Now that the shit has hit the fan, what will become of Boost.Build? I still prefer it to CMake even if Boost won't use it (whether I can convince my project to keep using it may be another matter). I have greatly appreciated Steven's, Vladimir's, and Rene's improvements over the years.

Rene, can you expound on what you mention here?

* I will continue to maintain Predef. But will be forking it into a
non-Boost form moving forward (in similar vein to ASIO).

* I will *not* continue to maintain B2 in the Boost form. I will be
rewriting b2 into a new form to address the needs of building software that
matches the industry I work in. I will be doing this immediately.

What does it mean to rewrite b2 in "a new form" that "that matches the industry I work in"?

I will be particularly irked if the Boost SC ends up hiring contractors to speed the conversion to CMake, when the issues in Boost.Build (documentation and it being quite hard to learn how to do advanced things with it) could be solved by far less contracted effort.

-Matt