Path: utzoo!attcan!uunet!stu!
From: stu@stu.scs.com (Stu Brown)
Newsgroups: comp.sys.ibm.pc.misc
Subject: Quick C Vs. MS C: results
Keywords: C
Message-ID: <722@stu.scs.com>
Date: 10 Dec 90 15:15:14 GMT
Organization: Mentor Graphics Corp. - Silicon Design Division
Lines: 54

Last week I posted a question about using Quick C & Assembler 2.51
versus Microsoft C 6.0 on a large application, the PCIP tcp/ip
package from MIT. Here is a followup.

I decided to buy the Quick C & Assembler package. I compiled and assembled the
PCIP package with small modifications to makefiles intended for regular
Microsoft C. All of the utilities I needed were included with the
Quick C package: linker, librarian, compiler, assembler, make (called nmake).

I compiled and assembled about 16,500 lines of code in various PCIP
libraries. The only "bug" I found is that the compiler complains
when you call a function which is later declared as static. For example:

main()
{
	...
	myfunc(a);
	...
}

static myfunc(x)
{
...
}

The compiler quits on a fatal error about a changing declaration. This is
fixed by declaring the function at the top of the program:

static myfunc();
main()
{
	...
	myfunc(a);
	...
}

static myfunc(x)
{
...
}

I linked up the ping program and it worked. I have not
tried the larger application, telnet.

As I understand from a comparison sheet that Microsoft puts
out that compares Quick C with regular C, the only significant difference
is the level of optimization available. I don't know if the library support
is different since PCIP relies very little on libraries
(it implements most functions itself). I was very happy to save about $300.
-- 
Stuart Brown
Mentor Graphics Corporation
Silicon Design Division (N.J.)
uunet!sdl!stu