I'm considering using literate programming to develop my next
embedded, real-time project. One of the advantages that I see is
being able to include the statecharts (state machine diagrams)
directly in the*source code.
My question: would it be ``better'' to create a simple statechart
drawing tool that reads/writes*diagrams, or to create a set of
*macros for specifying the statecharts directly?
I would really prefer *not* to save the statecharts as external binary
files due to the difficulties with revision control of binary files.
I know that a simple statechart drawing tool that can read/write*
directly out-of/in-to my nuweb source file is within my capabilities
as a software developer. And that was my first idea for a solution.
But then I saw the xypic package and was astounded by what could be
done with TeX graphics macros. And then I wondered if I could develop
a set of macros that would allow me to specify statecharts directly in
the*that would be readable and maintainable.
So what are your opinions? Would a ``statechart'' package be
feasible? Or is it better to stick with what I already know on this