« Music and breakfast | Home | I’ve been working on the railroad »
For the love of God, don’t call it Tagster
Posted by Mike Sugarbaker at 12:15 pm on 3/17/2005
Warning: this post is beyond buzzword compliant. We are doing COLD BUZZWORD FUSION here. So watch out. (And I’ve kept it approachable – so, the risk here is mere buzzword toxicity, not actual confusion. If you don’t mind that, read on.)
Buzzword #1 is social networking services. We’ve discussed them here before, but most recently in a spot where some of you likely missed it. To sum up, the current opinion is that social networking is the suck. Friendster and its nine million clones have failed to deliver on their promise to describe and augment human social connections, and they aren’t even that great at finding you people to hook up with.
Buzzword #2 is folksonomy – the currently-trendy notion that, rather than try and think up a taxonomy for how to arrange the world’s objects to cover your every user’s every need, you should just let your users make up labels for their own use and aggregate the results. Kind of a wisdom-of-crowds thing. Folksonomy is more commonly known as tagging, and it has replaced live chat as the most exciting and beautiful thing about Flickr. (There’s also this thing del.icio.us which combines tags with aggressive simplicity to become the first external bookmark server actually worth using. It’s still highly geeky, because it reflects its userbase so far, but it is indeed cool. Check it.)
So, put them together and what have you got? A social network mapper that can actually map the unforgivingly complex real thing. Draw yourself and the people you know as nodes, then draw little arrows between yourself and them. Those arrows don’t say “friend,” they don’t mean “co-worker” necessarily, they don’t give you a little menu of “crush,” “sweetheart,” and “friend with benefits” to choose from or anything like that. They are simply blank, one-way assertions of connection. But they let you tag them. Got a wife sitting around? Add her to the system, then point multiple arrows at her, tagging them with “wife,” “love of my life,” “blonde I have a crush on,” and “foe of my seekrit plans for Ken’s bachelor party,” respectively. (You might want to set that last one to “hidden.”)
On that note, there’s just one more thing to mix in. (Oh God, and I was already thinking that “remix” is the most overused word of 2005. Ah well: as I always say, if you’re not part of the problem, you’re part of some other problem.) The missing element here is friend filters. Every individual connection that you draw from yourself to someone else should be hideable, or in LiveJournal parlance “friend-locked,” not just depending on whether someone is on your list of friends but on an arbitrary subset or superset that you name. This needs to be per connection. So, you know, Kristine from accounting is linked to you as “co-worker” and everyone can see that, but that other “blonde I have a crush on” arrow you pointed at her really ought to be hidden from your, um, wife, and probably everyone else. And there will be more subtle situations too.
(Back at When.com we used to have an unofficial saying/design pattern: “we will not clean up your life for you.” We’d invoke this whenever it was pointed out that some feature we were cooking up might result in users getting conflicting appointments when they subscribed to too many calendars, or getting into some other awkward social scrape. We basically just decided that we weren’t in the business of preventing or solving social problems; we were a tool, not a butler. This kind of hands-off approach can lead to deeply worrisome phenomena, but I think LiveJournal shows us that, if you’re going to have social software that’s socially useful, you have to adopt it.)
You’ll note that, so far, it really doesn’t matter whether this is all centralized or de-. With the right tools, it could use all that tricky stuff developed by the FOAF people and the RDF people and whatever. Some of it relies on logins, which implies centralization, but we have a couple of options for that these days, and as I’ve written, social networking itself may eventually provide one. The key to adoption, though, is the user experience. People won’t buy into yet another social networking anything if it doesn’t get the real world any better than the competition does.
Computers, and their programmers, often don’t get the real world. Analysts of social networking services have told us this over and over. So let’s step out of the way and let the humans get the real world for us.
So: tell me why this idea sucks, or else I might build it. And nobody wants that.