Internet Hyper Application Introduction

What is HyperApp?

A HyperApp is a Web application bundle that contains everything it needs to run locally inside a virtual machine instead of on a remote server:

HyperApps are built as regular Web applications with Node.js, AJAX, PHP, ASP.NET, or servlets, etc.

Why HyperApps?

A HyperApp can serve static and dynamic content locally while communicating with a remote web server if needed. Thus, HyperApps can dramatically improve performance and the user experience while leveraging underlying hardware capabilities without compromising security in comparison with desktop/native applications. Note, HyperApps preserve the same-origin policy and can only connect to their origin web servers.

In essence, HyperApps are lightweight client-side proxy web servers that unify desktop/native and web applications. They are built on top of virtualization, thus isolated from each other and the host, and provide more security to end users and businesses.

Since HyperApps work locally, they can be viewed as MVC applications which allow leveraging the declarative nature of HTML to build View and delegating Model and Controller to the local virtual machine instead of pushing a large amount of JavaScript into a browser.

HyperApp states can be easily backed up to a cloud allowing HyperApps sharing between computers: office, home, laptop, tablet, etc. as well as restoring them in case of a hardware failure.

HyperApps are easy to clone, deploy, and test, thus they enhance application development.

How it works?

A user enters a URL in a browser, example: https://someserver/someapp.html
The browser loads the page and its script checks if its HyperApp exists on user’s machine

If so, it redirects to happ://someserver/someapp.html
Otherwise it checks if the HyperApp plugin is installed
- If so, it redirects to happ://someserver/someap.html
- Otherwise it asks the user to install the plugin

The browser passes URLs with happ scheme/protocol to the plugin
The plugin checks if a corresponding HyperApp is running
If not, it checks if a corresponding HyperApp image exists 1. If not, it downloads the bundle from a corresponding remote web server 2. The plugin creates the HyperApp image and a virtual machine
The plugin starts a virtual machine with the HyperApp image
The plugin channels HTTP requests to a corresponding HyperApp.

Offline mode:

A user enters happ://someserver/someapp.html in a browser
The browser passes URLs with happ scheme/protocol to the plugin
The plugin checks if a corresponding HyperApp is running

If not, it starts a corresponding HyperApp

The plugin channels HTTP requests to a corresponding HyperApp.

Get Started on Linux

This repository is a proof of concept example. The full implementation is ongoing and coming soon. Stay tuned.

Prerequisites: Firefox browser

Install Docker
Add yourself to the Docker group

$ sudo usermod -aG docker <YourLoginName>

In this prototype, we use the official Node.js Docker image. It is quite big, so we need to pre-load it.

docker pull node:0.12

Download and install the happ.xpi Firefox addon

Select the "Open" item from Firefox's "File" menu, it will bring up a file selection dialog. Navigate to the "happ.xpi" file, open it and follow the prompts to install the add-on.

Try the following in Firefox

happ://myapp1:8081/index.html

Here is the output:

Hello HyperApp

What just happened?

In this prototype, we access the HyperApp directly by using the happ scheme. So, the Firefox addon have intercepted the entered url, started a container for myapp1 with mapped port 8081, and channeled myapp1 trafic to the container myapp1. The web application is located at ~/.mozilla/firefox/<???>.default/extensions/{3D4F08B3-FA9A-44D1-BE2A-A5EFF83FC0B4}/resources/happ/data

Try also the following:

happ://myapp1:8081/app.jsx