Last week SourceForge quietly added support for OpenID to their site. The news is official now.

SourceForge implemented relying party support (as opposed to just being a provider) which is a trend not often seen by larger players. I wanted to talk with one of their developers to see what it took to make this all happen, especially in a large organization like SourceForge. I spoke with Luke Crouch who was the lead developer on the project.

In this podcast I try to cover some of the questions that large sites have to consider when adopting OpenID as well as ask a bit about the future for open technologies at SourceForge. Hope you enjoy it.

If you use open source software then you’ve probably heard about SourceForge before.  If you develop open source software then you’ve probably even used some of their infrastructure in the past.  Today they’ve made it even easier to login to SourceForge with OpenID.  SourgeForge.net isn’t acting as an OpenID Provider but rather is accepting OpenID logins; this is a good thing and reinforces the trend of sites like Ma.gnolia only accepting OpenID logins.

In their announcement OpenID on SourceForge.net they say, “OpenID is getting tremendous traction and we’re happy to be jumping into it. it’s bringing us back in touch with fresh web (2.0) technology. as a decentralized open-source standard, it’s a perfect fit for us - it allows us to streamline more user interaction and participation with our site, and hopefully more for the whole OSS community.”  As Steven Osborn points out, SourceForge.net is now one of the most prominent single sites that accepts OpenID to login.  Steven also goes on to talk about some of the more advanced things SourceForge allows you to do with your OpenID if you do wish to use your profile URL as an OpenID too.

Congratulations SourceForge for continuing to help get OpenID in the hands of open source developers on a daily basis!

Over the past two and a half days, nearly fifty different people came to the first ever OpenIDDevCamp hosted in San Francisco.  Nearly twenty people showed up Friday evening to start drawing up the agenda for Saturday from a list of possible projects.  Saturday we had our first full day of discussions and hacking around OpenID today.  About thirty-five people showed up and hacked, talked and shared.  Sunday started out a bit slower though my noon everyone was back to cranking on a new OpenID test suite, XRI debugging, and OpenID usability with some focus on mobile (partially inspired by Chris’s blog post).  Thanks to Vidoop for sponsoring breakfast, thanks MyStrands for sponsoring lunch and of course big props to Six Apart for hosting the event.  A bunch of photos can be found up on Flickr under the “openiddevcamp” tag.

Much was done … here is just a sampling:

• Brad Fitzpatrick, Ryan Barrett, and Sam Alexander wrote a new semi-automated test suite for OpenID relying parties and providers. More information can be found at the Google Code OpenID-test project page.
• Discussions of OpenID + OAuth and how they can seamlessly complement each other.  David Recordon and Joseph Smarr started to document this OpenID Extension which builds atop OpenID Attribute Exchange.
• A good group discussion about the possibilities around what we can land at OpenID end-points. (see Scott’s post on this). Consensus is that we’ve got the pieces and now its time to crank out some code.
• Joseph Smarr upgraded Plaxo to be an OpenID 2.0 relying party and debugged their support of i-names. (woohoo!)
• Christopher Allen helped to clean up the OpenID wiki and refreshed the page on setting up delegation.
• Evan Prodromou updated his OpenID plugin for MediaWiki to support 2.0 as a remote participant!
• Eugene Eric Kim and Matthew have a Perl XRI implementation (almost) done.
• Chris Messina talked about the DiSo project with its possibilities around XMPP and discovery.
• Randy Reddig continued his work from a few days ago on updating the OpenID logo especially to work better in small sizes as another remote participant.
• Chris Messina led a discussion on OpenID usability where we worked on developing relying party best practices around login interfaces and error messages.

Additionally, getting an entire group of OpenID veterans was a great way for people to debug their problems and learn new things:

• A few people got their Ruby on Rails blogs OpenID enabled for commenting.
• Getting the Java version of consumer library installed and figuring out what it takes to be a provider (hint: acting as a great provider is hard).
• Zentu now lets you login using OpenID!

Some takeaways and things the group would like to try to accomplish for next time:

• Translate vCard attributes to Attribute Exchange schema end-points.

Planning and organizing OpenIDDevCamp really wasn’t that hard so if you’ve ever thought about putting together an event like this, you really should!  Feel free to use this event’s wiki page as a template.