Question

I have the following setup for my Blog Data Model:

Blog               Articles             Categories
-----              --------          1  ----------
ID  \ 1         *  ID            /----- ID
Name \------------ BlogID       /       Name 
Owner              Name      * /      
                   CategoryID-/

I’m worried about how this data model matches the ‘best practices’ for Blogs. If a Blog is deleted, should all the articles for that blog be deleted? If a Category is deleted, what should happen to the Articles?

Is there a better way? Is there something I’m missing? What do ‘best-practices’ suggest to do?

Answer 1

This really depends on your use case. A lot of blog applications use tags with a many-to-many relationship instead of categories with a one-to-many relationsip because it makes more sense to see an article as related to (possibly) many tags than to see an article as “contained” within a category.

If the logic of your blog is that an article is strictly contained in a category, then the one-to-many relationship is justified. And in some circumstances a cascading delete would also be justified. After all, where would the orphaned articles even exist.

This is just a case where you must anticipate your user-base’s intuition. What would they assume is natural.

Incidentally, it’s a lot more work, but if it were me, I wouldn’t do a cascading delete. I would mark articles as orphaned and hide them from most views. I would also provide some interface for restoring orphaned articles. And I would consider providing a script that deletes anything that’s been orphaned for more than some period of time. That way, if you were wrong about what your users expected to happen, you can fix the problem.