Pull TextBlob out of capnpc-capnp and into KJ as kj::StringTree.  Use it in the stringification code to make that code cleaner and avoid using iostreams.

Allow SchemaLoader to use a callback to lazily load schema nodes, including dependencies of loaded nodes.

Enable some useful Clang warnings, fix some bugs it detected.  Especially, eliminate all dynamic initializers.

Use consistent naming style for structs that exist purely to specialize template aliases or functions.

Implement ability to set() struct and list fields by deep-copying from a corresponding Reader.  This copy requires no schema knowledge.

Mass rename 'reference' -> 'pointer'.  Been meaning to do this for a while -- all the documentation talks about 'pointers', not 'references'.

Dynamic API mostly done...  not tested.  This commit actually starts embedding the schema data in the generated code.

Implement 'Object' field type, which can be any pointer type.  The caller specifies the type when calling the accessors.

Basic code generation starting to work.  My machine just did something weird so I'm submitting and pushing in case of imminant hardware failure...