The internet is all about linking, but I think most people think of links in the wrong way — they think they are the same as URIs. A link is actually what it is in real life: link n. a relationship between two things or situations, esp. where one thing affects the other. Thus, what we’re bookmarking in our browsers, recording on del.icio.us, and sending to each other over IM are not links, but URIs. Since links are two-way, the URI we end up with is just one of the two endpoints of the link. This is usually what we want, but I wonder how useful it’d be if the recording mechanisms I mentioned above allowed you to work with the link as a whole. Would the extra context allow us to stop tagging our del.icio.us entries? Or would it merely complicate things? What if you could view your browsing history not as a time-sorted list, but as a directed graph? Might that be more interesting, if not more useful?

This isn’t all pie in the sky either. How do you think Google does its PageRank stuff? 1 It’s useful for businesses, and it just might be useful for consumers too.

Speaking of context, maybe you’d like to know what prompted this line of thought. It all started, of course, with a link — this one was from someone2 in #caboose, asking about a guy named Giles Bowkett, wondering why he wasn’t a #caboose member. It’s a pretty cool blog with some interesting ideas — and a few good links. One of which was to another blog by another developer who, like me, has an interest in Ruby and Haskell, among other things. So suddenly I’ve got two more RSS feeds in NewsFire, a bunch of backlog to read, and two more books in my wish list.

1 No, it’s not pigeons.

1 You see, I’ve already lost some of the context. I think it was technoweenie, but I’m not sure. What if I could track the link from the IRC conversation where it originated (from my point of view) to its destination?