Great post by Dare Obasanjo (MS employee) titled ‘New Features Are Not Innovation’ that everyone in software should read.

Along those lines, on the one hand, I’m thrilled to see that ekiga is working on improving their UI, but on the other hand, what I really want to do is lock ekiga, gaim, and evo in a room at GUADEC and not let them out until I have one data store for people and one UI to get at them. It isn’t rocket science, it just requires people reaching out beyond their communities and giving Linux users something that no other platform really has yet, which I think would represent actual simplification and innovation, in ways the various proprietary OS silos just can’t do. It seems sort of sad that orph might get closer, faster, to that with gimmie than the actual clients have. :/

[Edit later: I realize I said ‘data store’, but that was unclear- what I meant is that to the user it should appear as if there is one data store- i.e., the totally false and historical breakages between communication apps should be broken down. The data can be stored in 10,000 different places, or on one shared set of punch cards, for all I care, as long as when I get an IM from Jeff I can instantly and sanely decide to respond via IM, email, or VOIP, and when I decide ‘I want to talk to Jeff’, I open a Jeff object, and then pick an appropriate method, instead of then having to pick through several databases and tools to figure out the best method.]