On 28/11/2007, Richard L. Hamilton <rlhamil at smart.net> wrote:
> > On Nov 8, 2007 7:08 PM, Al Hopper
> > <al at logical-approach.com> wrote:
> > > On Thu, 8 Nov 2007, Shawn Walker wrote:
> > >
> > > > On 08/11/2007, Al Hopper
> > <al at logical-approach.com> wrote:
> > > >>
> > > >> Has anyone looked into what it would take to
> > make an indiana
> > > >> compatible pkg (ref http://pkg.opensolaris.org)
> > for the ksh93 shell?
> > > >> Is anyone working on this?
> > > >
> > > > ksh93 is already integrated into ON; so Indiana
> > already has it too.
> > > >
> > > > Are you talking about getting packages for the
> > newest versions going?
> > >
> > > Yes.  I'm interested in exploring how the packaging
> > system would work
> > > in terms of installing and updating a non-trivial
> > application and I
> > > thought that ksh93 would be a good starting point.
> >
> > Don't forget to update /bin/sh to ksh93 if you
> > release an update package.
> >
> > Josh
>
> Why?  /bin/sh doesn't need to be POSIX or anything other than 100.00% 
> backwards compatible.
> All that has to be true for POSIX is that the sh found on the PATH returned 
> by getconf PATH
> is POSIX compliant.  AFAIK, POSIX doesn't even require #! support at all, and 
> if it exists, the
> sh man page at http://www.opengroup.org/onlinepubs/009695399/utilities/sh.html
> recommends using a script to edit scripts that need the POSIX shell name in 
> them at install time.
>
> Assuming anything else regarding how the POSIX compliant shell may be found, 
> or expecting that
> you ought to be able to, is an error.
>

It only makes sense to make ksh93 the replacement for /bin/sh as ksh93
has already been tested to be /sbin/sh I believe. Until someone goes
through all of the hard work that Roland already has for ksh93, I
don't see the point of using any other shell.


-- 
Shawn Walker, Software and Systems Analyst
http://binarycrusader.blogspot.com/

"We don't have enough parallel universes to allow all uses of all
junction types--in the absence of quantum computing the combinatorics
are not in our favor..." --Larry Wall

Reply via email to