Subject: Re: lib/29980: the ctype.h functions are not documented correctly
To: None <lib-bug-people@netbsd.org, gnats-admin@netbsd.org,>
From: Klaus Klein <kleink@mibh.de>
List: netbsd-bugs
Date: 04/16/2005 14:30:01
The following reply was made to PR lib/29980; it has been noted by GNATS.
From: Klaus Klein <kleink@mibh.de>
To: gnats-bugs@netbsd.org
Cc:
Subject: Re: lib/29980: the ctype.h functions are not documented correctly
Date: Sat, 16 Apr 2005 16:28:33 +0200
Martin Husemann wrote:
> I disagree. There is a big CAVEAT section - that should be enough.
>
> This restriction is no suprise to any C programmer, the standard defines it
> this way, and gcc most of the times warns properly (now that we got rid
> of the casts in the header files).
My take on this: Personally, I think the current CAVEATS section is
fine for that purpose (but then, I can hardly avoid wearing a stdc
hat), and I also feel that pulling it right into the very first
paragraph (as originally proposed by Roland) would convolute the
text. How do you all feel about making this an extra paragraph
in DESCRIPTION, in the spirit of the POSIX edition?
Independently from the placement, it would be appropriate to change
the current CAVEATS text from "[undefined] result" to "[undefined]
behavior" (Roland: Note the A.E. spelling).
- Klaus