bug#23425: master branch: `message' wrongly corrupts ' to curly quote.

[Alan Mackenzie <acm@muc.de>]

Hello, Eli.

On Wed, May 04, 2016 at 08:03:43PM +0300, Eli Zaretskii wrote:
> > Date: Wed, 4 May 2016 16:48:49 +0000
> > Cc: eggert@cs.ucla.edu, 23425@debbugs.gnu.org
> > From: Alan Mackenzie <acm@muc.de>

> > > > That theory _is_ wrong, certainly amongst Emacs developers.  Even in this
> > > > bug thread, two other people have expressed negative sentiments about
> > > > this matter.

> > > So we have 3 people against that.  Is it really enough?

> > That's three people in this thread.  Over the months, more people have
> > expressed unhappinness about curly quotes, in the various ways they have
> > been imposed on us.  Yes, I think that is enough.

> I don't.  Please let's wait until 25.1 is released, the only way to
> change this decision is if we see a lot of complaints about this.

There won't be lots of complaints about this.  There will be a few people
who notice, and some of them will utterly hate it, just as I do.  Some
will have their Lisp rendered unworkable, just as happened to me.  There
are thousands of options in Emacs, and most users care about only a small
subset (although the subset is different for each user).  You could
remove any option, no matter how "important", and you'd get only a small
number of complaints.  This doesn't seem to me like the way to judge
things.

> > The comments from the other two people have been less than fully
> > respectful, and at least one of them is not known for making snide
> > remarks.  What does that tell you?

> That some people don't like this.  But we knew that back when it was
> decided to have this feature.

And when was this decision taken?  The answer is it wasn't.  Despite the
great opposition to it, Paul plowed ahead with several varied schemes for
curly quotes, not waiting for consensus first, and in the end, people
just got tired of opposing him.

> > Put frankly, the curly quote stuff was installed into Emacs without there
> > being any consensus in favour of it, and was deliberately done in such a
> > way that it could not be disabled (see below).

> Not true.  If there were no consensus, at least not a wide enough one,
> this wouldn't have been admitted.

That's not the way the Emacs project works.  Most contributors work on a
cooperative basis and first discuss contentious things and abide by
understandings reached.  You and I certainly do.  That leaves us open to
abuse by aggressive committers, who don't abide by the above convention.
This is what happened with curly quotes.  There is nobody in the project
who feels empowered to revert "premature" commits.

> The objections were not numerous enough and strong enough to prevent
> these changes.

The objections to curly quotes were extremely numerous (I've spent some
time today looking at old threads), and support for them was sparse.

[ .... ]

> > If it is too late to go into 25.1, can we put a fix into 25.2?

> I see no reason not to install a change along these lines on master,
> if the new value is not the default.

We'll see what we can manage.

> > And yet again, I suggest `text-quoting-style' should become a
> > configurable option, giving users the ability to chose how THEY want
> > messages to be displayed.

> And I yet again urge you to wait until 25.1 hits the streets, and we
> see what the Emacs community at large thinks about this.

-- 
Alan Mackenzie (Nuremberg, Germany).