Well, we finally got there. Xj3D has reached that mythical 1.0 version number. We were wondering just how long this has been going, but as best as we can tell, it was just over 8 years ago that Sun released the first version of Xj3D to the world. For the past 5 or so Alan and I have been principle maintainers and done the bulk of the development work - sometimes paid, sometimes gratis, always fun.
You can download Xj3D from either the usual place on Xj3D, but the stable release is on web3d.org. They have far more bandwidth than we do, so please download from here rather than xj3d.org:
http://www.web3d.org/x3d/applications/xj3d/
We’ve always taken a very different tact to the other browser implementations. We very strongly believe in conformance to the spec first and foremost, and forcing that conformance onto all users. Conforming content makes for portable content. If it runs on Xj3D, it will run on any other browser (excepting bugs or lack of implementation, of course). Even with this approach, we’ve demonstrated that you can be conformant and maintain good performance. For a long time we were the fastest browser out there - despite being written in a “slow” language like Java - faster than anyone on any platform in any language. Even today, we’re still right at the top of the performance stakes, usually loosing no more than 10% compared to the other implementations. We believe that by continuously pushing the envelope both in performance and conformance, we raise the bar for all browsers - if we can do it, why can’t you?
Xj3D is still the place where a lot of experimentation takes place. In the past we have prototype many extensions to the X3D spec - physics, fog, 3D textures, CAD, device abstraction, binary formats and more all began life as Xj3D extensions. We aim to continue pushing that envelope with our near term future. Almost immediately we are about to tear the codebase apart and implement the layering spec and use that experience to filter back the collection of massive changes that this is bringing to the core of the X3D specification.
Despite all this, and the open nature of the codebase, Xj3D has found itself used in all sorts of areas. It’s amazing where it pops up. For example, we are aware that several versions have been approved and deployed in highly classified active military systems. These are big votes of confidence in our work.
Finally, Alan and I want to pass on a HUGE thanks to everyone that has helped us over the years. Firstly to Brad Vender for all his work implementing the EAI and external SAI and to Rick Goldberg and Sun for the initial donation of the codebase. Next, to everyone that has tested Xj3D over the years, sending us bugs, the occasional fix, and putting up with the sometimes wayward release handling and capabilities that break all the time. Lastly, to all those companies and institutions that have helped the development of Xj3D through financial means: Naval Postgraduate School, NASA, NIST, University of North Carolina, Sun, North Dakota State University (Brad’s employer) and the others that ask not to be named.
Thanks to all, and we look forward to a bright future for Xj3D!