Compiler actually running plugins!  Hooked it up to capnpc-capnp (the loopback plugin).  Working on getting the output identical to the old capnpc...  still a few bugs left.

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

Improve const-correctness of readers and builders.  We enforce transitive constness, meaning you should not be able to construct a mutable Builder from a const Builder (you can always convert them to Readers instead).

Don't trust trusted messages so much.  Just trust that the pointers are valid and in-bounds, but don't assume the data matches the schema.  This makes it easier to safely create a 'trusted' message: just copy any message into a MessageBuilder, and if it ends up all in one segment, it's safe to use.

totalSizeInWords() method on generated struct readers/builders returns total size of the struct, not counting far pointer overhead.  Useful for allocating space for a flat copy.  Also a couple misc tweaks.

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'.

Restore ability to upgrade from List(Bool) to List(T) where T is a struct whose @0 field is of type Bool.  I previously disallowed this to reduce complexity, but it turned out to actually increase complexity.

Starting on dynamic message manipulation (manipluating messages without knowing the schema, or without compiling in the schema).

Delete Code Friday:  Inline structs/lists/data are too complicated for not enough benefit.  There goes a week of my life that I'll never get back.

Correctly handle inline fields that aren't present in the parent struct (i.e. because we're reading an older version of the struct that lacked those fields).