What is the use case for the system to monitor the MQ queue to automatically send mail

Boiled water <czmo***q.com> 2016/7/11 10:27:33
How to draw the use cases that are automatically executed in UML, for example, the system automatically sends emails
Pan Jiayu (3504847) 17:15:07
"System sends email" Yes Step, see which step is the step in the use case. Try not to use words such as "auto" and "manual".
Boiled water <czmo***q.com> 22:38:38 The
system automatically sends emails, that's it. Our system uses MQ for interaction. The system monitors the messages in the MQ queue. If a message comes in, the relevant personnel will be automatically notified for processing. In addition, Mr. Pan, I would like to ask how to distinguish between them or the steps in the use case? For example, when the system sends an email, it can be understood that the executor is the system. It can be executed by the human flesh system, but it was automatically executed by the system
. Pan Jiayu (3504847) 22:51:56
According to your description, the use case should be: time→monitor MQ Queue message. Sending mail is the step inside.
The few words you said later indicate that you need to carefully read the "Software Method" and review it .
Boiled water <czmo***q.com> 23:00:29
Okay. Thank you, teacher. Let me take a closer look at
Li Junjie (705***85) 23:05:02
"The system uses MQ for interaction. The system listens to messages in the MQ queue. If a message comes in, it will automatically notify the relevant personnel for processing." If Just describe the function to be realized in this natural language, will the development not be able to see it? Why do we have to struggle with how to write use cases?
Li Junjie (705***85) 23:10:18
Shouldn’t it be explained as timer-monitor mq queue or as system-sending mail? Will developers make mistakes? If the developer and the questioning brother have similar understanding of the use case, will he understand better what he needs to achieve? In what ways do these affect work?
Boiled water <czmo***q.com> 23:12:21
In fact, what I'm talking about here is whether there is a granularity standard for the use case.
Junjie Li (705***85) 23:18:00
depends on what the use case will do. ? If you want to estimate the cost or estimate the workload, you may have to set a granularity standard. I have seen the use case granularity standard for performance appraisal and defect measurement. As a result, it is very painful to write use cases.
Li Junjie (705***85) 23:19:28
If it is only used to express needs, it can be written and communicated, and can be confirmed. Is there a standard for granularity?
Li Junjie (705***85) 23:20:08
After all, the use case is also a kind of language. Can the language have a granular standard?
Boiled water <czmo***q.com> 23:20:20
is also Pan Jiayu
(3504847) 9:01:46 The
use case is to teach developers how to choose rigorous natural language description requirements. The so-called requirements described by many developers in ordinary times, under the irradiation of use case thinking, will appear to be riddled with defects. If you don't believe it, just post a paragraph of the so-called "requirements" written by your team to see.
Like the sentence above, "The system uses MQ to interact, and the system monitors messages in the MQ queue", there is a question here. Is "MQ used to interact" a design decision from the development team or a request from stakeholders? If it is a design decision from the development team, this requirement does not exist, and the use case (the value provided by the system to the stakeholders) should be found from the perspective of the stakeholders, such as: time → check new orders and remind them, even with "timing" !
For example, "employee submits application information", it can also be described as "employee presses the enter key", it can also be described as "Keypress event occurred on Form1", it can also be described as "[timing] scan the key position matrix row by row to confirm closure Is the coordinate of the key and the scan code corresponding to the key a bit like timing **MQ? Which is better depends on the core domain of the system.
Assuming that "interaction using MQ" is indeed the functional requirement of the system, the core domain knowledge of the system is indeed this domain. "The system monitors the messages in the MQ queue. If a message comes in, it will automatically notify the relevant personnel for processing." How to trigger the monitoring? Does each monitoring need to be triggered by an external executor, or is it performed in a certain time period? What is "news coming in"? In addition, "automatic" is also a redundant expression.
Pan Jiayu (3504847) 9:02:42 Pan Jiayu (3504847) 9:03:17 Pan Jiayu (3504847) 9:04:14 Li Junjie (705***85) 9:04:20 Thank you Mr. Pan Li Junjie (705***85) ) 9:05:21 My doubts should be the general doubts of a type of "combat" developers Pan Jiayu (3504847) 9:06:23 You can see the software method and the past Q&A records. First, I can put all the questions on the software method. You got it right Pan Jiayu(3504847) 9:09:24 Pan Jiayu (3504847) 9:09:56 Pan Jiayu (3504847) 9:10:04 Jenson(48***62) 11:59:12 I have benefited a lot, thank you very much Junjie With Teacher Pan.