# Brief: Dev Outreach in China

The purpose of this brief is to identify opportunities and challenges in meeting the needs of web designers and developers in mainland China.

This is a report of preliminary findings based on the findings of our Mozilla China team, and a few interviews I did with web industry people based in China.

Dietrich Ayala, August 22 2018

## Findings

* MDN Usage: MDN is not the top choice of front-end practitioners for multiple reasons, including translation availability, trust in translation quality, and search engine indexing.
* MDN alternatives: Chinese front-end practitioners need high-quality reference materials in their language. They primarily do this by searching Baidu, and finding results in tech blog posts.
* Front-end Community: WeChat groups for several front-end groups exist, and are popular.
* Deployment: Mobile web development is declining. WeChat Mini Programs are increasingly the primary or even sole focus of front-end development.
* Nobody we've talked to says there's been a comprehensive study of developer communities in China, so no valid estimates of size or composition. Anecdotes of 1m Wechat mini program developers.

## Recommendations

* Start a research effort to understand Wechat mini programs, in order to decide if and how we should engage that ecosystem and community.
* Invest in a comprehensive effort to localize MDN in Chinese, in order to become the top choice for web developers and designers in China..
* Invest in a developer advocate presence in China, in order to build the foundation of long term strategy and presence.
* Identify and interview the popular tech bloggers that show up in Baidu searches, in order to understand communications in that community and build relationships with influencers.

## Inputs

* Input from Mozilla China team at the SF all-hands.
* Interview with Yubo 裕波 of w3ctech.com, popular online forum for front-end discussion, and run series of events.
* Interview with lead JS developer for Le Wagon, a global bootcamp based in France, with locations around the world, who've found great success in China in their three locations there.

## Note: Wechat

* In conversations, it's not clear anyone at Mozilla has expertise in this, let alone basic understanding.
* Wechat centralization increasing
* Official IDE
* How would MDN serve the needs of those developers?
* Mini programs use subset of html, all of css, and subset of JS
* Can use frameworks, Vue is popular

## Inteview: Thibault and Forest, Le Wagon

* devs not using mdn
* not in chinese
* not using google
* baidu doesn't show mdn on first page
* good devs use google
* avg devs use baidu
* baidu shows tech blogs
* easy to test on baidu, search any css or whatever
* w3schools, csdn, oschina, tech blogs
* part time teacher at le wagon
* consulting/freelancer
* web dev, mini programs
* thibault: adoption of mobile - traffic moved to super apps
* forest:
* wechat mini program very huge
* closed system
* you'd need to work with tencent if you want document it
* css is same as web
* markup language is different, simpler - like xml
* strict because compiled to html
* js - own runtime
* own framework, has IDE
* based nw.js
* tell me about students needs
* well every company, not just students
* always choose mini programs
* we recommend
* like native app
* one client: very social mini program, requiring very long review
* no incentive to diy web because no audience
* mobile web not popular
* webkit based webview for links inside wechat
* no frameworks - no jquery inside mini program
* one framework provided by tencent
* but can use 3rd party js libraries, but not much used
* only for people already web devs
* students do not
* program is 9 weeks, only teach mvp
* no react in mini program
* don't teach perf
* don't care about responsive, because RPX
* special pixel system, 750rpx
* sometimes have to fetch screen height
* can use canvas, but then have to do pixels like normal
* Re: events
* RubyConf and GopherCon in China are popular
* Vue is popular, not sure if there is a conf or not

## Interview: Yubo

* founder of https://w3ctech.com and https://fequan.com
* counting the most popular front-end tech communities among China
* runs f-e events
* talked with friends and other organizations
* 1) there is no research on the total number of developers
* 2) there are thousands of front-end engineers for companies like Alibaba
* 3) we estimate that there are more than 100,000 front-end engineers in China
* Use # of Firebug installs in China as a reference for front-end developer population size
* Most developers now use the Chrome Developer Tools, and I think you can view the data from previous firebugs for analysis.
* Since firebug has stopped updating now, and Firefox already has its own developer tools, there are definitely fewer firebugs now.
* The total number of front-end engineers in several large companies should exceed 10,000, such as Alibaba, Tencent and Baidu. Tencent has thousands of front-end engineers.

