Everyone wants to (re)build their own pyramids (often from scratch) instead of just uniting to finish the top of the most promising one. But I don't judge too harsh the developers: my contribution to free & open source is quite small until now and free & open source it's a lot about free will and passion and not necessarily for the benefit of the world.
That's the thing -- what does Linux need at this point? Not much, and the things that are needed are not at the kernel level. The things on the kernel level that are needed are mostly from device manufacturers -- there is no shortage of people to do the work of writing a video driver, for example, if the specs were there.
Nearly every incompatibility we see right now is a result of a hardware company not releasing complete specs. It isn't for lack of coders that suspend/hibernate isn't working universally.
There are enough people working on the big pyramid.
That's the thing though, everyone has a different perspective of where they are working on the pyramid. One person's peak is so far below another's abstraction level as to be considered part of the foundation, and vice versa.
In oss, I think you can make contributions in two ways: refine or invent. To do either of these you have to learn the past.
Even if all the only result is an exact clone of BeOS an entire group of developers will have learned a whole lot about something state-of-the-art at the time it was created. That's knowledge that was previously available only to a tiny group of people.
