Access isn’t a relational database

Of course they are.

No, you just associate them with the model because most relational engines provide transactional control. I agree that transactions are an essential part of building databases for business applications, but they aren’t part of the relational model itself. They are an add-on to the model. You don’t have to believe me, Chris Date also thinks so: “[Transaction theory] doesn’t have much to do with the relational model as such” - Databases in Depth, P. 121, Chris Date, May 2005, O’Reilly – the book’s reviewed here.

OK, but there are other things that Access doesn’t have that are part of the relational model – like full support for all flavours of referential integrity.

Well, we can digress and argue about whether the different implementations of referential integrity are part of the relational model, but I accept the principle that Access doesn’t support the entire relational model. However, every mainstream database engine on the planet has bits and pieces of the relational model that it doesn’t support. For a start, they all use SQL which is not a particularly good match to the relational model. Chris Date again, from the interview quoted:

“So yes, I do think SQL is pretty bad. But you explicitly ask what its major flaws are. Well, here are a few:

• “Duplicate rows;
• “Nulls;
• “Left-to-right column ordering;
• “Unnamed columns and duplicate column names;
• “Failure to support "=" properly;
• “Pointers;
• “High redundancy.”

So all mainstream database engines fail to support the model properly which means that Access is no different from the others, such as Oracle, that you happily classify as relational.

But what about…

And so on and so on and so on... The argument will still be raging at closing time because it’s a database argument and, by definition, no-one ever updates their views.

We are all entitled to an opinion about this myth (since it depends so much on your definition of "relational"), but at least we can try to pare the discussion down to some fundamental points.