GNU bug report logs - #20385
Support curved quotes in doc strings

Previous Next

Full log

Message #128 received at 20385 <at> debbugs.gnu.org (full text, mbox):

From: Drew Adams <drew.adams <at> oracle.com>
To: Paul Eggert <eggert <at> cs.ucla.edu>, Nicolas Petton <nicolas <at> petton.fr>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 20385 <at> debbugs.gnu.org,
 Dmitry Gutov <dgutov <at> yandex.ru>
Subject: RE: bug#20385: [PROPOSED PATCH] Support quoting 'like this' in doc
 strings
Date: Thu, 23 Apr 2015 07:51:08 -0700 (PDT)

> > British and American usages tends to be reversed wrt which curly
> > quotes, single or double, are used at the first level:
> 
> Sure, and the tradition in many GNU projects is to use American-
> style quoting for regular English text, and British-style quoting
> for code.  For example, the Emacs 24.5 manual's info files quote
> ordinary text like this:
>    Emacs occupies a “graphical window”
> and quote code like this:
>    A description of an imaginary variable, ‘electric-future-map’.
> 
> Changing doc strings to look like the latter will better implement
> the longstanding style for documentation in GNU projects.  It's what
> quoting `like this' was originally intended to do (and *did* do,
> with circa-1965 draft ASCII).  It's an eminently reasonable way to
> quote code in documentation.

Sorry, but I disagree, whether or not you can claim that it has
been some kind of GNU convention.  It is foolhardy.  Both double
and single curly quotes are used for _ordinary text_ out there in
the real wide world.

Emacs is in that wide world, and increasingly so.  Emacs is a very
general tool, being usable for all sorts of editing and other
activities.  And that includes editing ordinary text.  And ordinary
text that often has embedded code terms.

Sacrificing being able to use two levels of quoting for ordinary
text in favor of stealing one of them for code quoting, just so
that we can wave our hands and say that we respect some old GNU
convention, is quite short-sighted.  Coming from someone who touts
his proposed change as bringing up to "nowadays" and "modern" ways,
I should be surprised.  But I'm not.

In short, Emacs should have all three: single and double curly
quotes for their usual uses with ordinary text, and some reasonable
and distinctive way to quote code that is embedded in ordinary text.

If this were HTML or XML (or even JSON) then we would not be trying
to sacrifice one of the first two for the third.  We would simply
adopt an unmistakable tag for code quoting.

My vote is (clearly) to keep `...' for quoting code in Info,
doc strings, Emacs-Lisp mode,...  It is simple, easy-to-use,
and versatile, and it has proven itself for at least 4 decades.

So far, we have heard exactly ONE (weak) argument against it:
you find it ugly.  I'd say prettify it in your own environment,
if it really bothers you, but otherwise leave it alone.  Let's
not sacrifice the usefulness of this proven convention for
someone's purely cosmetic concerns.

Or was there another argument in favor of your proposed change
that I missed?  I've asked several times now, but have gotten
no reply to the question.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.