Feature volume rises as applications and services merge and soon we will need the power of Google within single applications. Of course, there are reasons for this that lend to the idea that we will not have single applications in the future. As applications migrate into services, and services combine and interact, the whole of software is evolving into a massive software ecosystem. Every state of software can be integrate, broadcast, and pull from a host of other global services. The number of "features" available at any point is rocketing into unimaginable heights. Until we can automate the integration, filtering, and aggregation of the mass of services we have for working with the same data set, we do not benefit as fully from their availability.
Jeff Atwood brought this up in context of Office 2007's Ribbon and the Scout plug-in that may not see the light of day, for internal political reasons at Redmond. The apparent story is that adding a feature to search their interface, even optionally, would undermine their attempts at marketing the glory that is the Ribbon. Of course, a searchable Ribbon is leagues beyond the traditional mess of menus and toolbars. Embrace of this concept would do nothing but benefit them, and give a head start in giving users a compass to navigate the ocean of features coming to them. Usability is about to transform from a gentle drift to a tidal wave.
I want to expand on this, but it is for another post. Features adapt into web services. Microformats and service discovery replace Plug-in systems. The interfaces of our applications will become a search engine of features, contextualized to the present task. When I can gather some information and thoughts on these subjects, I want to produce something interesting to gather the ideas into one place.