Hi Jeremy,

At 2021-10-18T13:05:25+0800, Jeremy Kerr wrote:

quoted

quoted

+Addressing data is passed in socket system calls through
+.B struct sockaddr\_mctp

That escape is unnecessary.  Did you see it in another page perhaps?

I thought I'd seen some odd line-breaks at the underscore, but can't
replicate that now. Will remove.

My groff experiments don't reveal _ or \_ as being permissible
break points[1].  However, the structure tag _could_ break like this:

sock‐addr_mctp

In other words (if my UTF-8 gets mangled), after "sock".

To prevent that, you can prefix the word with the hyphenation control
escape sequence, "\%".  This escape sequence is extremely portable; it
dates back to 1970s AT&T troff.

Further, if you wanted to prevent breaking between the "struct" keyword
and the structure tag, you could use a nonbreaking adjustable space
escape sequence, "\~".  While this was a groff innovation (about 30
years ago), it's been adopted by Heirloom Doctools troff and mandoc(1),
so it's pretty portable to systems likely to install the Linux
man-pages.

So we might write

.B struct\~\%sockaddr_mctp

for instance in running text (that is, in sentences).  (When filling is
off, as in code examples and the synopses of most pages documenting C
interfaces, neither of these escape sequences is necessary; unfilled
lines are neither automatically hyphenated nor adjusted.)

Just a typographical FYI.  I know these issues sometimes frustrate
people.

Regards,
Branden

[1] I'm using groff's defaults.  It's possible to change the groff
    "hyphenation code" of any character, but few do so.