Here’s something that’s slowly making its way through the W3C working groups. CC/PP is a method by which a device — browser, PDA, SmartPhone, whatever— can describe itself and its user’s preferences to a server in the HTTP request. Think of it as a User-Agent string on steroids.
The device can specify its exact screen resolution, color depth, software, version, amount of RAM, plugins, etc. and specify the user’s preferred language, whether the user wants to see images or sounds, etc. The medium is an RDF-style XML string that presumably gets sent with the request.
All I can find about the project are W3C spec documents, which aren’t exactly scintillating to read but they do contain some samples which are pretty exciting when you think about the possibiities they’ll open. No more taking guesses about browser capabilities based on the User-Agent.
Content could be exactly customized for what the user wants to see and what their device is capable of showing them. Requests from PDAs could have a different stylesheet applied to them, drop images, etc. Requests from browsers without a JVM would be able to handle the ommission gracefully instead of some ugly “Download the Plugin” box.
A working draft was released on March 25, 2003, but I found other spec documents as far back as 1999, so this has been around for a while now. It’s way overdue in my mind.