[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Compiling emacs-20.7 with gcc 2.7.2.1 on m64k-next-nextstep4
From: |
Eli Zaretskii |
Subject: |
Re: Compiling emacs-20.7 with gcc 2.7.2.1 on m64k-next-nextstep4 |
Date: |
Thu, 14 Jun 2001 11:37:53 +0300 (IDT) |
On Thu, 14 Jun 2001, Pascal Bourguignon wrote:
> I solved the problem by replacing macros by functions.
Thanks for the info. I will put this into etc/PROBLEMS.
> As I said, I think that the usage of macros done in the source of
> emacs is preposterous. Most of them should be replaced by
> functions. Actually macros should almost never be used. They're just a
> kludge to paliate deficiencies of C. No high level language has such
> macros.
IIRC, in Emacs sources, macros are mostly used for efficiency, or to
hide implementation of data types in case it will change in the future.
> The only place where macros could be allowed, _was_ to define
> constants (now we should use "const int cst=1;"),
Emacs still supports old compilers which don't know about `const'.
> Well, and to
> protect from multiple inclusions of headers, since RMS can't
> understand that it's better to use #import (without outputing a
> nagging warning) instead of #include...
"#import"? Isn't that highly unportable?