[bug #57618] man/groff_char.7.man: page needs an overhaul

[Dave]

Follow-up Comment #7, bug #57618 (project groff):

It looks like most if not all the problems Branden identified in comment #1
have been addressed.  The page looks great now!

But of course I still have complaints.

(I realize this man page is very much a work in progress, so some of the below
observations may be addressed in changes that have yet to be committed.)

----

*AGL*

I still advocate for removing the AGL (formerly PostScript) column.  In
addition to the reasons I previously gave: with this man page now having a
convenient link to the AGL, and with both documents including the Unicode
value for every entry, that value becomes an easy cross-reference for those
very few users who do need to look up an Adobe name.

If the column is retained, something on the page should tell users what they
can do with this information.

What are the arguments for moving this information to grops(1) (as proposed in
comment #5) instead of removing it altogether?


----

*hyphenation-inhibition overkill*

While no harm arises from putting \% in front of things like "Latin-1" and
"UTF-8," they accomplish nothing: groff already won't break a line at a hyphen
with only numbers on one side of it.


----

*intro text*

In most subsections, if there is text before the table, it gives an overview
of that section or the data tabulated in the following table.  But a couple of
sections (e.g., "Supplementary Latin letters," "Logical symbols") lead off
with minutiae that might be better placed below the table rather than above
it.


----

*NFD*

(section "Special character escape forms"): "groff requires NFD (Normalization
Form D)"

In what sense is this true?  Groff doesn't take Unicode input directly at all,
but it does take Latin-1, the bulk of which's character set consists of
precomposed characters that in Unicode would be considered NFC.

And for Unicode input passed through preconv, groff seems happy with NFC but
not with NFD:


$ printf "co\xF6perate\n" | uconv -f latin1 -x Any-NFC | groff -Kutf8 >
/dev/null
$ printf "co\xF6perate\n" | uconv -f latin1 -x Any-NFD | groff -Kutf8 >
/dev/null
<standard input>:1: warning: can't find special character `u0308'
$


groff does handle composed characters given in \[uNNNN_NNNN] form:


$ echo 'co\[u006F_0308]perate' | groff > /dev/null
$ 


but this is not what preconv emits.

----

*non-printing characters*

The text refers to NBSP and SHY as "non-printing characters," but this doesn't
seem to be how they are classified (if ISO_8859-1(7) and Perl's Unicode
library are to be believed):


$ perl -e '$_ = "\N{NO-BREAK SPACE}\N{SOFT HYPHEN}"; s/[[:print:]]/X/g; print
"$_\n"'
XX
$ 


----

*verbiage*

(section "Rules and lines"): "Note that both the AGL and the Unicode-derived
names of these three glyphs are rough approximations."

0 ...as opposed to precise approximations?
0 The phrase "Note that..." communicates nothing.  Just tell the reader the
thing you want her to note, and she will note it.

----

*hyphens*

"Frequently-used glyphs" should have no hyphen.

"Playing card symbols" should arguably have one.

    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?57618>

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