Skip to main content
Software Engineering

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

Required fields*

5
  • I'm not going to downvote, but I think this is a little bit of a strawman argument, as currently phrased. It presumes that the team writing the DB schema is better/more correct/has a better understanding of the requirements than the application developers. Commented Sep 29, 2017 at 15:58
  • 3
    The key here is that the DB schema provides a single possible point of failure. Bugs in code are effectively infinite possible points of failure since an error may exist now or may be introduced in the future. Commented Sep 29, 2017 at 16:44
  • 2
    @Paul: Not at all, I assume they are the same people, since this is typically the case with this style of architecture. The point is that even if there is an error or oversight in the schema design, you would only get data corruption if there is an exactly similar error or oversight in the application code. And even then you are not worse off than without constraints. In any case, a schema is typically a lot less complex than application logic and therefore less error prone. Commented Sep 29, 2017 at 16:46
  • @17of26 and JacquesB I agree that there needs to be some form of schema enforcement, I just disagree that it has to be in the DB in all cases. The major advantage in my experience of having it in the DB is if the DB is shared by multiple applications. But there are plenty of ways to enforce a correct schema (and in some cases more complex rules than SQL is good at) in the application code as well, and also provides that single point of reference. There are arguments about the performance of schema enforcement in the application layer, but that's a race-your-horses problem to me. Commented Sep 29, 2017 at 16:57
  • 1
    @Paul: Of course the data constraints should be enforced at the application level also. That is a given. I'm just pointing out the risk of only doing it at the application level. Commented Sep 29, 2017 at 17:11