Over the weekend we made a bunch of bug fixes and improvements to Twitter which resulted in more reliability in message delivery—this increased overall traffic.
Heh, and as I speak, Twitter down. Like, completely down. Oh wait, it's up again! Nope, back down. Jimmy says it's back up for him. Wait, no, down.
Anyway, this has me thinking about the reliability of services overall, and how we're truly in the dark ages-- still. But the remarkable thing is that we persist. There's a graph someone must have made, in a world I've never seen, where a user's affinity to a service ultimately trumps the unreliability of said service.
Take MySpace, for example. Notorious for growing pains, it grew nonetheless. But MySpace provided a compelling value, and thus the user persisted. I know this is rudimentary to some, but it isn't readily apparent to others. And frankly, I think folks in corners of the tech industry would do well to remember these basic concepts.
Twitter may be ephemeral, and not entirely "useful" to a mass audience. Not yet anyway. But stuff like Vox, where you can essentially lock down posts for a certain audience, and open posts to a larger audience, plus the bells and whistles it provides (I really need to update my books, but I love that feature) make it a very compelling service.
Still, it isn't without frustrations. Last night my wife tried to add "widgets" to her page. She doesn't use the Dashboard in OS X, nor does she give a crap about Vista, so the very notion of widgets is essentially new to her. Unfortunately, the widgets in Vox presume a certain level of technical sophistication she lacked. It really broke down when she tried to add widgets from Weatherbug just as she'd seen on another person's blog. She couldn't, and after damn near an hour trying, she gave up-- chances are, like most consumers, she won't try again for a very long time.
Oh, just try doing a Google search for "widgets" and see what you get. Even if you add Vox to the mix, you aren't presented with a very useful explanation! Again, there's a lot going on here. There's the presumed tech know-how at SixApart. There's the unwillingness of the consumer to learn a whole range of nomenclature, no matter how cutesy you make the jargon. Ultimately there's the way any system handles an error or unexpected user input.
Tonight in game design class we played a number of really tight games: Assassin's Creed, Bioshock and Mass Effect. Know what was one singular thread throughout them, and any successful game? They don't presume a damn thing. They walk you through the system. They teach you, they accept faulty inputs gracefully and the really good ones let you succeed despite repeated failures. That doesn't mean they are facile or silly. It just means they are designed well. I fear many modern web services are failing to gain traction because they have lost sight of truly great design-- opting instead for whizzy new features that most people don't care about.
Personally, I'd rather have stability than the ability to see my Tweets do backflips over the moon.