First, what about the three major database paradigm is that we no longer say, many online blog writing are very good, such as the following links:
https://blog.csdn.net/qq_26878363/article/details/81533273
Copyright Statement: The above link is a blogger original article, follow the CC 4.0 by-sa copyright agreement, reproduced, please attach the original source link and this statement.
Now I want to talk about the focus of the first paradigm.
Although the database design paradigm is the design experience, compliance is not mandatory rules, however, and I strongly recommend, first paradigm to follow.
The first content database design paradigm, it simply is each column of the table to ensure atomicity, is the meaning of the expression to be unique, can not be broken down.
This is well understood, many people take for granted, it is estimated that no one would make a name and age, however, the law of the first design paradigm there may really exist, such as the design of a field in the table storing message content, this violation the first paradigm, because the meaning of the expression of the mail data storage does not have unique. Why do I emphasize the first paradigm to follow it? Like this situation I have written, to a certain extent, also caused no data redundancy, data in the mail call, no further data processing, but also convenient to use.
Data redundancy, and performance management will result in the burden of data processing errors do not have a direct impact, however, violate first normal form, there is a fatal problem, that is hard to find obscure may contain errors. Would simply say that when we insert data into the database, for those fields containing multiple meanings to be treated separately, for example, to add a company name in the mail, telephone, etc., if the company name is added wrong, when we conduct test We found the problem of increase is very difficult, because this information is hidden in the voluminous message body inside.
In summary, do not violate first normal form, or strict compliance as well, can bring at least three benefits, reduce the difficulty of data processing, data hiding may reduce errors, reduce data redundancy.
