Path: utzoo!mnetor!uunet!husc6!bloom-beacon!gatech!mcnc!decvax!decwrl!labrea!agate!gershwin.berkeley.edu!cswarren
From: cswarren@gershwin.berkeley.edu (Warren Gish;133B Biochem;x3-9219)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Macintosh 'C' Compilers
Message-ID: <8548@agate.BERKELEY.EDU>
Date: 10 Apr 88 09:26:04 GMT
References: <299@polygon.UUCP> <23561@ucbvax.BERKELEY.EDU>
Sender: usenet@agate.BERKELEY.EDU
Reply-To: cswarren@gershwin.berkeley.edu.UUCP (Warren Gish)
Distribution: na
Organization: University of California, Berkeley
Lines: 21
Keywords: C,Macintosh,compiler

In article <23561@ucbvax.BERKELEY.EDU> oster@dewey.soe.berkeley.edu.UUCP (David Phillip Oster) writes:
>2.) porting from Microsoft to the Mac is easier in LightSpeed C, than it
>is in MPW, because Microsoft, the Mac O.S, and LightSpeed C all assume
>that an int is 16 bits. MPW assumes 32, and it is a colossal pain.
No quarrel here, but I find it much more difficult to port code from a
32-bit-int system to a 16-bit-int system than vice versa.  Not only is
there a size change, but there is a loss of precision to be concerned
about.  As an added headache, the micro usually has less memory.

In my opinion, Apple took the far-sided view in establishing 32-bit
ints.  It's the ``natural'' size for the MC68020 and beyond, to paraphrase
K&R.  Having done most of my work in the last few years on VAXen and Suns,
this aspect of MPW C is a breath of fresh air in the micro world.

Do I hear 3 cheers from programmers for OS/2 on the Intel 286? :-)

Warren Gish
IS&T
UC Berkeley
cswarren@violet.berkeley.edu
cswarren@ucbviole.bitnet