Thu May 14 10:59:03 CEST 2009

low-hanging fruit

This partial eval stuff probably needs a rest.  It will involve a lot
of thinking, but what I need now is some easy success to boost

Maybe it's time to start porting to the dsPIC.  I have some parsed
manual lying around somewhere to get to the instruction set.  However,
it might be best to do it manually.

There is one thing I'd like to change though: the assembler should be
more compositional.  Right now there is a single table with all
opcodes, but it's probably best to separate out all function

This leads me to think that an opcode is really just an argument to a
more general instruction.  Anyways...

Hmm.. dsPIC isn't that low-hanging due to different instruction set
architecture.  It would be cool as a real platform though.  Chips are
PDIP and small.  Maybe I should take this as a hint: find a generic
approach to bootstrapping a new architecture.