Instant Messaging (IM) interoperability has long frustrated me. A bit of history: In 1997-98, I worked directly on IM standards at the Internet Engineering Task Force. Five years later, there has been some progress on standards but little on real interoperability. Clients like Trillian can support the AOL protocol in addition to the MSN protocol, the Yahoo protocol, IRQ's protocol and IRC, but you can imagine how much work it is -- and how frustrating when AOL takes action to block clients like Trillian.

Is there any hope for change? There's no technical reason IM can't be as interoperable as email, it's business reasons that hold implementors back. All the big IM services currently benefit from keeping consumer customers locked in and viewing ads, but business use could be a bigger revenue stream and business customers may encourage interoperability. EDS recently tried to block IM, only to face an uprising from staff using IM software to communicate with customers. You can't block such useful tools; if companies like EDS (and Samsung, and Alcatel) are worried about the security holes created by ad-hoc use of these tools, they must arrange for the service to be available in a secure way instead.

Yahoo, MS and AOL have all paid only lip service to the standards recently. AOL could really turn the situation around as the market leader. MS and Yahoo care most about breaking open AOLs larger user base. A Yahoo VP was recently quoted saying, shamefully, "It's important that interoperability conversations need to take place among the big three... Any other interoperability is secondary".

I'd like to point out a common type of error in the same article: it claims that a potential standard is Session Initiation Protocol (SIP), "Co-authored by Microsoft, Cisco Systems and Dynamicsoft". Ugh. In fact, the co-authors of the SIP standard are people, not companies. The four people worked for an AT&T research institute, Columbia University, CalTech and Lucent at the time. Microsoft, Cisco and Dynamicsoft employees have provided contributions to SIP and its offshoots. It's really lazy of the CNET article author to screw this up so badly when authorship attribution is so easy to look up on the RFC standard sites.