Re: POSIX misunderstanding

[Albert Cahalan]

On Wed, 2004-08-18 at 11:44, Paul Eggert wrote:
> Albert Cahalan <address@hidden> writes:
> 
> > a POSIX-conforming "head" program may
> > support a "-1" option. It may also support a
> > "-2" option, and so on.
> 
> That violates POSIX Utility Syntax Guidelines 3 and 11.  See:
> 
> http://www.opengroup.org/onlinepubs/000095399/basedefs/xbd_chap12.html#tag_12_02

Well, so does the --lines option.

I think "violates" is too strong of a word for
anything called "Guidelines", but anyway...

Guideline 11 isn't even a problem at all, unless you
artificially make it a problem. It's only number 3
that is of concern. (see "--lines" though)

You missed this bit at the bottom of the page:

   "On some implementations, the utilities accept
   usage in violation of these guidelines for
   backwards-compatibility as well as accepting
   the required form."

Well, there you go. The POSIX and UNIX standard
explicitly allows for the traditional behavior.

Please fix. You wouldn't violate anything, and
anyway, GNU's Not UNIX. :-)

> Most likely the underlying problem is that your distribution has
> decided to set _POSIX2_VERSION to 200112 in <unistd.h>.  This
> indicates that the distribution's maintainers want conformance to
> POSIX 1003.1-2001 as specified above.  To get the older behavior that
> you prefer, you can set _POSIX2_VERSION=199209 in your environment, or
> build coreutils with "configure DEFAULT_POSIX2_VERSION=199209".

Well, of course. Nearly everybody wants conformance.
We also want, to the maximum extent possible, all of
the historical behavior. You have to kill the historical
behavior when, and only when, it is in direct conflict
with the standard.

So if the standard defined a "-2" option that would
print the output twice, then yeah, you'd have to
do that. Standards conformance comes first, generally.
(you could still support "-3", but that's silly)