If it ain't programmable, it ain't a computer.

While I sort of agree, there is such a thing as a a "programmable computer", which would be rather oxymoronic is the two words meant the same thing.

Then there's also the question of what programmable is. Does it count if the problem is fixed, but I give different inputs? What exactly is the difference between a problem and an input anyway? Is it the same equation but for a different value of x? What if it's the same equation but you can alter some parameters? In some languages (Wolfram Alpha comes to mind), there is *no* grammatical difference between code and inputs and outputs. You could say it's only a computer if it's Turing complete, but technically no computer with finite memory is Turing complete.

While not quite semantics, the whole definition of these terms is rather woolly. It does make for good pub chat with a philosopher of mathematics if you are that way inclined however.