lugjg@vax1.cc.lehigh.edu (12/29/90)
I just thought that I should mention that there is a raging debate concerning 'why FORTRAN is object oriented' in comp.lang.fortran. And it seems like they could use some help.
burley@pogo.ai.mit.edu (Craig Burley) (12/31/90)
In article <130.277b6062@vax1.cc.lehigh.edu> lugjg@vax1.cc.lehigh.edu writes:
I just thought that I should mention that there is a raging debate
concerning 'why FORTRAN is object oriented' in comp.lang.fortran.
And it seems like they could use some help.
I disagree -- I've been responding to some of the questions and comments.
I don't think anyone is saying FORTRAN is object-oriented, just that, as
with any language (including assembly), the collection of programming
methods we refer to under the umbrella term "object-oriented" can be
applied to Fortran programming, and that it might be nice for Fortran
to someday offer features permitting more direct translation of the use
of these methods into Fortran code itself.
I did post an example for Fortran programmers of what good OO Fortran code
might look like, IF there was such a thing, and discuss why it was a bit
of an improvement over current Fortran techniques, and how it could
be handled optimally (by degrees) by compilers. But I don't think anything
I said could be reasonably interpreted to mean "Fortran is object-oriented".
It ain't!
That's not to say we wouldn't welcome some real comp.object jocks into the
discussion on comp.lang.fortran. And if I've missed some posting saying
Fortran IS object-oriented (even Fortran 90, which does come a bit closer
but does not include inheritance, for example), feel free to jump all over
it!
--
James Craig Burley, Software Craftsperson burley@ai.mit.edu