Doug and I talked about this off line. Doug predates all versions of
roff, he watched it being developed and used it. I think his opinion
matters.
In the message below the "Am I wrong wanting" and the specs are me,
his response is below that.
Anyone arguing for \sDD is just misguided. If you want two things
it is \s(12, if you want N things, groff gave you \s[1234].
Could we please just converge on this spec? I'm old and tired but
if we can get agreement on this and noone wants to do the code I'll
take a swing at it.
----- Forwarded message from Doug McIlroy <d...@cs.dartmouth.edu> -----
Date: Fri, 03 Apr 2020 18:36:39 -0400
From: Doug McIlroy <d...@cs.dartmouth.edu>
To: l...@mcvoy.com
Subject: Re: weird \s
Am I wrong in wanting
\sDwhatever - set size to D and print whatever
\s(DDwhatever - set size to DD and print whatever
\s[DD]whatever - set size to DD and print whatever (groffism)
That's exactly what the man page says--but not what texinfo says.
I am all for it. The current state, in case you missed it is
that \sDD sets the size to DD, provided the first D is one of 1,2,3.
Otherwise the second D is part of "whatever". This special pleading
was (barely) plausible for the original C/A/T phototypesetter,
which could not produce sizes greater than 36.
----- End forwarded message -----
--
---
Larry McVoy lm at mcvoy.com
http://www.mcvoy.com/lm
