AWorks should have these philosophical ideas

First, the difference in thinking

The key to Apple becoming the most profitable company in the world is that the product's performance exceeds the user's expectations, and because of a large number of reusable core areas of knowledge, the overall cost has been maximized. Yourdon and Constantine used economics as the bottom driver of software design in the book "Structural Design," and software design should aim to reduce overall costs. People find that the maintenance cost of software is much higher than its initial cost, because understanding the existing code takes time and is prone to errors. After the changes are made, testing and deployment are also required. Due to the lack of scientific software engineering methods, not only is the software difficult to reuse, but also it is difficult to expand and maintain, leading to high development costs.

More often than not, programmers are not coding but are reading programs. Because the reading program needs to be understood in detail and in concept, the investment in modifying the program will be much greater than the initial programming investment. Based on this consensus, the series of issues that concern us need to be constantly thought and summarized so that it can be reused. This is the origin of methodology.

Through the analysis of financial data, due to early mistakes in decision-making and lack of scientific software engineering methods, we have developed products with long cycle times, high technical difficulty, and extremely low rates of return. Not only is the software difficult to reuse, but it is also difficult to expand and maintain. The cost of development remains high.

It can be seen from the perspective of software development that software engineering and computer science are completely different areas of knowledge. The main difference lies in people, because software development is a human-centered process. If you consider the human factor, software engineering is closer to economics than computer science. Obviously, if we do not change the way of thinking, it is difficult to develop products that are both easy to sell and low in cost.

Second, the profit model

The BOM cost of the product is very low, and the gross profit is very high. Why is the profit of many listed companies unable to afford a suite? Who bought the house in the end, this issue is worth our reflection!

In addition to the vision, mission, and values, the great company’s core indicator is profit. One of the greatest pains for developers is the difficulty in developing products that can be sold accurately. Because many companies generally do not know how profits come from? So it is necessary to establish a profit model, namely "profit = demand - design." Demand is to solve the problem of how to sell products. Design is to solve the problem of how to reduce costs.

The quality of the code not only directly determines the quality of the software, but also directly affects the cost of the software. Software costs are composed of development costs and maintenance costs, while maintenance costs are much higher than development costs. The phenomenon of brute-force development abounds. Many hard-won funds are engulfed silently, and the entire society’s resources are wasted. .

Why not abstract the complex technology highly? AWorks is born in this context if it can be achieved by professional people doing professional things. Because of the integration of more software engineering techniques, it is possible to completely release the programmer from the non-core domains and focus on the core competitiveness.

Third, the core domain and non-core domain

In fact, a software system encapsulates knowledge in several fields. One of the domain knowledge represents the core competition of the system.

Force, this area is called "core domain", and other areas are called "non-core domains." Although the more popular terms are "business" and "technology", the use of "core domains" and "non-core domains" is more rigorous.

Non-core domains are the areas of others, such as the underlying drivers, operating systems, and components, even if you have some advantages.

It is also temporary and competitors can also obtain it through other channels. Although the improvement of non-core domains is necessary, it is still insufficient to dig deeper in the core domain so that competitors cannot easily obtain them from third parties. Because in-depth mining in the core domain achieves reuse based on the core domain, it is the fundamental means to obtain and maintain competitiveness.

To achieve reuse based on core domains, it is necessary to consider core domains and non-core domains separately. Because of the premature introduction of each collar

The mixing of knowledge in the domain will increase the unnecessary burden, which will lead developers to make no effort to think about the deeper problems in the core domain. As the scale of the problem to be solved becomes larger and the human brain has limited capacity and computing power, it is necessary to divide and conquer, because the knowledge of the core domain and the non-core domain is independent.

IV. Commonness and Differences

If there is no ARM company's IP authorization mode, it will certainly consume a lot of hard-won wealth when designing MCU. Although ARM's scale is relatively modest, it does not affect ARM as a great company. Its contribution to humanity is obvious to all.

In spite of this, without hardware support, hardware is a waste of iron. Due to the wide variety of needs, despite the tremendous efforts made by people to minimize the cost of development, it seems that people are looking forward to the realization of a beautiful dream that will be inevitably out of reach and that they will not be able to highly reuse the knowledge that humanity has accumulated through hard work. Driven by commercial interests, great companies are not great, attempting to tie their customers to their chariots, leaving their competitors desperate. Most successful companies are.

