Path: utzoo!mnetor!tmsoft!torsqnt!lethe!yunexus!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!olivea!uunet!shelby!neon!flamingo.Stanford.EDU!espie
From: espie@flamingo.Stanford.EDU (Marc Espie)
Newsgroups: comp.sys.amiga.programmer
Subject: Re: SAS Include FILES from HELL, film at 11pm.
Message-ID: <1991Feb7.220346.29943@Neon.Stanford.EDU>
Date: 7 Feb 91 22:03:46 GMT
References: <4034@orbit.cts.com> <1991Feb7.192644.25098@jato.jpl.nasa.gov>
Sender: news@Neon.Stanford.EDU (USENET News System)
Organization: LIENS, ENS, 45 rue d'Ulm, Paris (France)
Lines: 53

In article <1991Feb7.192644.25098@jato.jpl.nasa.gov> jdickson@jato.Jpl.Nasa.Gov (Jeff Dickson) writes:
>In article <4034@orbit.cts.com> chucks@pnet51.orb.mn.org (Erik Funkenbusch) writes:
>>
>>manes@vger.nsu.edu ((Mark D. Manes) writes:
>>>I recently purchased SAS/C and kissed goodbye the MANX C compiler that 
>>>I have used for years.
>>
>>I'm interested in why you switched to SAS.  I have been using Manx for years
>>as well, and i get nausea when i try and use SAS.  I was just wondering about
>>your reasons.
>>> 
[sigs deleted]
>
>	YEAH! RIGHT ON. I use MANX too. I have ever since 1985. I don't
>like having to conform to ANSI. Unfortunatly, MANX rid 16 bit ints and
>adopted the ANSI way of doing things. Wish they could release a new version
>in the spirit of 3.6 for us non comformists. Until then, I'm going to continue
>using 3.6 and my 5.x updates will just collect dust.
>
>					Jeff

Ok, this is a flame (kind of). I don't like the idea that I chose to
``have to conform to ANSI''.
What do you mean ``having to conform to ANSI'' ? ANSI is NOT something you
have to conform to. It just means you CAN do stricter type-checking. You
don't usually have to. The SAS/C compiler mostly returns warnings if
you indulge in type-punning, you can ignore them if you want, you can
even turn them off. On the other hand, you get the benefit of having
prototypes, which means you don't HAVE to use dumb castings the like of
sqrt((double)2) or (worse) float x = 2.57; y = sin((double)x);
You don't spend hours on stupid bugs because the compiler tells you when
you have done things wrong (when you INTEND to do things the wrong way,
you still CAN, but you'll get warned)---and don't tell me you don't indulge
in stupid bugs from time to time.
Spare us the macho part about ``real programmers don't need prototypes. Or
I've got the right to be non-conformist.''
Did it ever occur to you that everybody (UNIX too) is switching to ANSI C ?
There might be a good reason to that... Doesn't it bother you that your
source code will soon be a fossile, and is much less portable than ANSI-like
C right now ?

BTW, SAS/C supports 16 bit integers, as well as several models of floating
point. And there is something ANSI provides which traditional C has not:
if you use only single precision floating point variables, computations
will get done in single precision, instead of converting everything to
double first; the old traditional C efficiency-loophole.

You want to make us ANSI-C programmers feel like we wear a straightjacket.
This is wrong. ANSI-C means freedom. Traditional C means a high dependency
on the machine you use. Granted, you can get rid of such dependencies,
but in that case you don't need the dubious ``benefits'' of ANSI-C.
--
    Marc Espie (espie@flamingo.stanford.edu)