That's a fair criticism. But that is what I've called it for a long time.

> The real significant change would be switching to a NoSQL database but if you’re doing that then it’s not SQL you need to be “sanitising” anyway.

Right, that's why I started off by saying "Sanitize your inputs so that it can be safely stored in your data store", so that it could apply to any data store.

It's just a terminology argument at this point. But my main point was that you need to still think carefully about how you're going to store your data and do it safely, and then also make things safe on the way out, like the article suggests.

I think many people would agree that "sanitization" is loosely defined and I think that's exactly what led to the misunderstandings that the OP's article is trying to address.

> that's why I started off by saying "Sanitize your inputs so that it can be safely stored in your data store"

That could be interpreted like "escape quotes at the start of the request if you know that you're using a database where quotes have a special meaning" a la PHP magic quotes, which I'm guessing is not what you meant but it is what the OP is criticizing. The key is that the sanitization (or whatever you want to call it) shouldn't happen until you're ready to insert into the DB, otherwise that data will be coupled with database logic through the whole flow of your app

Sanitizing your input means changing them permanently. You don't actually want to do that. You want to store exactly what the original value was, but you want to do it safely. When you retrieve it again, it should be the same as it was originally.

If you "sanitize" the input, you won't necessarily have the original value ever again.

(I've seen the line of thought taken one step further: taking the realization that it's impractical to make strings universally safe for any context—even if you HTML entity-encode it twice, what if a recipient decodes it three times?—and concluding that security is hard and we can only approach it asymptotically, so shrugs XSS-like bugs are normal and unavoidable given finite time & budget.)

If the mindset is more like converting units, it becomes clearer. You can't concatenate HTML with a general Unicode string without converting the string to HTML first, any more than you can add inches and centimeters directly. "Cleaning" the centimeters would make no sense.

I think this is true, but doesn't quite reach the point GP is making: speaking correctly is also important. Calling parameterization input sanitization communicates the wrong message. And abstracting the wrong solution to apply it to a different problem isn't all that helpful. You could just as easily encode or hash input to fit the underlying data format without losing data (except in the case of truncation), but that isn't input sanitization, either.

Input sanitization is strictly checking front-end input against a ruleset and rejecting anything that does not comply. This is fundamentally different than dealing with anything thrown at you and handling it gracefully.

That's validation. Sanitization involves altering the data to make it safe (comes from the word "sanitary").

I think most devs think of sanitization as "make X safe", not "see if X is safe, if not reject" since that is usually called validation.

Using hand sanitizer does not remove your hands if they have harmful bacteria. The hands are still there, just cleaned from (some) of the harmful parts.