[bug #64440] [docs] minor corrections and updates

[Dave]

Follow-up Comment #5, bug #64440 (project groff):

[comment #4 comment #4:]
Hm, your working copy changes:

"Set the sizes of spaces... in twelfths of font's space width"

to

"Set the sizes of spaces... in twelfths of an em in @code{nroff} mode and
36ths of an em in @code{troff} mode."

But the original wording was accurate, and the revised is not; see the test
case in http://lists.gnu.org/r/groff/2023-05/msg00063.html (and for further
evidence, change TR's "spacewidth" parameter to different values and note how
this affects .ss's operation).

CSTR#54 said the units were 1/36m, but your own post in the email thread
(http://lists.gnu.org/r/groff/2023-05/msg00060.html) exposed that this wasn't
even true of "device-independent AT&T troff fonts in general," let alone of
groff.  This is where CSTR#54's error lies, and where groff's manual was
correct.

Your Texinfo-manual edit also adds: "The @acronym{AT&T} @code{troff} manual
misleadingly imples that the arguments to the @code{ss} request twelfths of an
em even on @code{troff} mode devices."

This doesn't appear to accurately reflect CSTR#54, which says that (a) the
units of the .ss parameter are unconditionally 1/36m for troff, and (b) the
request is ignored in nroff.  (It also refers to "arguments" to .ss in
CSTR#54, even though AT&T's .ss took only one argument.)  Nothing--not
CSTR#54, not our current groff docs--floats 1/12m as a unit--nor should
anything, because nothing _uses_ 1/12m.  grotty ends up using 1/12m as its .ss
units purely by the coincidence that this is the same value as 1/12 of the
space width.  But "1/12 of the space width"--as currently documented, and as
verified in that email thread--accurately reflects groff's behavior whether
troff or nroff is used.

Upshot: the groff .ss documentation is currently (barring comment #4's
proposed edit) correct; it is incomplete only in that it's missing the CSTR#54
erratum.  CSTR#54 (seemingly) erred in specifying 1/36m as .ss's units (though
as the email thread also reveals, this is accurate for Heirloom).

Comment #2 is still on my radar but I haven't gotten to it yet.


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?64440>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/