I was under the impression that an endpoint was defined in a config file as the list of possible clients but that makes no sense (in the sense that I assumed it said what computers could connet to the service) now I'm gathering that it's more of a definition, so would someone please explain what an end point is to me? I understand the concept of definining the contract interface and then implementing the contract but I get lost somewhere between there and actually having something useable.
What is an address in this context? the host address?
A binding is the communications method/protocol to use correct?
the contract is the "object being shared" essentially (yes i know that's so technically incorrect but work with me here)
Answer
An endpoint is what a service exposes, and in WCF terms, is made up of three things:
- Address
- Binding
- Contract
Address is the URL by which the endpoint can be reached.
Binding dictates transformations that are applied as well as the shape (to some degree) of the messages sent to the implementation of the Contract at the Address.
Contract dictates what operations are being exposed at the address. It's exactly what it says it is, it's a contract to indicate what calls are permissible.
Most of the time, people remember it as A B C.
Some things to note:
The binding is typically going to be a combination of channels with behaviors applied; channels being elements on the channel stack which modify the message and perform actions before they get to the service implementation.
While commonly represented by an interface in .NET, it is not a requirement that a Contract be represented in this manner. Some design-first advocates will define the schemas to the messages that are going to be sent for the request and the response first, which is what WCF transforms the .NET Contract interface into.