I am rather disappointed about this, because I can see the huge potential of WPF, like my favourite Zurich Airport WPF demo from a couple of years back. Any business will always review the popularity of certain products, and I’m sure the download figures for the WPF messenger must be abysmal. It will inevitably take some time before the reasons why Yahoo took this decision start to percolate, but lack of popularity must be a key aspect.
What is great about programs like Skype and Windows Live Messenger (formally MSN Messenger), is that they are so simple to use, and very responsive. The average end user of said applications is always in a hurry to either send a message, or respond to one, so-much-so that really fancy graphics end up being a hindrance, especially if they are used to something quick, which WPF has not always been. For me it is a case of “just give me a text box and the ability to send my message”. It is as simple as that.
As a developer, you need to be really judicious about when to elect WPF for an application, because what you think is pretty, may actually get in the way of usability, which is the key mistake Yahoo have made. They got to too fancy with the WPF UX and forgot about simplicity.
If I was tasked with building a Messenger client (assuming the design time experience had been improved to that of windows forms at least), I would build an exact copy of the previous UI, and gradually add the features that make WPF stand out when compared with Win32. This would result in an application that appears to an end as a succession of the previous, and not having to learn something new again, to achieve the same result i.e. send and receive messages.
Microsoft have also been erroneous, principally by not eating their own dogfood. It is so important for companies to “practice what they preach”, but they too have stuck with Win32 for Windows Live Messenger, is it a surprise that everyone else has as well?