CD> programmer, it made a lot of sense to me. I'd love to see more discussion
CD> like this. It's debate like this that moves technology forward.
Ahh, i would really like as well. And moreover, i got some topics (or
better say questions) to discuss. Wanna say beforehand, i greatly
respect all of u out here, so if anyone anytime would think that it's
not quite the place to do this kind of discussion, just let me (or,
hope, us) know this, so that we'd transfer it somewhere else.
The beginning:
--------------
Recently i tried the web demos (via the
Plug-in) and was quite impressed, by the physics, performance and SIZE
(most of all). I developed quite an advanced simulator in Java3D. Its
size is: ~100Kb of executables + 160Kb of graphics resources. I'm sure
the latter size could be reduced twice at least. The performance is
good enough as well. The dimension of the scene is 300m x 300m. Now,
i'm thinking of implementing smth with (or transferring this
application to) MathEngine. But before that i'd like to now if it's
worth it. Wanna say beforehand, i'm not familiar with MathEngine at
all. The questions are (guess quite a lot of people are interested in
knowing the answers):
1) MathEngine (later would be referred as ME) plug-in is just 200K vs.
J3D 2.4M. How this could be? Is the ME written over 10 times
efficiently then J3D, or does it have over 10 times less functionality
then J3D? Possibly the truth is somewhere in the midst?
2) The ME demos are great and rapid, but notorious HelloUniverse's cube
is rapid as well. We all know that performance tend to drop (quite
considerably) when there became more objects and behaviors in the
scene. How the ME would behave when the scene is increased hundred of
times (at least by the dimension of 300m x 300m), much more objects
and behaviors are put into the scene?
3) are the sound events supported? (quite a basic question, but...)
4) when (if ever) would the render used in Web plug-in be as efficient
as the one used in native .exe files?
5) is the ME technology aimed on Web based applications (games), or
the plug-in is smth like an advertisement, for it (ME) to be used in
native applications?
6) while the plug-in is small, 'the simple game written in one week by
one programmer' takes about 1Mb in archive and 1.5 as much in unpacked
form. Possibly, MeFastDynamics.dll, is a separate .dll which can be
used with all ME applications (which i doubt, actually), but even in
this case, the size of files to be downloaded is already over 1Mb. And
that's for a simple application, with not many textures and other
difficult-to-compress resources. How much could an advance simulator
take? At least only the executables, without any resources?
7) when the releases would be available and on what conditions (both 4
user and developer)?
8) is ME aimed on scientific applications or on entertainment (please,
not the advert slogans, but what the developers of ME do think)?
---------------------------------------
Well, i guess that's all 4 now. Would be glad to hear smth from ME
developers as well as from any of u out here who already gave a try to
this technology.
Thank u,
vladimir
-=V=-
>-------<=============>-------<
Join in Java community now!
http://JavaCafe.VirtualAve.net/
>-------<=============>-------<
In your previous letter u wrote:
--------------------------------
CD> Let me say that I enjoyed reading this, and although I'm a novice amateur
CD> programmer, it made a lot of sense to me. I'd love to see more discussion
CD> like this. It's debate like this that moves technology forward.
CD> [Casteel, Don]
CD> -----Original Message-----
CD> From: Anselm Hook [mailto:[EMAIL PROTECTED]]
CD> Sent: Tuesday, February 08, 2000 3:58 PM
CD> To: [EMAIL PROTECTED]
CD> Subject: Re: [JAVA3D] MathEngine is " Windows only " : Y U C K !
CD> Hmmm, I'll comment on the below message since I'm working away on various
CD> projects here at MathEngine including the OMG stuff.
