Re: [groff] modernize -T ascii rendering of opening single quote

[Ingo Schwarze]

Hi Jeff,

you seem to be reading too much into various sources, but fortunately,
it only tangentially affects what matters to the proposed patch:
there can be little doubt that fonts more commonly show 0x60 as a
grave accent today, and likely also in the past, and that that
practice better matches most (in particular international) standards.


Jeff Conrad wrote on Fri, Feb 08, 2019 at 01:20:09PM +0000:
> On Friday, February 8, 2019, Ingo Schwarze wrote
>> Jeff Conrad wrote:

>>> I think "historic" is pretty context dependent.  As nearly as I can
>>> tell, ASA/ANSI X3.4 has called for 0x60 to encode "accent grave".

>> Absolutely not:
>> 
>>   http://worldpowersystems.com/ARCHIVE/codes/X3.4-1963/page5.JPG
>> 
>> In that standard, 0x60 was still unassigned, and in the next version,
>> the ambiguity is already stated.

> I should have said "has long called for"; it looks to me like this was
> largely resolved by 1967:
> 
>     http://worldpowersystems.com/ARCHIVE/codes/#ASCII-1967

First off, that's not authoritative but a random commentary by a
random guy.  Besides, neither the text nor the table captioned
"ASCII diacritical marks" on the right hand side say that using
0x60 as an opening quote was deprecated; the table can hardly be
interpreted that way because clearly the double quote was *not*
redefined to mean a diaresis, and the comma was not redefined to
mean a cedilla.

> This is consistent with
> 
> http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.96.678&rep=rep1&type=pdf
> 
> (see Figure 70, p. 28).

While the Fisher paper appears to be a respectable source, it is
also not authoritative.  Besides, i could find nothing in the
surrounding text making any statement about the intended use of
0x60.  The mere shape of the glyph in the table is obviously not
saying anything.

> It's also consistent with ANSI X3.4-1986:
> 
>     www.columbia.edu/kermit/ascii.html

Again, not authoritative, but an interpretation by a random person.

I failed to find the actual X3.4-1967 standard document online,
but i see no reason to assume that it was different from later
versions with respect to 0x60, and later versions allowed *both*
usages of 0x60, as shown in one of my earlier mails.

>> That's why i qualified "traditional fonts" with "that provide",
>> implying that modern (i.e.  Unicode-compatible) fonts always provide
>> an "accent grave" at that code point.

[...]
> Perhaps it's better to just say that some implementations used 0x60 for
> an opening quote, while others used it for accent grave - without trying
> to assign a time frame.  The ratio at any point in time would seem tough
> to determine; I'm not sure what practical benefit would derive from
> having a definitive answer anyway.
> 
> The relevant question would seem what the ratio is today.  I suspect
> 0x60 is overwhelmingly shown as accent grave.  Moreover, as Ingo
> mentioned, the neutral single quotes are the proper match for neutral
> double quotes.
[...]

That part is probably accurate.

Showing 0x60 as a grave accent is clearly more common today and
only that practice is compatible with Unicode.  

Jukka Korpela quotes a private communication from Eric Fisher
  http://jkorpela.fi/latin1/ascii-hist.html#60
indicating that showing 0x60 as a grave accent may have *always*
been more common except possibly in the US.

Yours,
  Ingo