内容简介:Today, Microfrontends are no longer a proof of concept. If we search on the internet, we can find many case studies of adopting Microfrontends. Besides, most of these web apps are already in production, proving its robustness.However, it is essential to fo
Today, Microfrontends are no longer a proof of concept. If we search on the internet, we can find many case studies of adopting Microfrontends. Besides, most of these web apps are already in production, proving its robustness.
However, it is essential to follow the best practices and choose the right tools to succeed with Microfrontends. Otherwise, Microfrontends could quickly become an overhead for your teams, risking the overall growth of the project.
In this article, I’m going to share some of the best practices around Microfrontends and recommend several tools to establish these practices in your projects.
1. Code Structure and Components
When adopting Microfrontends, one of the critical decisions you have to make is to choose the approach used to structure the code.
The two conventional methods used out there are the Distributedrepo and the Monorepo. Other than these two are hybrid approaches that fall somewhere in between.
In the Distributedrepo approach, the Microfrontends are divided into multiple repositories with their lifecycle managed separately. In the Monorepo, all the Microfrontends will reside in a single repository.
Distributed Repo Approach
The Distributed Repo approach is the more flexible one but it has the great challenge of sharing UI components, to maintain a consistent UI across MFs (which usually means maintaining UI components as NPM libraries, which creates the overhead of maintaining different build pipelines, different repositories, and version mismatches).
One good example of both implementing this Micro Frontends approach and solving the problem of sharing UI components between them, can be seen in Bit.dev .
The Bit.dev marketing website is composed using two groups of React components published and managed by the Bit platform. The two groups were built and delivered separately. The “moment of integration” happens on build time, in a codebase that consumes components from both collections. Whenever a new component version is delivered, a new integration happens.
Hover over different components on Bit’s landing page to see each component’s “scope” or “collection”. Click on the component name (on top) to inspect the component and/or to install it in your project.
As mentioned earlier, the above page is built from components developed in two different codebases, on two different GitHub repositories. Components of each codebase are published to their respective Bit collections.
- base-ui: Github | Collection on Bit
- evangelist: Github | Collection on Bit
以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持 码农网
猜你喜欢:本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们。
基于内容图像检索技术
周明全 / 清华大学 / 2007-12 / 28.00元
《基于内容图像检索技术》从理论方法研究与实现技术角度,总结归纳了基于内容图像检索(CBIR)技术的研究与进展,并融入了作者多年来的相关研究与应用成果,系统地介绍了CBIR的主要概念、基本原理、典型方法、实用范例以及新动向。《基于内容图像检索技术》共有12章分为五部分:第一部分是概述,分析了CBIR的体系结构、技术现状和发展趋势;第一部分讨论图像特征提取,给出图像低层特征(颜色、形状、纹理、空间关系......一起来看看 《基于内容图像检索技术》 这本书的介绍吧!