[Chicago] Chicago Digest, Vol 97, Issue 24

[Martin Maney]

On Thu, Sep 26, 2013 at 06:05:06AM -0500, EDoxtator wrote:
> 
> > On Sep 26, 2013, at 5:00, chicago-request at python.org wrote:
> > 
> > ...but does anyone actually still use FORTRAN?
> 
> SPSS (statistical software, now owned by IBM) was written in the
> 1970's.  It's engine is FORTRAN77.  I understand the question was
> intened to show the impracticality of the book's examples, but it's
> interesting to see where little used languages show up.

Yeah.  Software Tools actually uses RATFOR, which is a much less
FORTRN-ish language; one that is pretty easily compiled into FORTRN -
one of the chapters/projects is writing the converter program.  It's
really an amazing book, and forged a record which I'm not sure any
other computer text has followed since, to our great loss.  Since this
was written by hard core hackers, the text was, of course, all on disk
(or tape, maybe.  or both, probably).  And they tested the code
extracted from the marked-up text, so the only bug any of us ever found
was one I noticed in the one bit of code that they *couldn't* test - a
bit where they showed how simply a certain function would be if only
the language supported recursion.  The much more complex implementation
of the recursive algorithm using explicit stacks of state was bug free,
but that example had an off-by-one or something along those lines...

Actually, I guess we've moved back in that direction these days, in
that the code that's presented is usually the code the authors tested. 
But it's rarely as solid and practically never as well described.  I
harbor a secret belief that Software Tools is what Knuth hoped literate
programming would equal.

-- 
The reason [limited term of copyright is] important is this:
Publishers are in the business of expanding capital.
The writers who supply them are in the business of
expanding civilization itself.  -- John Bloom