Path: utzoo!attcan!uunet!lll-winken!ames!mailrus!cornell!uw-beaver!uw-june!uw-entropy!dataio!bright
From: bright@Data-IO.COM (Walter Bright)
Newsgroups: comp.lang.c++
Subject: Complexity of C++ vs C
Message-ID: <1820@dataio.Data-IO.COM>
Date: 13 Jan 89 19:37:27 GMT
References: <6578@pogo.GPID.TEK.COM> <6590084@hplsla.HP.COM> <1360@cod.NOSC.MIL>
Reply-To: bright@dataio.Data-IO.COM (Walter Bright)
Organization: Data I/O Corporation; Redmond, WA
Lines: 23

In article <6590084@hplsla.HP.COM> jima@hplsla.HP.COM (Jim Adcock) writes:
>> Excuse me,
>I think a lot of people have been saying that a c++ compiler
>is a lot more difficult beast than a c compiler, but I'm not
>sure that's true.

I approached writing a C++ compiler with the idea that it was just a
few simple extensions to C, and that it would take about 3 months.
Boy, was I wrong!

If it was so easy, there'd be a lot of competition out there. There are
lots of C compilers for MS-DOS (probably > 20), but 1 C++ translator (a port
of AT&T's), and one C++ compiler (Zortech's). There would be a Borland C++,
a Microsoft C++, a Watcom C++, etc.

Doing a C++ is about the same level of difficulty as writing a global flow
analysis optimizer. I've been observing the industry for many years, and
adding a real optimizer to an existing debugged compiler takes about 1 year
(regardless of the number of people working on it!).

I would estimate that it will take a company one year to add C++ to an
existing C compiler. Add time to this if managers put more than two
people working on it.