[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Question on ProDOS SmartPort drive remapping



Eric Smith wrote:
I wrote:

There are plenty of legitimate complaints one could make about Fortran,
but I've never before heard anyone claim that it is not an HLL.  How
do you justify that?


Michael J. Mahon wrote:

The original FORTRAN had limitations quite transparently derived from
the 704 architecture, and FORTRAN IV had the limitations raised in
accordance with the extensions of the 7094!

As an example, the three-level DO loop nesting limit was a consequence
of the 704's (and up to 7090's) three index registers.


The arithmetic IF was a consequence of the 704/709/709x architecture as well,
but nevertheless FORTRAN compilers were developed for other machines.

Right--after everything is Turing equivalent.  ;-)

The point is that FORTRAN was designed to fit the 704 like a glove,
and no other machine was considered.  I don't really see any problem
with a *machine-oriented* high level language.  ;-)

Perhaps "not an HLL" is too strong, but it certainly was not conceived
as machine-independent.


The amazing thing about a dancing bear isn't how well it dances, but the fact
that it dances at all.  By modern HLL standards the original Fortran is quite
crude, but in 1957, Fortran must have been pretty f@#$ing amazing.  For that
matter, so was Commercial Translator, the predecessor of COBOL.

FORTRAN certainly was amazing!  It's developers knew that the code it
generated would be compared to hand-generated assembly, and so they
chose the language elements so that translation would map well to the
704 and even included an optimization pass that did a dynamic simulation
of program flow to figure out where to make tradeoffs.

The FREQUENCY statement that provided the probabilities of conditional
branches for this simulation was relatively short-lived, partly because
we now know that programmers tend to be *terrible* estimators of branch
probabilities!

The great lengths they went to to generate good code were a direct
result of the total lack of credibility regarding machine-generated
code--and FORTRAN won over almost everyone eventually.

-michael

NadaPong: Network game demo for Apple II computers!
Home page:  http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it's seriously underused."