In the database world, developers are sometimes tempted to bypass the RDBMS, for example by storing everything in one big table with two columns labeled key and value.
Ironic, if you look at the Wikipedia schema. (At least last time I looked, which I admit was a few years ago.)
Good point, except the the irony is the reverse of that.
I think that many observers would say that the Wikipedia system works quite well considering they are one of the most widely used sites on the Internet. Since this article is published on Wikipedia, it would seem an ironic disproof of the article's contents that the publishing platform uses the very strategy that is claimed to be an anti-pattern.
I would go a step further and suggest that the technique listed uses the relational database such so that in the end it is more like a columnar database, although I have to think about that further before I feel more sure.
On the other hand, perhaps Wikipedia proves the point, and perhaps it could be argued that something other than a SQL DBMS would be more appropriate. I think the more rabid of the noSQLites would go that way.
Personally I prefer to be more descriptive than prescriptive and judge them on their success under extreme duress.
Ironic, if you look at the Wikipedia schema. (At least last time I looked, which I admit was a few years ago.)