delorie.com/archives/browse.cgi   search  
Mail Archives: djgpp/2000/11/12/13:45:26

From: Julian Hsiao <madoka AT novastar DOT com>
Newsgroups: comp.os.msdos.djgpp
Subject: DJGPP and Borland's compiler generated code size
Date: Sat, 11 Nov 2000 16:41:10 -0500
Organization: Carnegie Mellon University
Message-ID: <madoka-6C398A.16411011112000@news.supernews.com>
User-Agent: MT-NewsWatcher/3.0 (PPC)
X-Complaints-To: newsabuse AT supernews DOT com
Lines: 24
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

Hi,

I'm currently taking a programming class and being short on budget, uses 
DJGPP and Borland's free compiler and XEmacs for assignments.  The 
reason I uses both compilers is because I try to avoid using certain 
constructs that only GCC or Borland provides (well, I guess the only way 
to completely avoid that is to code while reading the C++ specs paper, 
but I'll pass on that...).

Either one worked quite well for my purpose (except for some reason, 
violating the const declaration only results in a warning in both 
compilers, but an error in CW, which is what my class uses), but I 
noticed that DJGPP's generated binary size is considerably larger than 
that of Borland compiler.  With DJGPP, I pretty much always get ~200K 
binaries while with Borland's compiler I get ~40K binaries.

Being a fairly introductory class, most of the assignments are trivial 
(implement a priority queue using a linked list, etc.), and no STL is 
used.  I turn on optimization when compiling with both compilers.

Can someone please explain why this is the case?

Julian Hsiao
madoka AT novastar DOT com

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019