Sun Sep 20 14:59:20 CEST 2009


Looks like the overall design is done.  In the mean time Dave Herman
fixed the show-stopping bug in c.plt, and I've started integrating it
in Staapl.  The plan is now to gradually move the libprim
implementation in C to a higher level description, so it can be
compiled to C or another substrate (i.e. a stack machine).  This would
also enable the linearity constraints to be machine-checked.

EDIT: the goal here should be to keep the current functionality
working at all times, and migrate it to a different implementation.