## Etherpad from SF2018 All-Hands Session

From: https://public.etherpad-mozilla.org/p/dev-outreach-in-china

Captured here because etherpad is being turned off at some point
Dev Outreach in China

The aim of this session is to get a better idea about the Chinese web development market and Devrel activities that might be useful for us to undertake.

### Hector Zhao to do a short presentation about Chinese community/browsers
* Slides: https://docs.google.com/presentation/d/1ROX3C8t5Ac997-2rO4jTmDtjPMIcYpBXj-Ypf91NUKI/edit

* Chromium/Chrome seems to dominate, b/c some of the other browsers don't identify via user agent strings

* top desktop: QQ (tencent), others, browsers are dual - you can switch btwn Blink and ? because older technologies don't work in more modern engines.

* Chinese devs are MISSING Firebug, heavy usage till last possible minute - "if i need to learn new tool i'll switch to Chrome"

* top mobile: UC, QQ , 360, baidu, sogou

* global resources : github, stack overflow.

* local resources: CSDN, oschina.net, segmentfault (localization happens, quality/recency??)

* systematic learning: imooc.com,

* Great Firewall: YT and most of google, Facebook, Twitter, Wikipedia (zh-CN)

* affected: AWS, Edgecast CDN, Github (lockbox won't work)

* NPM @Taobao (hosts mirror server of npm, of python)

* circumventions happen - most savvy devs know workarounds for GFW (great firewall)

* Google Devs for China: https://developers.google.com/china/

* MDN in Chinese - english search results, concern about quality bar on MDN

* complaints about local site search on MDN, they prefer searching google for MDN, using KW search

* WeChat Mini Programs: basically a packaged web app with 170mm daily users/1 mil devs

* - predefined components, web like tech

* .json, wxml/.wxss/.js

### Dietrich to share his insights from working in the Chinese community

* Slides: https://autonome.github.io/elephant/public/moz-china-sf.html

* Shenzen to Hong Kong - 1 hour distance, but worlds apart.

* WebConf Asia, Charis Rooda - https://twitter.com/charis https://www.webconf.asia/

* hong kong: Finance/Banking and Advertising

* hong kong: web dev communities - devs have no choice in tooling, front end design looks fresh/new, but down the stack is older tooling like jquery.

* must support multiple browsers and browser like UI

* assumption - most users are mobile, not necessarily desktop, embedded browser is more likely than not (we seriously need to improve the mobile experience of MDN, and take the time to integrate service workers to enable offline ability)

* super-excited/ super-sad

* Open Source Conf in HK: totally isolated from web community. (4-5 yr old event) - compare to Fossasia or OSCON. most attendees were local.

* Shezhen - 1st economic zone in China - 12 mm, more like 15-25 mm - 90% of electronics in the world come from here.

* 100s and 100s of meetup groups: AI, cloud, python (huge), machine learning, almost no web stuff.

* kickstarter changed everything for Shenzhen. $100K from KS isn't really that much, but now small orders are possible

* SZOIL - shenzhen open innovation Lab; small incubator lab, non-profit. Part of the Fab Lab family. https://www.szoil.org/

* Lots of labs/spaces available to allow people to come in and build their own hardware. The spaces could do anything from you, from planning to prototyping to manufacture and distribuition. they take a percentage cut of your project, depending on how much they do for you

* Easy to create anything, as long as it is similar to something that had gone before

* Lots of new low cost chips coming out of Shenzhen, e.g. to power low cost Google Home/Alexa knock-offs

### Q&A

* How many web devs are there in China? And what percentage speak English?
-- We don't really know; Dietrich having useful meetings soon.
-- some big meetups can have up to 1500 devs there

* How do people make websites? With the IE/Chrome swtching browsers, do people target IE, or Chrome, or just making legacy sites?
--No, they are not really targetting old browsers? They should make sites that work in modern browsers as well as IE. IE cannot be ignored, as so many people still use it.

