API Question

Seeing as the new Java3D library is moving right along, I was wondering if there was any chance that the latest version of JME might be moving to using the primitives in that library vs. managing the source (namely the code in the vector math library).  For example, moving from JME's use of Quaternion to vecmath's Quat4F or using many of the other available primitive types.  Seeing as I'm still a newbie here, it's only an observation that I've made while playing around with the library on a project I'm working on.



Cheers!

I think it would be better, to put the JMonkey Primitives to the vecmaths Library  8)



Vecmath is not part of the JRE, and when you look in there sources, you will see there is to much overhead!



JMonkey has normal method and Local methods, in the vecMathstuff I always have to look in the docs to know how the it works

squid69 said:

Seeing as the new Java3D library is moving right along, I was wondering if there was any chance that the latest version of JME might be moving to using the primitives in that library vs. managing the source (namely the code in the vector math library).  For example, moving from JME's use of Quaternion to vecmath's Quat4F or using many of the other available primitive types.  Seeing as I'm still a newbie here, it's only an observation that I've made while playing around with the library on a project I'm working on.

Cheers!

Why would you want jME to use Java3D primitives?

The main reason to use the Java3D primitives is because Java3D is an API supported by Sun.  Therefore, Java3D and its supporting libraries are backed by JSRs or very precise specifications that dictate the functionality.  In Java, they are a standard regardless of whether or not they are in the JRE.  In my experience (and I've been coding for over 25 years, 10 of it in Java), adhering to standardized APIs and libraries helps adoption of the products being written because it tends to reduce the learning curve.  I realize that the vector math stuff is probably not the ideal example of this, it was just one that was more apparent to me as I was going over the jMonkeyEngine.



Again, it was more of a question than a suggestion.

squid69 said:

The main reason to use the Java3D primitives is because Java3D is an API supported by Sun.  Therefore, Java3D and its supporting libraries are backed by JSRs or very precise specifications that dictate the functionality.  In Java, they are a standard regardless of whether or not they are in the JRE.  In my experience (and I've been coding for over 25 years, 10 of it in Java), adhering to standardized APIs and libraries helps adoption of the products being written because it tends to reduce the learning curve.  I realize that the vector math stuff is probably not the ideal example of this, it was just one that was more apparent to me as I was going over the jMonkeyEngine.

Again, it was more of a question than a suggestion.

I see.. Well the math package in jME is only a small subset of it's feature set and functionality, the rest would still be loose on definition. In fact, many parts of jME are completely undocumented. Besides that it would be very costly to rewrite the engine entirely to use vecmath (there are differences both in API & implementation), not to mention all the extensions and user made code that would become incompatible.