Google developed AngularJS in 2009 and released its first version in 2012. Since then it saw enthusiastic support and widespread adoption among both enterprises and individuals. AngularJS was originally developed for designers, not developers. While it did saw a few evolutionary improvements in its design, they were not enough to fulfill developer requirements. The later versions, Angular 2, Angular 4, and so on have been upgraded to provide an overall improvement in performance, especially in speed and dependency injection.
The new version is simply called Angular, a platform and framework that allows developers to build client applications in HTML and TypeScript. It comes with declarative templates, dependency injection, end to end tooling, and integrated best practices to solve development challenges. While the architecture of AngularJS is based on model-view-controller (MVC) design, Angular has a component-based architecture. Every Angular application consists of at least one component known as the root component. Each component is associated to a class that’s responsible for handling the business logic and a template that represents the view layer.
Cross-platform mobile development frameworks
Electron was created by Cheng Zao, a software engineer at GitHub. It was initially released in 2013 as Atom Shell and then was renamed to Electron in 2015. Electron enables web developers to use their existing knowledge and native developers to build one codebase and ship it for each platform separately. There are many popular apps that are build with Electron including Slack, Skype for Linux, Simplenote, and Visual Studio Code, among others.
Tensorflow.js consists of four layers, namely the WebGL API for GPU-supported numerical operations, the web browser for user interactions, and two APIs: Core and Layers. The low-level Core API corresponds to the former deeplearn.js library, which provides hardware-accelerated linear algebra operations and an eager API for automatic differentiation. The higher-level Layers API is used to build machine-learning models on top of Core. It also allow developers to import models previously trained in Python with Keras or TensorFlow SavedModels and use it for inference or transfer learning in the browser.
An application built with React 360 consists of two pieces, namely, your React application and runtime, which turns your components into 3D elements on the screen. This “division of roles” concept is similar to React Native. As web browsers are single-threaded, the app code is separated from the rendering code to avoid any blocking behavior in the app. By running the app code in a separate context, the rendering loop is allowed to consistently update at a high frame rate.
AR.js was developed by Jerome Etienne in 2017 with the aim of implementing augmented reality efficiently on the web. It currently gives efficiency of 60fps, which is not bad for an open source web-based solution. The library was inspired by projects like three.js, ARToolKit 5, emscripten and Chromium.
AR.js requires WebGL, a 3D graphics API for the HTML5 Canvas element, and WebRTC, a set of browser APIs and protocols that allow for real-time communications of audio, video, and data in web browsers and native apps. Leveraging features in ARToolKit and A-Frame, AR.js makes the development of AR for the web a straightforward process that can be implemented by novice coders.
The creator of Gatsby, Kyle Mathews, quit his startup job in 2017 and started focusing full-time on his side projects: Gatsby.js and Typography.js. Gatsby.js was initially released in 2015 and its first version came out in 2017. It is a modern site generator for React.js, which means everything in Gatsby is built using components. With Gatsby, you can create both dynamic and static websites/web apps ranging from simple blogs, e-commerce websites to user dashboards.
Gatsby supports many database sources such as Markdown files, a headless CMS like Contentful or WordPress, or a REST or GraphQL API, which you can consolidate via GraphQL. It also makes things like code splitting, image optimization, inlining critical styles, lazy-loading, and prefetching resources easier by automating them.
Next.js basically allows you to write a React app, with the SSR and things like code splitting being taken care of for you. It supports two server-side rendering modes: on demand and static export. On demand rendering means for each request, a unique page is rendered. This property is great for web apps that are highly dynamic, in which content changes often, have a login state, and similar use cases. This mode requires having a Node.js server running. While static export on other hand renders all pages to .html files up-front and serves them using any file server. This mode does not require a Node.js server running and the HTML can run anywhere.
Nuxt.js was originally created by the Chopin brothers, Alexandre and Sébastien Chopin and released in 2016. In January 2018, it was updated to a production-ready 1.0 version and is backed by an active and well-supported community. It is a higher-level framework inspired by Next.js, which builds on top of the Vue.js ecosystem and simplifies the development of universal or single page Vue.js applications.
Under the hood, Nuxt.js uses webpack with vue-loader and babel-loader to bundle, code-split and minify your code. One of the perks of using Nuxt,js is that it provides a nuxt generate command, which generates a completely static version of your Vue application using the same codebase. In addition to that, it provides features for the development between the client side and the server side such as Asynchronous Data, Middleware, Layouts, etc.
Under the hood, NestJS uses Express, and is also compatible with a wide range of other libraries, for example, Fastify. For most of its abstractions, it uses classes and leverages the benefits of decorators and metadata reflection that classes and TypeScript bring. It comes with concepts like guards, pipes, and interceptors, and built-in support for other transports like WebSockets and gRPC.