* Is there anything that is similart to MDN available in China?
--Not really. Not as complete

* Is there a lot of framework usage in China? E.g. Vue/React?
Vue has great Chinese documentation - it's like

* Might be an idea to sert up a separate service in China for MDN? E.g. developer.mozilla.cn?
-- This would be difficult if the site is not registered in China. Cannnot use things like CDNs outside China. If you don't have a Chinese license, then your service will be subject to restrictions, or maybe even blocked altogether.
-- There are equivalents to things like AWS in China

* What does China need from us? How could we help them? What do we need to do for Mozilla to be in China?
-- it is a big uphill battle. Both in terms of MDN, and in terms of translation strings in the product. E.g. devtools person tweeted screenshots of devtools in China, asking for feedback.

* Chinese localization outdated, not as good or complete as English version
-- Yup, main things are that pages don't look the same as the English content, and people who read Chinese&English say the Chinese version is just wrong.
-- English versions updated every day, Chinese version not updated so often. REsults in Chinese version being bad.
-- people often look at translated versions. Often in Chrome, as it has built-in translator service.
-- we have very small amount of community memebers doing Fx l10n into Chinese.

* How did Dietrich find the labs and other people he talked to in China?

* How many Mozilla employees in China?
-- 25
-- they do everything

* In china, most usage of the web happens on Mobile. Does most testing happen on mobile?
-- not too sure, but often people just test against Chrome, as most browsers aere based on it.
-- mobile browsers, WeChat app very important. Doesn't use web standards exactly, although similar.

* PWAs in China?

### Other potential agenda items:
* Peiying to share her insights as a user visiting China
* Chinese l10n community

### Background knowledge

From Mary Meeker's annual 2018 trends report just released:

https://www.recode.net/2018/5/30/17385116/mary-meeker-slides-internet-trends-code-conference-2018

China report from new research : Slide 237-261

Also this in topline findings:

China is catching up as a hub to the world’s biggest internet companies. Currently, China is home to nine of the world’s 20 biggest internet companies by market cap while the U.S. has 11. Five years ago, China had two and the U.S. had nine.

### Rough notes

From Florian:

In particular, we would like to discuss or learn about these things:

China is growing a lot. Google seems to start looking into this market, but doesn't seem relevant there (yet). I don't have insights, who at Mozilla has? Is this/these person(s) at the SF all hands?

What is blocked in China? How can Dev Outreach still reach out?

I run the browser-compat-data project and in China there are mobile browsers called UC and QQ with a good deal of marketshare. I would like to understand how web developers work with and develop for these browsers. And I would like to have contact to these companies so they can add their compat data (I don't speak Chinese, though).

I want to understand how good the Chinese translations of MDN docs are. Chinese traffic has been growing a lot on MDN, but we aren't investing anything in our Chinese content or l10n infrastructure. Are we missing out a giant opportunity?

From Dietrich:

Last year I went to Shanghai, Chongqing, Shenzhen, Hong Kong and Beijing, and my personal experiences showed usage in mobile that is less about "browser" than it is about the web as an interstitial/transient part of tasks done inside messenger apps.
The apps you listed are somewhere between messenger apps and "lifestyle operating system", and pretty far from what we generally consider a "browser".

### Questions

From Sheppy (feel free to eliminate any duplicates as things evolve here):

What can we do to help foster a localization community for MDN content in China?

Is our content quality localized for China making sufficient progress, or do we need to consider supplementing with contracted or otherwise paid resources to improve/accelerate the work?

What kinds of content matter to Chinese developers? Do they need tutorial and how-to content localized more than reference material, vice-versa, or do they need both equally?

We have an inherent time zone problem when trying to work with localization community in China, and the same problem when trying to interact with developers who might want to be involved in MDN's mission, or to get help from the MDN team with some issue. What ideas can we find to help mitigate this?

We've made moves toward using Discourse for communication rather than mailing lists and other email-based solutions, which has worked well, but it doesn't resolve our issues with language barriers. Can we find ways to help make it easier to translate Discourse conversations to help improve communication across locale barriers?

What if any differences are there in terms of Web technologies that are of interest to web developers in China as opposed to the US or European markets? Are there things in need of attention that we are not focusing on because they're relatively unimportant here that are substantially more important to Chinese users?