Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!umd5!purdue!gatech!ncsuvx!csclea!wolf
From: wolf@csclea.ncsu.edu (Thomas Wolf)
Newsgroups: comp.sys.atari.st
Subject: Mark Williams C
Message-ID: <1442@ncsuvx.ncsu.edu>
Date: 2 Feb 88 23:08:52 GMT
References: <8802021829.AA25954@ucbvax.Berkeley.EDU>
Sender: nntp@ncsuvx.ncsu.edu
Reply-To: wolf@csclea.UUCP (Thomas Wolf)
Organization: Computer Science , NCSU, Raleigh NC
Lines: 26


It seems that Mark Williams C is one of the more popular C compilers for the
ST.  Well, I've had my copy (an early v1.2) for more than a year now and am
ALMOST happy with it:  I can't stand the speed (or lack thereof) of the
compile/link phase.  Even after optimization (letting the compiler put temp
files in ram-disk), my hard-disk based system still took over 5 minutes to
'make' a 3-module (total < 1200 lines) program.  That's an average compile/link
time of 220 lines/minute!  MW-Quick C and Turbo-C for the PC's supposedly
compile/link 10,000+ lines/minute.  Could that be?

Am I just TOO impatient?  After all, it gives me time to fix a cup of tea while
I wait for compiles/links :-)

I didn't bother upgrading to v2.0, since I didn't detect any significant
speed improvements (and the cost was $50.00 -- for an UPGRADE!) when I tried it
out on a friends machine.

Could some kind soul post (or send to me) some compile benchmarks for the 
newest MW-C ?  (Is it v2.17?)  If possible, I'd hope to see a benchmark for
APSKEL.PRG -- but even a 'hello world' result would be appreciated.  Perhaps
this info will help me decide whether I want to upgrade to the new MW or just
wait for Laser-C by Megamax.

Thanks in advance for any info.

Tom Wolf