From keld@dkuug.dk Fri Oct  4 23:43:57 1991
Received: by dkuug.dk (5.64+/8+bit/IDA-1.2.8)
	id AA10484; Fri, 4 Oct 91 23:43:57 +0100
Date: Fri, 4 Oct 91 23:43:57 +0100
From: Keld J|rn Simonsen <keld@dkuug.dk>
Message-Id: <9110042243.AA10484@dkuug.dk>
To: dominic@british-national-corpus.oxford.ac.uk, wg15-uk@xopen.co.uk,
        wg15rin@dkuug.dk
Subject: Re: (wg15rin 146) request for "ident" specification part in LC_CTYPE
Cc: donn%com.hp.fc.hpfcrn@hplb.hpl.hp.com, hlj@posix.com
X-Charset: ASCII
X-Char-Esc: 29

Ten days ago, Keld wrote
> I think it is very unfair that people are now trying to stop
> ISO ballotting procedures. DS have not had a chance to officially
> ballot on this ISO subproject since draft 9. 
> Something has gone wrong with the IEEE and ISO syncronisation plan,
> if people are trying to beat up DS to keep quiet on this one.
> And anyway this is also liaison requirements from other ISO WGs.
> 
> An formally these people (which at least include Donn and Dominic)
> cannot do it. So should DS shut up while all other national
> member bodies come and give their ballots?

Since then, there has been a large exchange of private mail between
Donn, Keld and myself.  This has been useful, and has helped all three
of us to clarify our positions, and to distinguish what we really meant
from what others thought we had said.  I think it is a fair summary to
say that

1. We are agreed that nobody was attempting, or should attempt, to
   stifle debate.

2. We are all presenting personal opinions, not national positions.

3. Donn and I believe that, while it would be highly desirable for the
   shell and little languages support a richer character set in variable
   names, we consider that to make this change now would introduce a
   long delay into the production of 9945-2, a standard for which
   there is a pressing need, whether or not this feature is present in
   its initial version.  The feature can and should be added in an
   early addendum.

4. Keld believes the feature to be sufficiently crucial that it should
   form a part of the initial version of the standard, even if this
   introduces a delay.  Any such delay, he believes, would be short.

5. We all agree that such a facility should form a part of 9945-2.  We
   just disagree on when.

6. Keld argues that the proposed change is small conceptually.

7. Donn suggests that the consequences of the change will be large in
   practice, given the current code-base for the shell and tools.

We have also begun to discuss technical aspects of the issue.  Let's
continue to do this in public.  The matter can be discussed formally
under item 4.7 of the recently-circulated Stockholm agenda, 2nd CD of
9945-2.

-- 
Dominic Dunlop