technology is dead?

I say go to web sites such as www.octaga.com and look at the work being done in X3D. Osgood, you gave up too early. Your's is a common complaint among those who did. Is XML uglier than classic VRML syntax? XML is uglier than almost any alternative syntax. But it comes with millions of lines of supporting code so while you are in a vocal minority, you are also in a very small one. As the song says, "you made your move too soon".

If all X3D is is a transcoding, I would agree. It is an ugly transcoding. On the other hand, significant improvements were made in the VRML design that you overlook including much improved nodes, a much better object model, a much better API (SAI) that supports AJAX.

Blaming the standards group for making your move too early is like blaming your parents for your last failed relationship. They may shape your decisions but they don't make them for you. You turned on it early when it disappointed you, and you won't look ahead or give those who stayed the course credit for what they have achieved. That's mean-spirited and not just a little self-immolating.

If VRML had failed, it wouldn't still be the single largest supported format for all tools combined when importing and exporting. If VRML had failed, companies like Octaga and BitManagement wouldn't be doing multi-million dollar 3D projects today. If VRML had failed, worlds built ten years ago wouldn't still be running in browsers built in the last two. If VRML had failed, the US Homeland Security department wouldn't be using it in HLS projects for training emergency responders nor would it be a US Navy approved standard.

What VRML could not do as has been pointed out so often is make 3D designers out of web page builders. It couldn't make real-time 3D as easy as document design. What it can do and is doing is provide a royalty-free unencumbered means to create 3D on the web and off. It saved a information space for use by those who cannot afford Maya but can master tools such as Flux Studio because they have the time and are willing.

You are missing the point, Elliotte, because you need a whipping post for another project that isn't winning mindshare as fast as you want it to. Like so many I've talked to about Open Office, you don't want an open unencumbered royalty free means to create and maintain information, you want to beat Microsoft. You want to hurt them. You want it so bad that you'll tell any story you have to and make the point on any back you can find. You are a better man than that.

X3D/VRML is succeeding. Do your homework and look at the URIs. Look at the projects. Look at the software. I don't want EVERYONE to use X3D. We learned in VRML that too few have the talent and that ensures just as you find in SecondLife that a lot of the 3D is terrible. If they want to try, we provide the tutorials for free, software for free, and guarantee the lifecycle. That is all that can be expected. If they can't make a go of their business or their relationships, that's life. The world will let you die, it will let you starve, it will let you lose. All a standard can provide is a chance to live, to make money, and to win at the business or hobby or art that you choose, but if you don't, it can't.

If you don't need 3D, don't use it. If you do, look at the costs of building it and maintaining it because it is by nature one of the highest cost content types, so if you do use it, figure that out fast because you will burn your customers just as the early wysiwygers before HTML and XML burned customers with closed encumbered systems that died without even a chance of rehosting. Then take another look at VRML/X3D and ask if the only metric for winning in some markets is surviving. Osgood has an axe to grind. Elliotte, you have a cause and may you get what you need, but don't do it on the back of a worthwhile effort that has done everything necessary to succeed with its values intact and its content still wrling a decade later. There aren't too many pre-dot.bomb standards that can make that claim and those that can did because their real users kept them alive. Don't insult them because you didn't.

Caveat vendor.

1 comment:

Elliotte Rusty Harold said...

Len,

First I didn't say it was dead. I said it failed. That's a very different thing, and one I still hold to. It did fail, rather obviously.

I get to say that because I did use it (or try), did write about it, and was very much a part of the community in the early days. I and others in Silicon Alley spent a lot of time trying to make this stuff work. Unfortunately it didn't work, the time was wasted, and if we wanted to get anything done we had to look elsewhere, whether we were doing 3D or web sites or something else. VRML was a pointless, useless technology.

Is it still a pointless, useless technology ten years later? The jury's still out, but at this point the burden of proof rests squarely with people who claim it's not. If you can do something useful with VRML now, great; but don't deny history. VRML failed in the 1990s. Unless you understand why it failed and repair those problems this time around, it will only fail again.

Comment Policy

If you don't sign it, I won't post it. To quote an ancient source: "All your private property is target for your enemy. And your enemy is me."