Re: bug in pwd POSIX-compliance

[Chet Ramey]

Eric Blake wrote:

> On further research, the requirement for `pwd -P' to set PWD was intentional, 
> and not a mistake.  See POSIX XRAT A.3 
> (http://www.opengroup.org/onlinepubs/009695399/xrat/xbd_chap03.html) under 
> Symbolic Links line 1148, which states:
>
> Historically, several shells have had built-in versions of the pwd utility. In 
> some of these shells, pwd reported the physical path, and in others, the 
> logical path. Implementations of the shell corresponding to IEEE Std 1003.1-
> 2001 must report the logical path by default. Earlier versions of IEEE Std 
> 1003.1-2001 did not require the pwd utility to be a built-in utility. Now that 
> pwd is required to set an environment variable in the current shell execution 
> environment, it must be a built-in utility.
>
> Also, see the followup mails to my defect report, as posted on the Austin 
> mailing lists: http://www.opengroup.org/austin/mailarchives/ag/msg08028.html.  
> Just because no non-compliant shell does it that way is not an excuse for bash 
> to not do it, at least when bash is installed as the compliant sh.

I understand that, but no shell does it.  Period.  Even those claiming
compliance, like bash and ksh93.  It's an artificial requirement.
Certainly it's non-intuitive for a builtin that's not performing any
modifying action to change $PWD as part of its execution.

Don Cragun's reply didn't include any rationale for pwd's behavior. I
suspect there is none.  Even the stated desire for consistency is
not compelling.

If I change bash's behavior in this case, it will be only in posix mode.

Chet
--
``The lyf so short, the craft so long to lerne.'' - Chaucer
( ``Discere est Dolere'' -- chet )
                                                Live...Laugh...Love
Chet Ramey, ITS, CWRU    chet@case.edu    http://cnswww.cns.cwru.edu/~chet/