OpenID’s biggest problem is its learning curve. Using it as actually really simple, but if you’re not technical the amount of stuff you have to know before you can understand it is enormous. If you are technical, it just doesn’t seem like it should work—there are a bunch of questions that come up every time OpenID is discussed anywhere (“but surely there’s nothing to stop someone else from spoofing your ID”) which OpenID has answers for, but which are easily misunderstood.
The magical moment with OpenID comes the first time you log in to a site by typing your OpenID and clicking a button. In my experience, that’s the point where people convert from OpenID skeptics in to OpenID believers.
To that end, I’ve put together a screencast:
It’s 5 minutes 46 seconds long and demonstrates creating an OpenID (using MyOpenID) and logging in to a number of websites with it, then explains how other providers can be used to avoid having a single point of failure. It also touches on delegation; I decided not to cover that in detail for fear of giving people too much information up-front.
It’s the first screencast I’ve done, and I can confirm what everyone else has said: those things take ages! It took around seven hours to edit together less than six minutes of footage. If anyone’s interested I can write up some of the things I learnt in the process. Mark Pilgrim’s notes proved invaluable.
I’m hoping to get the screencast in front of as many people as possible, so please link to it if you find it useful. Link directly to the screencast page and not to this entry; the screencast page is being served statically by nginx which should hopefully help avoid a repeat of the meltdown I had on Tuesday.