Creating applications with Aqua language

The official way to write applications for Fluence is using Aqua programming language. Aqua compiler emits TypeScript or JavaScript which in turn can be called from a js-based environment. The compiler outputs code for the following entities:
    Exported func declarations are turned into callable async functions
    Exported service declarations are turned into functions which register callback handler in a typed manner
To learn more about Aqua see aqua book
The building block of the application are:
    Aqua code for peer-to-peer communication
    Compiler cli package for aqua to (java)typescript compilation
    Initialization of the FluencePeer
    Application specific code (java)typescript in the framework of your choice
In the next section we see it in action

Facade API

The main entry point @fluencelabs/fluence is Fluence facade. It provides easy way to start and stop the Fluence Peer. The facade API is enough for the most of the uses cases.

Fluence peer in JS

@fluencelabs/fluence package also exports the FluencePeer class. This class implements the Fluence protocol for javascript-based environments. It provides all the necessary features to communicate with Fluence network namely:
    Connectivity with one or many Fluence Node which allows sending particles to and receiving from other Peers
    The Peer Id identifying the node in the network
    Aqua VM which allows the execution of air scripts inside particles
    A set of builtin functions required by Fluence protocol
    Support for the typescript code which is generated by Aqua compiler
Even though the js-based implementation closely resembles node there are some considerable differences to the latter.
FluencePeer does not host services composed of wasm modules. Instead it allows to register service call handlers directly in javascript. The Aqua language compiler creates a typed helpers for that task.
Due to the limitations of browser-based environment FluencePeer cannot be discovered by it's Peer Id on it's own. To overcome this FluencePeer must use an existing node which will act as a relay. When a peer is connected through a relay it is considered to be client. The FluencePeer routes all it's particle through it's relay thus taking advantage of the peer discovery implemented on the node. A particle sent to the connected client must be routed through it's relay.
The js-based peer does not implement the full set of builtin functions due the limitations described previously. E.g there is no built-ins implementation for kad or srv services. However op service is fully implemented. For the full descriptions of implemented built-ins refer to Api reference
In contrast with the node implementation FluencePeer can initiate new particles execution. Aqua compiler generates executable functions from func definitions in aqua code.
Last modified 25d ago