Android development in 2019, my journey of change

[[264555]]

Mobile development has not cooled down, nor is it as tragic as the "romance of the end" that many people have predicted. It is just like the Internet industry, where the bubble has gradually dissipated and it has entered the rational second half.

Here I would like to share with readers how we can find our own technical direction and cultivate our differentiated personal competitiveness in the second half of mobile technology.

What I want to point out here is that many people are panicking about the wave of layoffs and worrying about being laid off, but in fact they are overly anxious. There are still many good companies recruiting in the market, and the salary they offer can basically reach 20K to 50K, with an average of 14 to 16 months of employment.

Therefore, I hope that at least my readers can look at the current development of the industry rationally. On the one hand, do not ignore the technological changes that are happening in the industry, but on the other hand, do not blindly follow the trend and criticize the market. Otherwise, you will be like an unscrupulous retail investor in A-shares who blindly follows the big Vs to buy and sell frequently. In the end, you will lose interest or even the principal.

There is a saying that goes: there are always a large number of people who are anxious about following the trend, while only a small number of people are quietly awesome. We hope to be that small number of people who can look at trend changes rationally and maintain objective and calm self-growth.

So, how should we find our own direction for mobile development in 2019?

Mobile technology in 2019

When it comes to technology, many authors will talk about completely empty industry trends for the next ten years; some will simply encourage everyone to switch to technology, to front-end, back-end, or AI; some even directly encourage readers to abandon technology, switch to products, operations, or even leave the Internet industry. In my opinion, these ideas are somewhat irresponsible and do not consider everyone's actual situation.

Let me explain my views below.

1. Don’t leave the Internet industry easily

It's very simple. Most people come to work to make money. Compared with many traditional industries, the salary in the Internet field is still absolutely leading, and the ceiling is relatively much higher. At around 30 years old, you can get a P7 or P8 level in Alibaba, or work in a medium-sized enterprise as a technical manager. It's not a problem to have an annual salary of one or two million, which is enough for you to cope with various financial pressures of yourself and your family. In comparison, it is difficult for other industries to guarantee such treatment.

In addition, if you feel that you have saved enough money, you can also choose to start a business on the Internet. The disruptive nature of Internet technology determines that it has a very large room for innovation. Moreover, computer technology has continued to develop in recent years, and technologies such as the Internet of Things, 5G, AI, and big data have continued to mature and gradually landed, which provides a lot of room for imagination for friends who want to start a business.

2. Don’t leave the technical direction easily

Some readers have been working in technology for a long time and have encountered bottlenecks, so they will consider leaving technology to work on products and operations. Of course, I do not deny this choice. I support technical people to learn product thinking and user thinking. However, if you abandon technology and switch to product directly, the risk may be very high. This means that you have to abandon your years of technical accumulation and start from scratch in a new direction. If you want to switch back to technology in the future, you may be eliminated.

Of course, there are also successful bigwigs who have successfully transformed. Huang Zheng of Pinduoduo started out as a technical person, then switched to product development and finally started his own business. If you also plan to start a business in the future, you can indeed consider this route.

3. Don’t change your technical direction easily

This is what has been discussed in the field of mobile development in recent years. Some people want to switch to the big front-end, thinking that the big front-end should unify the user terminal development; some want to switch to the back-end, thinking that the back-end has high concurrency and high performance, and feels very technically deep; some want to switch to algorithms and AI, thinking that this is the trend of the future and must be grasped. This kind of thinking is what I mentioned before as the "retail investor thinking": buy whatever is popular, chase the rise and sell the fall. Friends who know about stocks know that this often does not have a good ending. What I prefer is "value investment": don't be affected by short-term fluctuations. If you are optimistic about a stock and believe in its intrinsic growth value, you must hold it for a long time, and the final return will definitely be higher than most people.

Don't think that big front-end, back-end, and algorithms are necessarily high-end. Each field serves the business, and they also have to write a lot of business code. They are not busy dealing with high concurrency, high availability and other seemingly glittering problems every day. As for whether to choose big front-end or native, it is just a form of UI display. In the end, it is all business development, which will not bring much sparkle to your resume.

As for this "stock" on the mobile side, I am still optimistic.

