By Noah | February 8, 2009
The Self-Describing Web was published yesterday as an official W3C TAG finding. I am the editor, which means I did almost all of the writing, and I was among the TAG members who contributed significantly to the debate about the content. After two years of work, many drafts, and lots of interesting discussion, it feels good to have it done. The TAG is the senior technical body responsible for the Web, and I have been a member since late 2004. One of the things the TAG does is to help users of the Web, and those who publish information on the Web, learn how to use the Web better. Publishing “Findings” like this is one of the ways we try to set down insights and guidelines that we think may be helpful.
Why this finding? Several years ago Tim and others pointed out something that should have been obvious to me: if you’ve got a system like the Web where you can try to access most anything, and without coordinating in advance with the other end of the connection, it better be possible to know what to do with any response that comes back. The Web wouldn’t work if before accessing, say, the New York Times site, you had to call up the administrators at the Times and ask: how are you encoding the news today? Your browser must know what to do no matter which site you try to access or what comes back. So, it’s very important that the information sent through the Web be, in some important ways, self-describing. That’s the reason, for example, that each Web response carries with it an indication of the format (the media-type) of the information returned. Is it in HTML? Is it a jpeg picture? Your browser can tell by looking when the response comes back. When you start looking into these issues more deeply, though, quite a few interesting and important subtleties emerge, and that’s what the new finding tries to explore.
If you haven’t seen it before, you might also want to check out the Architecture of the World Wide Web, which the TAG was wrapping up just about the time I joined in 2004. It gives a quite readable summary of how the Web works, with lots of insights on how to use the Web well. Some sections go into lots of technical detail, but overall it’s written so even those who don’t know Web technology in detail can get a lot from it.
Also, a very big thank you to everyone who took the trouble to comment on earlier drafts of the new finding!