CD> First of all, Java3D is not intrinsically platform independent - you have to
CD> ask the vendor for a port to your platform. Fundamentally it's no more nor
CD> less platform independent than MathEngine. J3D has to bind to target
CD> hardware on each platform, and MathEngine has to support numerical methods
CD> optimally on each platform. The PS2 port of MathEngine for example is a
CD> total rewrite optimized at the assembler level for the rather unique
CD> characteristics of that device. There exist implementations of Java3D for
CD> Windows and Linux and Solaris and Irix. MathEngine runs on those platforms
CD> and on the PS2 - albeit not yet in commercial release. Secondarily Java3D
CD> deployment is still not mature. Deploying J3D based applications requires
CD> more than a single click, and this makes J3D untenable for web based
CD> content. The problem fundamentally is an idealistic attitude at Sun, which
CD> eschews interoperation with legacy frameworks such as Windows in favour of
CD> their own entire model of reality. Much of the good technology coming out
CD> of the Sun effort is difficult to install, difficult to deply. Technology
CD> is irrelevant if deployment is difficult. Thirdly, the first iteration of
CD> MathEngine technology was free to download at the very least - nobody else
CD> has put themselves up to that kind of general scrutiny including Ipion. The
CD> proposition has always been to only charge people for genuine value - if
CD> there was genuine value and a genuine relationship then dollars should be
CD> negotiated. Anyway, direct scrutiny has resulted in intense technical
CD> feedback and intense improvement to MathEngine as a result. The new engine
CD> is orders of magnitude faster, energetically stable, much smaller, and uses
CD> entirely new algorithms and methods not based on Baraff. GDC 2000 (
CD> www.gdconf.com <http://www.gdconf.com> ) w
CD> I think everybody agrees on and believes in the fundamental value of
CD> addictive web based multiplayer games with rich emergent behavior. We've
CD> all have a vision of making that kind of killer web app - it's kind of a
CD> holy grail. I actually suspect that interactive experiences / games will be
CD> the dominant form of entertainment and education within a decade and that
CD> cinema will fade away. But ever since starting to work with dynamics last
CD> year I've realized that graphics isn't really as important as underlying
CD> behavior in achieving these kinds of goals. A formal approach to behavior
CD> systems based on newtons laws could cover most of the typical cases that
CD> people imagine when they imagine building immersive experiences. Best of
CD> all it would allow different game developers to make different objects that
CD> actually could interact together. We've never had a "lingua franca" for
CD> describing the interaction of behavior systems. Many efforts to provide
CD> game building toolkits have failed because different models of object
CD> behavior and interaction were insufficiently compatible with each other. I
CD> believe that fundamentally most objects can interact adequately through
CD> physical properties such as position, collision and the variety of other
CD> forces and constraints that we are so used to perceiving as human beings.
CD> Perhaps the place this is all going to happen first is the PS2. It may very
CD> well be the first umbiqitous computing platform.
CD> - Anselm Hook
CD> (Note that we visited Sun and talked about the points below ( that we
CD> provide Java Dynamics for them for free). They suggested that we go through
CD> the java community standards approval process, which we may do but haven't
CD> yet committed to. We also wanted a port of J3D to PS2 but it turns out that
CD> this is harder than first appears so that is also on hold. So solutions
CD> today are still somewhat fragmentary. It's clearly our ambition to solve
CD> this and everybody is keenly aware of these issues. )
CD> -----Original Message-----
CD> From: Discussion list for Java 3D API
CD> [mailto:[EMAIL PROTECTED]]On Behalf Of Matthew Gahan
CD> Sent: Sunday, February 06, 2000 10:50 AM
CD> To: [EMAIL PROTECTED]
CD> Subject: Re: [JAVA3D] MathEngine is " Windows only " : Y U C K !
CD> Mathengine isn't the only runn www.ipion.com <http://www.ipion.com> should
CD> be considered a major contender.
CD> Not least because Ipion's costs a tenth of what Mathengine's kit costs
CD> ($5000 ~ish for Ipion with some royalty's...).
CD> According to the doc's this is a highly modular native c/c++ library, which
CD> I am guessing would lend itself well to JNI integration...
CD> The Open Media Group <http://www.openmediagroup.com> managed this with
CD> Mathengine. The results were ok...but still slugish by current standards...
CD> Ipion has a very low overhead (20% cpu usage estimate). It also has some
CD> very nice optimizations for dealing with very large numbers of Free body
CD> systems (ideal for particles).
CD> -----Original Message-----
CD> From: The Casteels < [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> >
CD> To: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> <
CD> [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
CD> ===========================================================================
CD> To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
CD> of the message "signoff JAVA3D-INTEREST". For general help, send email to
CD> [EMAIL PROTECTED] and include in the body of the message "help".
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA3D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
