>Date: Wed, 18 Oct 2006 15:21:31 -1000 (HST)
>From: Joseph Kowalski <Joseph.Kowalski at eng.sun.com>
>
>> From: Richard Lowe <richlowe at richlowe.net>
>...
>> (I do still harbour some concern regarding any future ksh(1) or sh(1)
>> migration, however, but that's another case).
>
>Very interesting question.
>
>Don (as in Cragun), would having a default edit mode for ksh93 prevent
>it from eventually becoming the SUS conforming ksh?
No. He would not. There is a difference between having a
configuration file which can be adjusted by your system administrator
to set a default editing mode (OK according to the standard), and
having ksh93 always set a default editing mode even if no configuration
files are present on the system (not allowed by the standard). The
standard, however, only talks about what happens when the shell is
named "sh"; the standard says nothing about a shell named "ksh93".
This case is about "ksh93"; not "sh".
The name(s) of the configuration file(s) will change when (if) ksh93
becomes /usr/bin/sh or /usr/xpg4/bin/sh such that they will be
compatible with the standard. My understanding is that ksh93 already
does this if the last component of the pathname where it is installed
is "sh". But, again, that is not this case.
- Don (Or in this case: Donald W. Cragun, Chair of the IEEE Portable
Applications Standards Committee's Shell & Utilities Working
Group.)
>
>- jek3