In other words, unless one day users no longer use mobile phones, or a terminal that is more powerful than mobile phones appears, then you will need to worry about your job. But I think it is more likely that even if the terminal is upgraded and the mobile phone is eliminated, I think the operating system running in the new terminal will still be called Android.

My personal technical plan for 2019

Below, I will talk about my personal technical plan for this year as a reference, combining some good mobile development directions in 2019.

1. Dynamic

At present, in the field of mobile development, dynamic solutions are mainly divided into three categories:

Enhanced H5 container, based on WebView, this solution is the most widely used and relatively inexpensive; coupled with domestically developed WebView kernels such as UC kernel and X5 kernel, it can achieve higher stability and lower platform differentiation; the popular WeChat/Alipay mini-programs in the past two years also currently use this solution;

Based on JsCore, JS and Native communicate to achieve dynamicization; represented by React Native and Weex. The difference between it and the WebView solution is that the former uses Native View as the rendering layer, while the latter uses WebCore to parse Html and CSS to achieve rendering. The former has better performance;

The Flutter solution abandons Js because its frequent communication with Native leads to performance degradation. Instead, it adopts Dart that supports AOT (ahead of time compilation). The latter is faster at runtime, while JS is JIT, which means it is compiled at runtime. In addition, Flutter directly uses the Skia rendering engine, skipping the frequent interactions between JS and Native View, and directly uses Dart for rendering, which has much better performance.

Dynamicization has always been a hot topic pursued by mobile applications. Whenever a new dynamic solution is launched, it will be a hot topic of discussion. This also fully demonstrates the important value of this technology to various business types, and it is worth everyone's long-term investment of time for research and study.

2. Mobile terminal architecture

I know that most students' daily development is still focused on business development, mostly developing UI pages or customizing Views, or encapsulating some third-party frameworks such as OkHttp to do some light customization. Unfortunately, such development experience is not considered a highlight, especially when you have worked for three years or even longer, these will become your shortcomings: pure business development, lack of technical depth.

Therefore, I have always encouraged readers to study mobile terminal architecture in depth, which can not only expand your knowledge, but also cultivate your technical depth and thinking density, so that you can consider a stable architecture that can run in various complex scenarios. Moreover, in terms of mobile terminal architecture, many large companies have accumulated years of experience. We can learn by studying their open source technical frameworks and systems in depth. With these studies, if you want to work in a large company in the future, it will be a natural thing.

Here are some architectural points for reference, and you can do related research and study later:

• Cross-platform gateway system: For example, Mars of WeChat is implemented in C++, which encapsulates universal technologies such as long connection, weak network optimization, encryption and signature verification, and traffic management. The generated code can be applied to multiple terminals at the same time, including mobile terminals and PC terminals.
• High-performance log collection system: It can store as many running logs as possible on disk without occupying too many CPU, IO and other resources. Even if the system crashes, it must ensure the integrity and security of the data. It must be able to compress the data or even report it in pieces, or manually retrieve it through a long connection.

Due to space limitations, I won’t list too many here. Interested readers can follow me and reply [Learning] to get advanced Android learning materials.

3. Terminal intelligence (mobile terminal + AI)

Some readers may not know much about this, but this field has gradually become popular in the past two years. The basic idea is to move the operation of AI algorithms directly from service clusters with huge computing power to mobile terminals with limited resources such as CPU, so as to achieve higher real-time performance and facilitate the implementation of algorithms in real time.

The basic process is to conduct big data training on the server side, calculate a set of executable mathematical models, compress them and send them to the mobile side; the mobile side collects the current user's data in real time, inputs it into the mathematical model for calculation, and obtains the results immediately.

The advantages of edge intelligence compared to conventional AI are:

• Real-time: No need to upload data to the backend and wait for the results to be returned. For example, the process of turning photos into paintings in the Prisma App a few years ago was long. If it can achieve end intelligence, it can quickly convert photos even without a network.
• Personalization: In recent years, recommendation algorithms have been designed to achieve high personalization. With end-to-end intelligence technology, there is no need to upload your browsing history to the backend for analysis. Instead, your preferences can be calculated locally while you browse, and the content you are interested in can be quickly sorted and selected for you.
• Data privacy: Many apps collect a large amount of user data and report it to the background for user portrait analysis. However, with end intelligence, it is entirely possible to perform real-time calculations on user data, extract features, and then report the feature data without uploading real user data.

