[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: release candidate 5 available
From: |
Ben Pfaff |
Subject: |
Re: release candidate 5 available |
Date: |
Mon, 01 Aug 2005 07:12:14 -0700 |
User-agent: |
Gnus/5.1007 (Gnus v5.10.7) Emacs/21.4 (gnu/linux) |
John Darrington <address@hidden> writes:
> "Be conservative in what you produce; liberal in what you accept."
> Using -ansi in development catches things that are not part of the standard
> and therefore might not be accepted by other compilers.
[...]
> Notwithstanding that, the -ansi flag does go part of the way.
> For example, without -ansi, gcc will merily accept // style comments which
> are not part of the ansi standard (until recently).
But it makes lots of code not work at all. It broke off_t on
Solaris and made -Wformat more or less useless even on GNU/Linux.
Autoconf intentionally never chooses to use the -ansi option.
Furthermore, the way we were adding -ansi only at build time, not
at configure time, is a bad idea. The -ansi flag turns off
plenty of features provided by the compiler and headers, but
"configure" had already checked that those features were there.
Thus, inconsistencies arose.
I don't think -ansi is appropriate for a program that wants to
use compiler and system extensions.
--
"Unix... is not so much a product
as it is a painstakingly compiled oral history
of the hacker subculture."
--Neal Stephenson