[alt.religion.computers] Perl may be great, but...

tale@pawl.rpi.edu (David C Lawrence) (11/23/89)

In <1989Nov22.153901.3503@splut.conmicro.com> jay@splut.conmicro.com
(Jay Maynard) writes: 
Jay> I'm watching the world do great things in perl, and I'm jealous, knowing
Jay> that Larry has succumbed to the Richard Stallman Syndrome: "That's not a
Jay> real computer, and I won't program to it."  Richard is as blatant about
Jay> it as he is about the GNU Manifesto's real objectives.  I don't really
Jay> think Larry has it in for 16-bit machines, but then again, perl could
Jay> have been written to avoid the more obvious limitations...as it is now,
Jay> perl crashes and burns spectacularly.

This is actually pretty amusing.  Now it's a pathological disorder to
write good software that runs on a variety of machines, but not all of
them.  Now you might believe that Richard needs treatment for other
things, but I don't think this is one of them.

[Obligatory religious cut: You'd have to ship a few MTS/Plus
programmers that way first.]

Dave
-- 
 (setq mail '("tale@pawl.rpi.edu" "tale@ai.mit.edu" "tale@rpitsmts.bitnet"))

kjones@talos.uucp (Kyle Jones) (11/25/89)

Jay Maynard writes: 
 > I'm watching the world do great things in perl, and I'm jealous, knowing
 > that Larry has succumbed to the Richard Stallman Syndrome: "That's not a
 > real computer, and I won't program to it."  Richard is as blatant about
 > it as he is about the GNU Manifesto's real objectives.  I don't really
 > think Larry has it in for 16-bit machines, but then again, perl could
 > have been written to avoid the more obvious limitations...as it is now,
 > perl crashes and burns spectacularly.

David C Lawrence writes:
 > This is actually pretty amusing.  Now it's a pathological disorder to
 > write good software that runs on a variety of machines, but not all of
 > them.  Now you might believe that Richard needs treatment for other
 > things, but I don't think this is one of them.

I disagree.  Take GNU Emacs as an example.  While Emacs runs on lots of
platforms, no one who's looked at its code would ever dream of calling
it portable.  GNU Emacs has proliferated mainly because of the hard work
of those who have ported it to other systems, not because it was
originally written with portability in mind.

Now, while I think it's wonderful that Emacs has a strong enough following
that such jackleg maintenance is possible, this does not change the fact
that it is lousy software engineering practice to write systems that *have*
to be maintained this way.