In 2018, Alibaba established a dedicated end-to-end intelligence team dedicated to researching the business implementation of end-to-end AI. Currently, there are two major scenarios:

• Vision: This is widely used in audio and video scenarios, such as face recognition in Douyin, scanning blessings in Alipay, and Pailitao in Taobao, which are all terminal intelligent technologies that everyone has come into contact with in recent years;
• Data: This part is to collect user usage data in real time on the terminal, so that some recommendation decisions can be made immediately without uploading the data to the background for calculation, eliminating this long process. Real-time intelligent recommendations are often more effective than conventional recommendations.

4. Others

In addition to the above three points, there are some other directions for reference:

• Audio and video technology: In the past two years, TikTok and some stranger social products have focused on real-time audio and video experience, attracting a large number of young users. I believe this trend will continue for a long time. After all, audio and video can carry much more content and entertainment than text and pictures. Therefore, audio and video can be considered as a long-term technical point. Of course, the disadvantage of this over-depth technical point is that due to over-specialization, the range of choices when seeking employment is relatively narrow, and you can only choose companies in the audio and video field;
• VR & AR: With the gradual advancement of 5G, I believe that the large-scale application of VR & AR will not be far away. Interested readers can make plans in advance and study the AR-related development tools launched by Apple and Google. A group of companies will definitely grow up in the future;
• Mobile blockchain: After a few bloody dramas in the past few years, blockchain is now beginning to be implemented step by step. Blockchain is essentially an innovative technology, and its combination with mobile terminals also has great potential for imagination.

Writing, reading, investing

In 2019, in addition to the continuous exploration of technology, I will also invest my spare time in three aspects:

1. Writing

Since I started writing, I have always loved writing. The written form can last the longest, and can be continuously read by people, creating value. More importantly, keeping writing can not only allow me to reflect on the knowledge I have learned, but also deepen my understanding of something. After all, many things can only be discovered when writing. The flaws and loopholes in your thinking.

2. Read

If you want to continuously output high-quality content, you must constantly input new knowledge into yourself. This is not only a respect for the content itself and the readers, but also a requirement for self-growth.

In terms of reading, I mainly focus on technical articles and books. I will maintain a certain amount of reading every week. In the future, I will share my recommended book lists and some reading experiences with you in the official account "wingjay".

Of course, some people may ask me if I don’t have enough time. In fact, I will take two to three hours to read and write after get off work every day. It often lasts until late at night or even one or two o’clock. However, this kind of learning does not make me feel tired or anything like that. It is more of a growth and a solid feeling.

3. Investment

Investment and financial management is probably a skill that many young people lack, because they have been instilled with the idea that saving money, depositing money in the bank is safe, and buying fixed deposits is reliable. But in fact, China has launched a 4 trillion flood irrigation plan since 2008 to cope with the financial crisis, which has caused the depreciation of the human currency to be very serious. Many people complain about rising housing prices, but in fact, if you have the mindset of investment and financial management, you should have found a way to put your money into real estate to preserve its value when the 4 trillion plan was launched.

Of course, once the opportunity is missed, it is missed. Don’t think that financial investment is far away from you. When you are worried about the high housing prices, you are already deeply involved in it.

Here is a small prediction. A-shares plummeted at the end of last year, but started to rise again this year. Some people think this is just a short-term recovery. I personally think that China may usher in a real long bull market, similar to the super long bull market in the United States that started in the 1980s and continues to the present. Wait for your face to be slapped.

The best time to cultivate financial investment thinking was ten years ago, and the second best time is now. This year I will set aside some time to study economics, finance and financial management. I have always believed that finance and the Internet are a golden pair, and a person who understands both technology and finance will definitely have a broader space for development.

A person's wealth is a reflection of his or her personal cognition. In 2019, breakthroughs in technology will improve cognition.

Adapt, break through, and grow against the wind

Times are changing, and technology is constantly changing. As a technician in today's era, we must first learn to adapt to changes, rather than being eliminated by changes. We must constantly pay attention to changes and respond at the same time. We must dare to jump out of our comfort zone, break through our bottlenecks, and constantly seek new technical growth points. We must keep learning and stay calm even in the cold winter, and grow against the wind.