The userland would be harder to write a long lasting book on. It's prone to change between major version numbers.. for instance clang/c++ RT/etc. It would be

The userland would be harder to write a long lasting book on. It's prone to change between major version numbers.. for instance clang/c++ RT/etc. It would be cool to see a volume on this stuff, but it would probably need a ton of authors and be more like a series of exposes (thinking a book like Architecture of Open Source Applications)