This is quite a step-up for Papervision3D and will allow more developers to start exploring the platform. There is a montly subscrition you'll need to pay to watch the entire series but if you do let us know how it worked out for you and if it's worth it

Have you read it? Let us what you think in the comments.

Are you unsure about getting it? Get the sample chapter here.

Here's a demo of Remotely controlling a desktop AIR 2.0 app with an Android G1. The scene that is controlled is a Papervision3D scene and this is a featured example in the book.

Pretty cool ham?

Sometime along the development of the application I came across the need to convert 3D coordinates into 2D screen coordinates and also the opposite. It's nothing new and it's already implemented in papervision, but since it took me a little while to find it, I decided to share it.

Getting the 2D Screen coordinates from a DisplayObject3D

Since papervision 2.0 that every DisplayObject3D has a property called "screen", that will return the coordinate of the object on screen. Whenever I tried to access the screen property the number3D I got back was always (0,0,0). Probably because it's an expensive operation and it's something you only need sporadically, papervision will not calculate this automatically. You need to ask papervision yourself to calculate this coordinates. To do so, just use the "calculateScreenCoords" of the do3d

myDisplayObject3D.calculateScreenCoords(camera)

Now when you try to access the "screen" property you will get the number3D with the screen coordinates for your DisplayObject3D.

Getting the 3D coordinates from a 2D point (x,y)

For this one I found a great post by zupko that tells you exactly what you need to do.
In a nutshell, you need to take your 2D coordinates (x,y) and convert them into a direction vector. Having the vector and the camera position you can create a ray (line) that will go from the camera into infinity. Next you need to specify the Z, or in another words, where do you want to stop on that line. Zupko uses a 3D plane to do that using the "getIntersectionLineNumbers"

Just remember papervision's 3D Zero (0,0,0) is at the center of the screen and not on the upper left corner like the 2D Zero (0,0)

pv3d_gallery_as3project

They include an air contact manager, and two sound equalizer how-tos with one of them showing how to get those cool pv3d equalizers.

Enjoy!

So I finally decided to go have a look at Papervision's destroy methods, and surprise surprise they weren't being called, no harm there, it's alpha these things are meant to happen

So here's what you have to do to force those destroy methods to execute, and clean your memory.

In org.papervision3d.core.proto.MaterialObject3D  set the destroy method from protected to public, do the same to all the materials that you use, like org.papervision3d.materials.MovieMaterial . When you're done with these just the destroy method in your class. For me it looks something like this:

Hope this helps.

We've all been waiting for this moment for a very long time, to have physics
and 3D come together, so without further due here's a little demo I've thrown together in an hour or so.

The code will follow shortly, I just need to tidy it up .

demo

Enjoy.

So here's a little demo on the new Phong shader :) As soon as i have the time tonight i'll put on a much nicer one. This is just to show that altough PV3d just got a new whole set of features it's still easy to use

PhongCube

So what you'd expect to happen above is to have a new bitmap that looks the same as my9SliceEnabledMc,
but what you get instead is the original 40x40 movieclip.
This doesn't make sense at all, i have no idea what Adobe was thinking when they let this one through.
This was driving me crazy for a while now, and today i finally found the answer, as simple as it seems,
putting the my9SliceEnabledMc inside another sprite corrects the issue!!

//my9SliceEnabledMc is a 40 by 40 9-slice enabled movieclip on the library
var myContainer = new Sprite();
var my9SliceEnabledMc = new my9SliceEnabledMc();
myContainer.addChild(my9SliceEnabledMc);
my9SliceEnabledMc.width = 200;
my9SliceEnabledMc.height = 300;
var myBitmapData = new BitmapData(200,300);
myBitmapData.draw(myContainer);
addChild(new Bitmap(myBitmapData))


Also i should add that this is a major issue with PV3D since you cannot directly use a 9-slice item as a texture.

The most important of these branches is the one called Phunky. It is the branch that will soon became the 2.0 version of Papervision3D, which in the SVN trunk is now tagged Beta 1.7 (dated 20.08.07). So, if you're using trunk, you should expect only bug-fixes. It's a bit different from other open-source projects where trunk normally means the latest revision of the project. For Papervision3D, you can think of trunk as a stable version (whatever that means for a software that's in Beta).

Someone said on the list that "switching from trunk to Phunky shouldn't break anything (doesn't impinge on the API) but should net you a performance increase".

MaterialsRestructor was a branch created to test code regarding Materials. Once changes are tested the most likely is that they go into trunk or Phunky.

I don't know much about the 'ascollada' branch besides the fact that it was created by Tim Knip and that he tries to maintain it in sync with ascollada, a actionscript library for parsing COLLADA files. I've tested loading some COLLADA files with Beta 1.7 and it works fine so I don't have the need to use this branch yet but Ivan is using it (the DEA class), so he might have something to add here.

The 'frustum' (definition) branch has a 'BVH mocap importer class' with a beautiful sample here, but I was unable to find the real purpose of this branch. You can check the examples folder for the sample sources.

'timk_frustum' is a branch that appeared when Tim Knip put up for testing a new FrustumCamera3D, in early August 2007. This class is now also at the 'frustum' branch.

And that's it. If you know more about this and want to share please do. Thanks.

Update: John Grden posted a comment explaining a bit more about the branches, thanks!