*BSD News Article 50455


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!pravda.aa.msen.com!cssun.mathcs.emory.edu!emory!gatech!howland.reston.ans.net!agate!agate!matt
From: matt@godzilla.EECS.Berkeley.EDU (Matt Austern)
Newsgroups: misc.jobs.offered,comp.lang.c,comp.lang.c++,comp.object,comp.lang.eiffel,comp.lang.misc,comp.unix.bsd.386bsd.misc
Subject: Re: Beginner to C/C++ looking for some good books
Followup-To: comp.lang.misc
Date: 01 Sep 1995 18:52:01 GMT
Organization: University of California at Berkeley (computational neuroscience)
Lines: 44
Message-ID: <MATT.95Sep1115201@godzilla.EECS.Berkeley.EDU>
References: <40b4i2$egf@ixnews5.ix.netcom.com> <brockmanDDLJDL.2IJ@netcom.com>
	<41alha$b1g@cisunix1.dfci.harvard.edu> <brockmanDE5u28.5Lr@netcom.com>
	<19950831T090632Z@naggum.no> <427eaj$1ri@ucsbuxb.ucsb.edu>
Reply-To: matt@physics.berkeley.edu
NNTP-Posting-Host: godzilla.eecs.berkeley.edu
In-reply-to: loren@beauty.ucsb.edu's message of 1 Sep 1995 17:02:11 GMT
Xref: euryale.cc.adfa.oz.au misc.jobs.offered:205012 comp.lang.c:112123 comp.lang.c++:123025 comp.object:31435 comp.lang.eiffel:9164 comp.lang.misc:17256 comp.unix.bsd.386bsd.misc:165

[Please note the crossposting and the followup.  I have redirected 
followups to comp.lang.misc.]

In article <427eaj$1ri@ucsbuxb.ucsb.edu> loren@beauty.ucsb.edu (Loren Koss) writes:

> Okay, here is where I am going to have to be almost rude and say "are you 
> a bonehead?"  Perl is 0% portable.  Why write a program that can only 
> work Unix machines?  I've never heard of a Windows Perl interpretor.

I have.  At least I've used an MS-DOS Perl implementation, and I know
of no reason it wouldn't run under Windows.  Perl has also been ported
to OS/2, and I strongly suspect that it either has been or will soon
be ported to NT and VMS.

> And 
> that brings up point number two.  Why write a program in an interpreted 
> language?  Interpreted languages are slow, make bad use of memory, and 
> often tie up useful resources while going through its lexical analyzer.  

Again: all other things being equal, an interpreted program will be
slower than a compiled program.  But all other things never are equal!
If your program spends most of its time making calls to functions in a
a regexp library then the interpreted versus compiled issue will be
more or less irrelevant (Amdahl's Law, anyone?), and the main
efficiency issue will be the quality of that library.  In cases like
that, Perl may well be an efficiency win.  Note also that run-time
speed doesn't always matter terribly much.  I care about speed when I
write (say) Monte Carlo programs, or large-scale neural simulators,
but I also write lots of programs where I don't care about it.  And,
finally, note that there are many programming techniques that tend to
be easier if you're using an interpreter than if you're using a
compiler.

> BASICALLY, C++ is language of choice.  If you get confused programming in 
> it, drop down to C.

And this is what I strongly disagree with.  I don't believe in the
concept of a "language of choice": I believe in using the right tool
for the right job.  I use C++, Perl, and several other languages; I
don't believe there is any one language that's best for all purposes.
-- 
  Matt Austern                             He showed his lower teeth.  "We 
  matt@physics.berkeley.edu                all have flaws," he said, "and 
  http://dogbert.lbl.gov/~matt             mine is being wicked."