Something new.

dedmin

Daniel S

Mike Lucey

This sounds great!

Wacov

I'm gonna have fun with this one!!

So, will all the newtonServer external commands be available for scripting, even if they're not implemented? I'm just wondering how we can get access to the materials

hobbnob

Kinda confused to be honest, mostly about the behaviour thing and ruby. Is sketchyphysics still usable for people who don't know ruby? A lot of the features mentioned need ruby to make them work, which is kinda worrying. . .

camokid11

@hobbnob said:

A lot of the features mentioned need ruby to make them work, which is kinda worrying. . .

To be honest, i dont like the idea, i hate scripting and the easier its made the better. Im still catching up on the current Sketchyphysics scripting (thanks to LazyScript). However the idea is good as its something difference but personaly, i think i will struggle with it.

Having said that, it wouldbe better to use a different Physics engine for sketchup.i guess idea really have to try it first to really give my proper opinion.

CPhillips

@wacov said:

I'm gonna have fun with this one!!

So, will all the newtonServer external commands be available for scripting, even if they're not implemented?

Yes. Thats the way it works.

CPhillips

@hobbnob said:

Kinda confused to be honest, mostly about the behaviour thing and ruby. Is sketchyphysics still usable for people who don't know ruby? A lot of the features mentioned need ruby to make them work, which is kinda worrying. . .

No, if anything you will need to know less. It will come with all the basic SP features and only new features will require ruby code.

Regarding the question of changing the physics engine. When I started this I took another look at all the physics engines out there and decided they were unsuitable for one reason or another. I even went so far as to get Physx running only to discover that it does not support cone or cylinder shapes. And convex hulls can only have 256 faces. Ug.

Wacov

@cphillips said:

When I started this I took another look at all the physics engines out there and decided they were unsuitable for one reason or another.

Is it switched to Newton 2.0... or is that not stable enough?

CPhillips

BTW. I had never heard of LazyScript . I must of missed it somehow. I am looking at it now. But from a glance it looks like an attempt to fix the same problem SketchyBehaviors is. Allowing others to write code that can be easily used by others.

Excellent work guys!

CPhillips

@unknownuser said:

Is it switched to Newton 2.0... or is that not stable enough?

Yeah, its the latest beta of 2.0.

camokid11

so will the scripts be like the ones show in the picture, or will they be similar to whats being used in sketchy physics at the moment

CPhillips

@camokid11 said:

so will the scripts be like the ones show in the picture, or will they be similar to whats being used in sketchy physics at the moment

The behavior scripts are an evolution of what was in sp3x. ontick{}, ontouch{} etc.

camokid11

so hopefuly not too hard then, ill have to try it when it comes out because ive got mixed feelings about it. One last thing, will it be compatable with sketchyphysics, eg: youcan run old sketchyphysics models on it, or have a bit of both worlds?

hobbnob

I don't think it will be tay, as it uses a whole new physics engine.

CPhillips

@camokid11 said:

so hopefuly not too hard then, ill have to try it when it comes out because ive got mixed feelings about it. One last thing, will it be compatable with sketchyphysics, eg: youcan run old sketchyphysics models on it, or have a bit of both worlds?

Sorry. Maybe simple models could be converted. But anything that used a controller field or scripted elements wouldnt work.

Wacov

Out of interest, how is the joint data gonna be stored... in other words, is the JCT making a comeback?

CPhillips

@wacov said:

Out of interest, how is the joint data gonna be stored... in other words, is the JCT making a comeback?

Joints will just be another behavior(s). I imagine most joints will not need a joint connection tool because they will connect themselves automatically via touching or something like "wheel" auto connects to "body".

When you do need to manually connect a body to a joint you will do it in the joints properties. Probably a button like "Attach all selected bodies to this joint". Only not as wordy.

CPhillips

Update. I put up a video of the current version. Its now working pretty well and there are a few interesting behaviors in place.

First the video shows a cone being created and then a behavior being created for it. Toggle color. This simple behavior allows an group/component to be two different colors. Another behavior can tell it to toggle. It starts out the first color. Once behaviors are started using the button at the top the cone changes to its first color. When behaviors are stopped it returns to its original color.

Thats kinda worthless by itself. But the second behavior added in the video allows the color to be toggled by clicking on the object. Note the Event and Signal field both contain "ToggleColor" That means when the click happens the toggle color behavior changes color.

Next is Interpolate. It allows a object to smoothly move from one "pose" (position+rotation) to another by moving a slider. Note that the slider works even if the behaviors are not running.

Next is Follow Path. The behavior moves the object, Sang in this case, smoothly on the curve selected. The speed is adjustable even when the behavior is running and he can be set to ping pong or loop.

Last are a few objects with the physics rigid body behavior.

Video best viewed in 480p or higher. Otherwise you cant make out much. Not sure why the frame rate is so low in the video its smooth in real time.

[flash=620,500:1zvyzk27]http://www.youtube.com/v/U1bar8gKVsI&hl=en_US&fs=1?rel=0&hd=1"[/flash:1zvyzk27]

Wacov

The rigid bodies seem to collide with a plane at ground level - will this be optional? Looking fantastic though, this will have many uses