All of the schools were closed today due to it being an icy, snow covered atrocity outside. Unfortunately, the school district didn’t announce the closure until relatively late in the morning. I take Oscar to school in the morning, so I’m working from home today which roughly translates into pecking away at the tasks that require the least amount of continuous attention in the minutes long intervals that I have between interruptions. They’re pretty continuous. Pair this with a relatively unresponsive VPN connection to work and you’re in for a frustrating disaster of just wishing that you could get anything accomplished other than at home tech support issues like how to get past a level of a game. I’ve also got a fairly spectacular headache, but that’s besides the point.

Because I couldn’t do any of the things I was supposed to do, I tried to catch up on the half a billion unread things in my feed reader. Most of it, like always, was crap. Jeff Atwood wrote this rundown of how to sanely handle web (really it could be any type of login, but since web apps are apparently the only type that exist right now) logins and save the user from suffering and I really enjoyed.

He spends a fair amount of time discussing the background behind his rationale for planning a sane and usable login system and I enjoyed reading about that as well. I’m pretty sure that no one will ever develop an open and accessible authentication service that meets all privacy scrutiny, support every language, and be relatively secure so why not make a best effort to utilize the ones you can? People have already agreed to heinous EULAs and are apparently capable of remembering those.  It really isn’t the responsibility of an application user to remember which of the million authentication methods you offer they used when registering initially and I’m glad to see other people wrestling with what some would label minutia.

If you’re building a cumbersome login system for something that people aren’t already familiar with, you’re probably doing it wrong. I guess that never stopped anyone in the past, but it’s worth reading if only for the amount of thought invested in methodology.