Dandelion · JELLY Technology Weekly Vol.29

Front-end intelligence refers to the expansion of the front-end with the help of AI and machine learning capabilities, so that it has some characteristics beyond the current front-end capabilities, which will be an important change in the future front-end direction. At present, various Internet manufacturers have their own projects towards this general direction of continuous practice, to challenge the future, and Ali is one of them, but also left a lot of impressive products…

High overlook

The height of the sky, feel the universe is infinite

The front frame

A new mechanism for parsing JSX

React 17 introduced the new JSX syntax conversion mechanism, which not only eliminates the need to import React objects in advance when writing JSX, but also provides some performance optimizations and conceptual simplification. For example, if the key is passed independently of other props, it is not recommended to use defaultProps in functional components. This paper introduces the related background and general modification content, as well as upgrading and compatible methods. For details on the motivation and implementation of this change, see this RFC.

The design philosophy

Front end intelligence in Ali those things

The direction of front-end intelligence is proposed to bring about changes in front-end technology and expand the front-end with the help of AI and machine learning. What are the problems encountered in the process of promoting this goal? Ali gave a summary and reflection on the direction of front-end intelligence in 2020.

Graphics programming

A high quality interactive experience website “Radar”

This site will randomly jump to some excellent websites based on CSS3, Canvas and WebGL technology, with amazing interface presentation and interactive experience, you can experience the powerful ability of Web rendering technology in your busy work

Artificial intelligence (ai)

AI expert knowledge map

What is artificial intelligence? What subdivision modules are there? Here you can find all knowledge points related to artificial intelligence, and the contents listed in each knowledge point will be connected to corresponding materials, such as Wikipedia. Whenever there is a new research in the field of artificial intelligence, the corresponding module will also be updated, hoping to give students who are interested in artificial intelligence a direction. Also to have a certain accumulation of students to provide some help.

Application of machine learning in static code analysis

Nowadays, machine learning has been deeply applied in various fields of life. This paper raises an interesting point: can machine learning help software developers dig bugs in programs? The answer is yes. The authors start by introducing the industry’s known machine learn-based static profilers, such as DeepCode, Infer (from Facebook), SapFix, Embold, CodeGuru (from Amazon), and more. The author then illustrates the difficulties and limitations of machine learning in the field of code analysis with the example of implementing a code analyzer that uses machine learning techniques to find code defects. Reading the whole paper, the biggest benefit is that the author combines the machine learning technology with the analysis of application scenarios, which is conducive to enrich our imagination in ML technology.

Tool to promote

AntV: Visual data solution

AntV is ant Financial’s data visualization solution, among which G2, The most frequently mentioned, is a graphics-based visualization engine, which adopts The development concept Of The Grammar Of Graphics, is data-driven and supports highly interactive solutions. The latest version 4.0 uses TS to rewrite. Compared with eCharts, which is easy to use (the bottom layer relies on the lightweight vector graphics library ZRender), the development document is not perfect, so it has a higher threshold for data visualization, but has greater advantages in expansion and creativity. Currently AntV has a series of products including G6 (relational data visualization engine, supporting interaction, analysis, animation, etc.), F2 (mobile solutions, supporting H5), L7 (geospatial data solutions). Extensions include G2Plot, Graphin, and ChartCube.

The sea gleanings

The sea, product kuibu and thousands of miles

Machine learning 30 minutes introductory point north

Intelligent front-end is the direction of the future and want to keep up a quick introduction to machine learning, maybe you still feel no suitable work in the project practice, or feel for academic, professional and so on factors did not have a chance to find a job as a machine learning, so quick to follow the teacher mark and see it together, 30 minutes to show you understand the basic knowledge of machine learning, Break the stereotype of AI, and take a look at some practices and projects of machine learning in Bump Lab.

Kinetic effects are actually quite simple

Front-end intelligence ultimately hopes to liberate engineers from “meaningless low-level labor,” but it is not the only way, and there are many solutions in many areas of development that can significantly reduce r&d costs. For example, in the process of dynamic effect development, we often think that designers have sorted out the whole dynamic effect logic, but debugging is still time-consuming and laborious. Why can’t we directly reuse the designer’s dynamic effect logic? Which do you prefer, machine or manual parameter retrieval?

“Dandelion” magazine, updated weekly, we focus on mining “basic technology, engineering, cross-end framework technology, graphics programming, server development, desktop development, artificial intelligence, design philosophy, front-end framework” and other general direction of the industry hot issues, and professional interpretation; Not only that, we will also recommend the selection of concave and convex technology articles, to show you the research direction of the team.

Look up, dandelion seeds will take root and germinate, such as summer flowers gorgeous; Ge Ge Zhi zhi, we climb to overlook, the sea gleaned, in order to product silicon step to thousands of miles.

Dandelion JELLY Technology Weekly contribution Guide