On Sat, Feb 23, 2019 at 04:51:19PM +0000, Jason McIntyre wrote:
> On Sat, Feb 23, 2019 at 05:29:31PM +0100, Ingo Schwarze wrote:
> > Hi,
> > 
> > Anthony J. Bentley wrote on Wed, Jan 30, 2019 at 04:01:30AM -0700:
> > > Jason McIntyre writes:
> > 
> > >> in the man page you have used Sq. that will make it mark up the same as
> > >> it already does:
> > >>
> > >>  (`*')
> > 
> > > There's a difference: Ted uses a UTF-8 locale where Sq shows up as
> > > pretty Unicode quotes. Literally typing `' doesn't (not in the terminal,
> > > anyway).
> > > 
> > > mandoc should render Sq as ' ' in ASCII mode. We fixed Dq the same way
> > > in both mandoc and groff a couple of years ago (`` '' to " ").
> > 
> > To finally conclude this thread: after a thorough discussion on the
> > list <groff at gnu dot org>, this patch, and the corresponding patch
> > to groff, has been rejected.  While some liked our arguments of
> > making output nicer with Unicode-copatible fonts and avoiding ASCII
> > usage that conflicts with Unicode, no consensus could be reached
> > because substantial objections were raised, too.
> > 
> > The main objection is that using 0x60 for the opening single quote
> > (as an alternative to using it for a grave accent) and 0x27 for the
> > closing single quote (as an alternative to using it for an apostrophe),
> > according to the latest version of the ASCII standard, is a valid
> > way to use the ASCII character encoding and has been valid since
> > these bytes were frist assigned meaning in the 1965 edition of ASCII -
> > even though this usage is no longer compatible with ISO-LATIN nor with
> > Unicode.
> > 
> > Moreover, it is the traditional way in which *roff used ASCII, it
> > is compatible with the way many traditional terminals (including
> > DEC VT-100 and VT-220) display these ASCII characters, and several
> > groff developers are still using fonts for ASCII display that display
> > ` ' as symmetric single quotes.  Even our own Gallant console font
> > does that.  Several groff developers dislike the idea of changing
> > established groff behaviour, even for the somewhat old-fashioned
> > ASCII output device, unless there are strong technical reasons, and
> > i had to concede that this change would be more of the "nice to
> > have" kind and isn't strongly required for anything.
> > 
> > The case Dq -> ``/" is substantially different because in ASCII,
> > double quotes - both opening and closing - were always clearly
> > defined as ".
> > 
> > 
> > Consequently, people using a modern Unicode-compatible font for
> > ASCII output will have to live with the slightly ugly single quotes,
> > likely for good.
> > 
> > Needless to say, if you can use the -T utf8 output device, for
> > example by running with LC_CTYPE=en_US.UTF-8, the issue doesn't
> > occur in the first place, and you already get nice single quotes
> > right now, without needing any code changes.
> > 
> > Yours,
> >   Ingo
> > 
> 
> thanks for the follow up.
> 
> i personally don;t mind the wonkiness of how Sq marks up, but maybe
> you'd like to mail a diff for anyone that really wants '' to use
> themselves?
> 
> jmc
> 

doh, the diff's already there! sorry!
jmc

> > 
> > > Index: usr.bin/mandoc/chars.c
> > > ===================================================================
> > > RCS file: /cvs/src/usr.bin/mandoc/chars.c,v
> > > retrieving revision 1.48
> > > diff -u -p -r1.48 chars.c
> > > --- usr.bin/mandoc/chars.c        15 Dec 2018 19:30:19 -0000      1.48
> > > +++ usr.bin/mandoc/chars.c        30 Jan 2019 10:55:08 -0000
> > > @@ -100,7 +100,7 @@ static struct ln lines[] = {
> > >   { "rq",                 "\"",           0x201d  },
> > >   { "Lq",                 "\"",           0x201c  },
> > >   { "Rq",                 "\"",           0x201d  },
> > > - { "oq",                 "`",            0x2018  },
> > > + { "oq",                 "\'",           0x2018  },
> > >   { "cq",                 "\'",           0x2019  },
> > >   { "aq",                 "\'",           0x0027  },
> > >   { "dq",                 "\"",           0x0022  },
> > 
> 

Reply via email to