Xref: news-dnh.mv.net comp.os.msdos.djgpp:691
Newsgroups: comp.os.msdos.djgpp
Path: news-dnh.mv.net!mv!news.sprintlink.net!howland.reston.ans.net!EU.net!sun4nl!phcoms4.seri.philips.nl!newssvr!kunst
From: kunst AT prl DOT philips DOT nl (Pieter Kunst)
Subject: Re: size of A.OUT files
Sender: news AT prl DOT philips DOT nl (USENET News System)
Organization: Philips Research Laboratories, Eindhoven, The Netherlands
References: <DAu0nF DOT 4It AT jade DOT mv DOT net>
Date: Wed, 28 Jun 1995 09:20:20 GMT
Lines: 17
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Dj-Gateway: from newsgroup comp.os.msdos.djgpp

In article <DAu0nF DOT 4It AT jade DOT mv DOT net> "A.Appleyard" <A DOT APPLEYARD AT fs2 DOT mt DOT umist DOT ac DOT uk> writes:
>  I have a big Gnu C++ program (an Emacs text editor) that I wrote. I make its
>a.out into an .exe file by prepending go32.exe . If compiled without -g, the
>resulting .exe file is 520953 bytes long, but when I call the DOS command `MEM
>/C' from inside the program it says that the program as loaded (presumably
>including go32.exe) occupies only 131584 bytes of conventional memory (and no
>upper memory). (My go32.exe file is 78810 bytes long.) If thus only 25% of the
>.exe file is binary program image, what occupies the remaining 75%?

Did you strip the executable (or link with '-s') ?
If not, then a large part of the executable file will be
the symbol table (for debugging purposes).

Try: strip a.out

Pieter.