Where HTML got angle brackets from


http://developer.yahoo.com/yui/theater/video.php?v=crockonjs-4
.
.

. "One of his ideas was to take all of the display-oriented stuff out, so you were only talking about the structure of the document. You can recognize there are tags in this thing, right, and some of them may appear eerily familiar to you. That coincidence is not accidental. The one that you might not recognize is the EOL, but you can probably guess what that one's supposed to do. As GML went through its evolution, first he had custom tags for closing, then later he came up with a double colon as a thing that would close it. Then at some point the discovered angle brackets. Like oh, angle brackets! That sort of changed the whole thing, and you all know what happened with that.

Where did the idea for angle brackets come from? It came from here, Brian Reid's scheme. While Goldfarb was doing his stuff, Reid was a student at Carnegie Mellon. He developed a document processing system called Scribe, which was the first time in history that anybody had gotten the separation between structure and presentation right. Brilliant system he came up with, an elegant, minimal language for expressing documents.

There's only one reserved character in his language, which is the @ sign. If it was followed by a word, then that word would be the name of an environment. Then you could have some block of text, or nested stuff that came after it, which would be the input to that environment. In order to give authors lots of options, he had six sets of quoting characters that you could use to define these blocks, one of them being a pair of angle brackets, which really impressed Goldfarb. Another nice thing he had was that for dealing with very long things or deeply nested things, he had special forms begin and end, in which you plug in the name of an environment, and then you don't have to worry about accidentally closing out of the contents. Scribe was a really elegant, very, very nice way of creating documents. It was very easy to write in it, and it was very easy to get things right.

Another thing he had was support for bibliographies. Here are examples of declaring a tech report and a book. The interesting thing about the way he does that is it looks like JSON. This is where the idea that you could represent data in a document format came from. This idea eventually got moved into the XML community, but unfortunately they didn't get enough of Reid's stuff to get it right. But what they did get from this was the idea of attributes. Later we're going to see attribute equal sign equals some value, and this is where it came from.

Scribe was very influential. Scribe inspired SGML, Standardized Generalized Markup Language. It also influenced LaTeX, which was a document processing mode for Donald Knuth's TeX, which is a brilliant typesetting system. The difficulty with TeX is it's probably the most powerful ever made, because it's not just a document description language, it's actually a programming language. So its complexity is maybe beyond what a lot of people are able to deal with, but it's really good stuff, and Lamport's LaTeX stuff made it significantly easier to use. SGML was a popular product at IBM's Federal Systems Division, but not much else. So if you happen to be in a Governmental, or quasi-Governmental organization, there's a chance you might have been using SGML. The rest of the world said nah, we don't want that. So it was doing OK in its niche.

It turns out one of the quasi-Governmental institutions that was using SGML was CERN, and CERN was where Tim Berners-Lee was. Tim Berners-Lee adopted SGML as the language for his World Wide Web because that was what he knew, and it was already used in-house. If CERN had been using Scribe, or if Lee had been more up to date on what had happened in the history of document processors, the world might be a much better place today than the world in which we are, but that's the way it turned out. Now, the SGML guys were shocked at what he did to HTML. They were completely disgusted, he had broken their rules… Well, we'll get to that in a moment.

HTML was not state-of-the-art when it was introduced in the late 20th Century. It was intended for simple document viewers, and nothing else. That was its complete mission. It was not intended to be an application platform, it was not intended to be any of the things that we require it to be now. All of that stuff came later. A lot of people looked at the World Wide Web as it was first proposed and thought it just didn't have what it takes. I was one of those. And in fact, we were right. But things happened, and it changed."
.
.
.
On the video it's minutes 8 through 12.