Is there a way to crack it? That is an abstract tool for "analysis of commonalities and differences." In fact, no matter what kind of core MCUs are used, and regardless of which company's OS, the design principle is the same, but the implementation method and entity (hardware and program) are not the same, but as long as the commonality is abstracted as a unified interface, The difference can be handled with a special interface.

Based on this, we may wish to make a bold assumption. Although the PCF85063, RX8025T, and DS1302 are from different semiconductor companies, the commonality is RTC real-time calendar clock chips, which can share the same drive interface with a high degree of abstraction. The differences are handled by special drive interfaces. Although FreeRTOS or μC/OS-II or sysBIOS, Linux, and Windows are different, they are all OSs. Multithreading, semaphores, messages, mailboxes, queues, and so on are their unique commonalities. Obviously, QT and emWin can be highly abstract as well. For the GUI framework. In other words, no matter what MCU, regardless of whether or not to use the operating system, as long as the corresponding header file is modified, the application code can be reused.

This shows that regardless of the choice of MCU and OS, as long as AWorks supports it, you can achieve cross-platform operation on the target board. Because no matter what kind of OS, it is just a component of AWorks. For different OS, AWorks will provide the corresponding adapter, then all the components can be interchanged as needed.

Because AWorks has developed a uniform interface specification and has highly abstracted the built-in functional components and peripheral devices of various MCUs, no matter whether you choose ARM or DSP, with highly reusable software design principles and only for interface programming. The premise of the idea, the application software can achieve "a programming, lifelong use, cross-platform", apparently AWorks gives you the greatest value is not to re-invent the wheel.

V. Ecosystem

If only the OS and the application software framework constitute the ecosystem, this is not enough. In the era of IoE, a complete IoT system also includes sensors, signal conditioning circuits, algorithms, and access to cloud technologies. It can be said to be extremely complex and all-encompassing. This is not something that a company can achieve within a few months. , it takes a long time to accumulate.

ZLG made a long-term layout at the beginning of its establishment. We did not position ourselves as a chip agency or design, nor positioned ourselves as an instrument manufacturer, nor did we position ourselves as a solution provider, but over time and The development of the times, after painstaking efforts, has naturally become the leading brand of "Industrial Internet Ecosystem". This is not deliberate, but it is the result of long-term struggle.

ZLG designed high value-added modules, boards and high-end measuring instruments through "chip + AWorks", and accessed ZWS (ZLG Web Services) IoT cloud processing system through wired and wireless communication interfaces to realize big data processing and constitute the industrial Internet. ecosystem.

Its business model can not only sell hardware, but also can sell the platform. It can also provide system services to end users for a targeted industry. At the same time, ZLG will establish a joint industrial Internet ecosystem laboratory at 50 universities across the country. Through the model of industry, research, and development, it will cultivate talents to serve the industrial sector. It will also be possible to create a ZLG-based angel investment to help more people achieve greater success. Promote the rapid development of the “China Smart 2025” program.

V. Experts and Generalists

The success of any organization and system is inseparable from the cooperation and struggle of experts and generalists. It has been very difficult for us to come along for 12 years. However, it is gratifying that the AWorks ecosystem has been developed and some experts and core talents have been trained. I deeply appreciate the value of excellent people. I can't describe the value of excellent people. Therefore, my main job in the future is to find and discover outstanding talents and provide them with a worry-free platform that can calm down and innovate. To change a small part of the world, welcome self-referrals or readers to recommend me to ideal and ambitious people. If you are interested, you will find me.

Six, series introduction

This series of books is named "Embedded Software Engineering Method and Practice Series". At present, it has completed "Programming and Data Structure", "Programming for Ametal Frameworks and Interfaces" and "Programming for AWorks Frameworks and Interfaces" (I). The follow-up will also include "Programming for AWPI Frameworks and Interfaces," "Programming for AWUI Frameworks and Interfaces," and "Programming for ZWS IoT Frameworks and Interfaces."

(Author: ZLG Zhiyuan Electronics chairman ZLG)