But how many people read them? Once you wind up having to point to the contract, the relationship is already south bound. Today people are readily writing software in languages with no specification other than the prototypical implementation. When communication was expensive, investment in big documentation was usually sound.
Today, I don't need an atlas of Georgia Counties. I've got Wikipedia and the US Census Bureau and free long distance on my cell-phone plan and minutes that go unused every month. If I have a question about Ubuntu, I don't need a manual. There's StackExchange and Google. There are lots of lightweight channels.
Don't get me wrong, I love books and deep knowledge. But the six months that it takes to produce, publish, and disseminate a tome doesn't offer slam dunk ROI. The agilists have a point.
You're confusing a bunch of stuff together here. Agile and powerpoint aren't the same thing. And BDUF/waterfall isn't the same thing as writing specs nor is it the same thing as using reports/essays/documentation to drive discussion and design.
The problem with PowerPoint is that it incentivizes laziness and promotes pitch-culture to the detriment of quality in-depth discussion.
As I wrote in another post, how I would drive getting off of the pitch-train is through internal blogs, wikis, and possibly forums. The way things are today you tend to have 2 main tools in use: the pitch and then reactionary response (the email thread, bug trail, backlog, what-have-you). That's very limiting, myopic, and reactionary. It makes it difficult to have meaningful discussions at a meaty level of detail on things (designs, architectures, directions, etc.) Those things tend to happen in email threads which get unwieldy rather quickly.
You certainly don't need to have a report/document/memo/manifesto about every little thing. But when you need to have a discussion about, say, the direction of your whole product stack, major engineering changes, that sort of thing, it hurts when what should be a conversation at a level of an RFC is reduced to a pitch-deck that people then either thumbs up or thumbs down.
Today, I don't need an atlas of Georgia Counties. I've got Wikipedia and the US Census Bureau and free long distance on my cell-phone plan and minutes that go unused every month. If I have a question about Ubuntu, I don't need a manual. There's StackExchange and Google. There are lots of lightweight channels.
Don't get me wrong, I love books and deep knowledge. But the six months that it takes to produce, publish, and disseminate a tome doesn't offer slam dunk ROI. The agilists have a point.