In those days, FORTRAN IV was handy as a "portable" language just
because it was the one language (other than COBOL) available
*everywhere*. It could be used as a sort of "high level assembler" too.
My first job out of college was working on a large database system. It
was written using a pre-processor (think of RATFOR) and a bit of
per-platform assembler for things that couldn't be done in standard
FORTRAN (e.g. random access files, OS calls). It was a real, high
performance database system that ran on CDC, UNIVAC, IBM 370, Xerox
SIGMA, PDP-10/20, and VAX. Each new platform usually turned up some
minor quirks. We always said there were no portable programs, only
programs that had been ported.
Gary
On 4/4/25 23:44, Chuck Guzis via cctalk wrote:
On 4/4/25 20:15, ben via cctalk wrote:
I thought FORTRAN IV was the portable programing
language.
They have talked about having smarter high level programing languages
for years. Has that gotten anywhere?
Back in the heyday of FORTRAN, the universe of architectures was a lot
more diverse and foreign to languages such as C. Non-binary (decimal)
numeric representations, Binary ones' complement math, vendor-unique
character sets (it's why FORTRAN (and COBOL) is written using a very
small common character set)--and lack of character-handling operations,
as well as Boolean functions.
I think that the original PALASM was written in "portable" FORTRAN. One
of the more common ways to start a "portable" program was to READ a card
with the character set punched onto it in A1 format. You manipulated
characters by referring to their positions in the alpha A1 array.
--Chuck