The first memo already had applications in it

Every time somebody says the web was originally created as a system for sharing documents there’s a small stabbing sensation somewhere behind my eyes because it isn’t true and it’s been repeated long enough to become furniture.

Go and read the original proposal yourself. Berners-Lee talks about pages built on the fly, and about special links out to external applications. The exact wording is that hypertext allows documents to be linked into live data so that the information is retrieved every time the link is followed and that if you sacrifice portability you can make following a link fire up a program so that diagnostic tools could be linked directly into a maintenance guide.

That’s an application launched from a link, described in the founding document, before any of the technology existed to do it well.

Dynamic applications running inside the browser weren’t really a thing until the later nineties and that’s the part people are half remembering when they make the claim. But reaching applications through the Common Gateway Interface, with HTML rendered fresh on every request, was there from very early on. That’s where I started my career and it wasn’t an exotic corner of the web at the time. It was how anything that needed to know something about you worked.

There’s a stronger version of the argument that gets missed. If the web had really been about static documents, it wouldn’t have needed to exist. We already had networked access to document systems. There were several of them and they worked. The thing that made this one worth building was precisely the parts that weren’t static and a proposal for yet another way to fetch a file would not have gone anywhere.

Then there’s the passage of time which is the part that should settle it regardless of what the memo says. That document is thirty four years old. The web in the sense people mean when they say the modern web has been running for at least twenty of those years. Even if the original had been purely about documents, there’s no reason to believe a design decision from 1989 still governs what the thing is in practice.

The reason you should care isn’t pedantry about history, though I’ll admit some of it is. It’s that the claim gets used to argue that everything since is a corruption of the original purpose and that if we went back to documents we’d have a better web. That argument needs the origin story to be true and it isn’t.

Plenty of the criticism you’ll read about the modern web is fair and I’d sign most of it. Pages are too heavy, too much of the interactivity is doing nothing for anybody and a good deal of what gets shipped as an application should have been a page. All of that stands perfectly well on its own.

It just doesn’t need a founding myth to support it, and building on one that’s wrong makes the whole case easier to dismiss by anybody who’s read the memo.

If you want to make the case that too much of the web is needlessly dynamic, and you should, then make it about what users get rather than about what the web was for. You’ll win that argument far more often, because you’ll be describing a page somebody is looking at right now instead of a design intention from before most of your readers were born.

There’s a smaller version of the same error that turns up constantly in this industry, which is treating an early design document as a statement of what a thing is permanently allowed to be. Protocols get repurposed, tools get used for jobs nobody imagined, and the people who built them are usually pleased about it rather than offended.