People have different theories of software. There's 37 Signals and then there's Microsoft's.
I'm not a huge fan, but I acknowledge that it does a lot. It's Microsoft's server strategy for the next decade or so, most likely: a place where office docs can live and move around in automated business processes. That means it has a learning curve -- a learning curve that most users will never climb. Which means yes, it's going to suck for a lot of people.
As an aside, the "do one thing and do it well" camp is a great idea in theory, but in practice sucks. Big companies make purchase decisions, like it or not, based on big feature lists. Products with big feature lists score better than those without.
I don't especially like this situation, but there it is. Simply complaining about it is not going to make it go away.
> As an aside, the "do one thing and do it well" camp is a great idea in theory, but in practice sucks. Big companies make purchase decisions, like it or not, based on big feature lists. Products with big feature lists score better than those without.
That is, in itself, a very interesting observation. Reminds me of PG's talk on Viaweb and the use-case for their template language: "users always
want an upgrade path, even though as a rule they'll never take it."
That's exactly why it sucks balls.