All in the <head> – Ponderings and code by Drew McLellan –

OpenID Has Failed. So What's Next?

37signals, a fairly early and significant adopter of OpenID, has announced that they’re dropping OpenID support from their products. Whilst one swallow doesn’t make a summer (or whatever the reverse of that is), it’s a fairly notable moment when a company like 37signals strips out a technology from their platform due to it causing more harm than good. They’re not just making it unavailable for new users, they’re migrating all users off OpenID and onto their native login system. They want it gone, citing the customer support issues it creates.

You know what? It’s fine. OpenID has never really caught on in a major way. I use it pretty much only on StackOverflow and 37signals products, and I really like it. Regular (non-web-building) users find it puzzling, and don’t have a good understanding of URLs, let alone ownership of them. So it works for geeks, but not for the general populous, and therefore isn’t a good viable solution for most sites.

That’s fine. No point fighting it. This iteration hasn’t worked, so let’s make a note of why it failed and start work on the next. The problem OpenID was attempting to solve hasn’t gone away.