4tH compiler

I've uploaded an improved version of F10**, extending the range to 1e24, knocking off an FP operation on average and making the DO..LOOP a bit tighter as well. The trick is to use the "unused" space between 1 and 1e9 by starting off at 1e9 from the beginning and covering the range from 1e10 thru 1e24 by FP and the range between 1 and 1e9 by

BTW, ErrNo is much like the C "errno". It's a message number and can be used as an index to the array in errs_4th.c. It is used in both compilation and runtime. In 4tH, it is usually manipulated using better not manipulated from a host C program. Hans Bezemer

So do I. As a matter of fact, I NEVER wanted "native" support for double or floating point number, because: (a) Nine out of ten times you wont need it; (b) It's quite complex, especially for beginners; (c) I didn't want to drag the overhead around. Note due to 4tH's (Harvard) design I would have to assign a whole new segment to FP and

I always wanted to write and maintain a compiler.. ;-) Hans

Sounds like you're doing great work, Hans. Now we know why Chuck Moore always preferred scaling fix point to floating point. Best wishes, LRP -----Original Message----- Sent: Wednesday, March 14, 2012 3:34am To: "4tH-compiler" ... [More] <[email protected] > Hi 4tH-ers! I've almost finished the ANS FP overhaul. Lots of - sometimes tiny - [Less]

Just to let you know I'm reading this 4th stuff and even saving some of it for future (if unlikely) study. In between times I'm having to print out music for Jennie's Chickens and Dill Pickle Rag for the upcoming practice session of the Barley Shakers celtic band. Meanwhile downstairs I'm trying to cut out a wooden plaque with my home-brew CNC