Workflow System Design-new problems and new ideas

Workflow System-new problems and new ideas

The next-generation workflow system can form a new system only after solving the following four problems:

Dynamic Flow Data Structure large-scale flow topology Data Structure new flow automatic operation Controller Embedded formula and asynchronous Dynamic Data Driven

Every new design is very difficult and requires a relatively high level of mathematics. Which design is the most critical part? What are the commonalities behind these designs? What types of mathematical tools and theoretical knowledge are required for these designs?

Ah .. It's really complicated .... I think it is complicated to think about it. However, to do this kind of thing, you cannot worry about it. You need to stabilize it step by step. This is not a small project for the boss. You can collect the money in three months.

There are still many problems with the old process management system. First, the underlying data structure is very primitive-the linked list structure, which cannot process the dynamic process topology, in addition, if the linked list structure is used to identify the symmetric topology and asymmetric topology in the process topology structure, it is not as convenient as the process matrix data structure, however, the elements in the data structure of the process matrix represent the connection lines between nodes, rather than the nodes themselves, which makes it difficult to use the node status for the process operation control model .. The chain table structure and matrix structure are used to represent the process. Each has its own advantages and disadvantages. It seems that it is impossible to design a new process data structure in a single way,

That is to say, to design a new process data structure, we must consider various situations comprehensively. After fully analyzing the advantages and disadvantages of Matrices and linked lists, creative adoption of new ideas to design a bottom-layer data structure that can easily express the process and avoid the disadvantages of the linked list structure and matrix structure, such a new process data structure must emphasize that the new process data structure must be concise and clear.

If the mathematical model used is more complex, the difficulty will be greatly increased for programmers who write code later. This will bring a great obstacle to the process from design to development, if the previous theory is complicated, it will be more difficult to implement it later. After all, for the current programmer level in China, most technical staff are not particularly competent in mathematics and theory, there are many first-line programmers in high school and junior colleges, especially those with only 2-5 years of development experience. The number of first-line programmers with a bachelor's degree or above is not very high, most front-line technicians have a limited theoretical level. If the researchers propose a complex mathematical model for front-line developers to implement it, it is estimated that it will be relatively difficult. This is a national condition, the situation cannot be changed in the short term. This situation exists not only in the software industry but also in many industries and fields, that is, the relationship between front-line technicians and the following researchers is not very close, communication between technicians and researchers is far from enough. Even with high-level research results, it is not necessarily possible, and first-line technical staff must move forward to the high-end field, without the support of subsequent researchers, it is impossible to move from low-end to high-end by its own strength. This is a problem that is hard to overcome in the short term, therefore, when designing a new process data structure, we cannot simply consider the advanced nature of theory, it is also necessary to consider the degree of difficulty in actual development and the ability of developers to digest and absorb new theories and ideas, as well as to implement these new ideas as the basic basics of specific coding.

Therefore, when designing a software system, we must consider the front-line developers' ability to understand and accept the design and their ability to implement the design. Otherwise, the new data structure and model designed by the designer may encounter technical difficulties during coding and cannot be broken through for a long time, this makes our boss very anxious... If a country can print money infinitely and use natural resources infinitely, all economic problems will not be a problem. Unfortunately, no country can print money infinitely and exploit natural resources, this joke means that theoretically, we can design an infinitely complex system, but not necessarily implement it... Because the capabilities of human and human-created tools are always very limited.