Re: man/man8/ldconfig.8: document system-wide tunables

"G. Branden Robinson" <[email protected]> writes:
>> > The man(7) package doesn't impose a rigid stylesheet on its output.
>> 
>> Perhaps, but the fact that I now know about the rule, and the
>> exception, means that it's not entirely flexible either ;-)
>
> The man(7) macro language is not highly flexible, no,

Sorry, I meant the man-pages style guide rules, not roff itself.

>> In this case, it does sound like the man pages need a different thing
>> for "example code" (.EX) and "preformatted text" (html's <PRE>).
>
> A common error is to assume that `EX` implies `<PRE>`.  It doesn't.

I just meant for semantic context, like "this is a code snippet" vs "I
want this to be formatted the way I typed it."  Might be helpful when
outputting HTML/CSS, like the examples would be in boxes and shaded with
the usual "copy to paste buffer" button, but <pre> would just be
monospaced and unformatted.  But thus is the slippery slope that led to
XML, and we don't want that again.

>> glibc uses texinfo, which is TeX but terminals don't have **** off ;-)
>
> Yes.  And info(1) sometimes produces...intriguing results.

My preference here is PDF, as we make sure all the cross refs and TOC
work correctly, and it formats very nicely (duh, TeX).  Info is useful
in emacs only because my fingers are already on the keyboard.

>> I don't write programs in assembly either (well, usually ;).
>
> No indeed.  The real "assembly language" of troff is its
> "device-independent output", which for short I call "trout" if it's
> Kernighan's original spec from CSTR #97 in 1982, or "grout" if it
> features groff's extensions.

I've been accused of abusing my own namespace here.  I would have called
my version "djrout".

> Yes, but people get _really wedded_ to the specifics of rendering even
> after setting aside the huge issue of terminal width.

I once worked for a guy who, at one point, wrote a typesetter that went
between satellites and the New York Times.  The typesetting had to be
perfect, every time, and run at line speed.  I learned a lot from him,
and he's the reason I have my own copy of the Chicago Manual of Style
and know what Reverse Boustrophedon is.

Also, Team Oxford Comma!

>> (wait, when did "man -l" happen?  Have I been missing that all along?
>> Is my script really that old?)
>
> In the man-db implementation of man(1), it seems to date back to _at
> least_ 2002, as I see a Git commit for it (imported from CVS or

I might predate even that.

> If you are/were using Red Hat/RPM-based systems, you may have been using
> Brouwer/Lucifredi man(1), which as far as I know never implemented this
> `-l` option.  After that implementation lay moribund for many years,

RHEL 9 has "man -l" via man.man-db at least, but that's today.