> Using this code without changes, sfront warns that "the statement
> never runs near newfreq" and there is no sound output. When the line
> "ksig vib, newfreq" is changed from "ksig" into "asig", everything works
> fine and sound is generated. Shouldn't it run ok without this modification?
The correct semantics for k-rate and i-rate statements in aopcodes,
and for i-rate statements in kopcodes, was the subject of much
discussion on the saol-dev mailing lists over the summer, and from
my understanding of the discussion Eric is planning to post a first
draft of corrigenda which lays out exact semantics for these cases.
My plan with sfront is to wait for this document to appear, and
then do a preliminary re-implementation of these semantics -- I
really don't want to re-implement this semantics twice, once now
and once after the corrigenda appears, which I'm afraid is what
will happen if I try to divine a "consensus semantics" from those
mailing-list archives and implement that now ...
The present sfront behavior is "only i-rate statements in iopcodes,
only k-rate statements in kopcodes, only a-rate statements in aopcodes",
which is semantically is pretty straightforward ... and will result
in SAOL code that will almost certainly be "backward compatible" with
whatever semantics result.
-------------------------------------------------------------------------
John Lazzaro -- Research Specialist -- CS Division -- EECS -- UC Berkeley
lazzaro [at] cs [dot] berkeley [dot] edu www.cs.berkeley.edu/~lazzaro
-------------------------------------------------------------------------
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 11:46:41 EST