Tue Aug 24 21:55:11 CEST 2010


A bit too much freedom to pick from so many targets.  One thing I
wonder about is whether the PIC was actually a good choice.  It is
really different from standard RISC.

ARM cores are getting really cheap.  Programming them in C isn't a big
deal as I've recently learned;  It feels like "normal" programming.

So why am I doing the whole tool stack myself?  Maybe I should keep
these questions away from the project and worry about it in the
libprim/meta projects.

    * Staapl is about the PIC18.

I don't really have time to make it work on different architectures
(tools + libraries).  Maybe one day LLVM, but that's it.

    * Staapl is about Scheme and Forth.

To make Staapl more interesting community-wise, it might be a good
idea to get the standard Forth interpreter going: automatic
"unrolling" bootstrapping.

About the dsPIC.  This is an interesting architecture from an
application side, but probably better for the other metaprogramming
project.  In any case, I should not attempt to do anything before
writing an app in its machine language..  Then the road will become
clear pretty fast.