This tutorial is going to cover bindable nodes. Now, these aren't new nodes, just a new way of using them. Certain nodes that affect the general appearance and behaviour of the world are bindable, that is they can be bound and unbound. When a node is bound, the settings within it affect the state of the world. When a node is unbound, another node of the same type must be bound, so the settings in this other node affect the world. In this way, you can switch between a number of different settings for a world by setting which of the nodes is bound.
We've already met all of the bindable nodes, so you should know what they do already. These are the ones:
As you can see, we've come across all these before, and as you should know, they all affect the overall appearance and behaviour of the world. Now, it's time to discuss the internals behind node binding; how it all works.
OK. The browser maintains four structures inside it's inner workings, one for each type of bindable node. These are called bind stacks. When a VRML file is parsed, the first node of the appropriate type is placed on top of it's respective bind stack. Subsequent nodes are added to the bottom of the stack. The currently bound node (ie the one that is active) is the one one the top of the stack. Once the world has loaded, we can manipulate the stack with a certain set of events.
All bindable nodes have a couple of events in common. They all have the set_bind eventIn and the isBound eventOut. These are both SFBool values. When you send a set_bind event to a node with a value of TRUE, it is moved to the top of the stack, and so becomes the bound node. This node will then send a TRUE isBound event. The node that was on the top of the stack sends a FALSE isBound with the same timestamp. One exception, which is that if the node is already on top of the stack, nothing happens and no events are sent.
If you send a FALSE set_bind event to a node, it is removed from the stack for good. If this is the bound node, it sends a FALSE isBound, and the next node in the stack is made the bound node. If not, it is simply removed without further ado.
If you send set_bind events to a node that has been removed from the stack, it has no effect and nothing further happens. If a node is deleted, it acts as though is was sent a set_bind FALSE event.
Now that you know how to use bindings, what are they useful for? Well, being able to change the background and fog styles is quite useful, as you can imagine. Also, by using different NavigationInfo nodes, you can resize your avatar, change the movement style, all sorts of things.
However, the most useful node is the Viewpoint. If you bind the user to a viewpoint, he will be moved there instantly in the style defined in the node (ie its jump field) This means you can have scripts moving the user around the world, by binding to different viewpoints. Another effect of binding to a viewpoint is to set the user's position in the transformation hierarchy. So, if you are moving an object around the world, and you bind the user to a viewpoint grouped together with that object, the user will be animated at the same time as the object.
This example and it's associated code shows the uses of bindings. All the user interaction is done with bindings, no scripts or anything. The small yellow sphere is a switch that temporarily binds the user to a Viewpoint on the moving platform around the outside. The effect is only temporary as the TouchSensor generates a FALSE event on release that moves the user back to the middle. This can be fixed, but we need a bit of scripting to do it. For the meantime, hold down the button to get the effect. The three coloured switches temporarily swap Background nodes around in the same way.
Right then, that's all you need to know about binding nodes. Next time I'm going to give a brief introduction to script nodes and how to use them in your worlds. We're not going to cover writing them, just how to plug them together into your worlds.