On Fri, Jul 10, 2020 at 02:41:13PM +0200, Ingo Schwarze wrote:
> Subject: Re: Groff macro to make .UR and .UE links clickable in PDF?
>
> > .URL https://foo.bar.com/fred/juki/
> >
> > would be displayed (in PDF, HTML, and nroff) as simply
> >
> > foo.bar.com/fred/juki/
> >
> > Steve, what do you think of this practice?
> >
> > Everyone: Would anyone object if .URL used this strategy for cleaner
> > typesetting?
>
> Yes, i do strongly object.
>
> I think it is very bad practice to omit the protocol from an URI.
> For one thing, it results in invalid URI syntax. On top of that,
> the fact that this week, the web is a monoculture of https:// neither
> means that other protocols don't exist nor that other protocols
> cannot become used.
> .... And finally, the omission of the
> protocol can - depending on the context - cause confusion because
> it removes an obvious indicator that the thing printed is a URI in
> the first place, an indicator that the document author may have
> relied on.
Ingo, I think you're overreacting. I can't remember the last time
I had a problem in omitting the protocol in a browser location
bar.
But the question here is how to *display* a URL in text and my
rule of thumb is to reduce the size of the displayed URL as much
as possible. If we are talking about interactive documents, like
a PDF with clickable (terrible term) links, then I would argue
that the underlying link syntax should include a fully compliant
address, with protocol. But when it comes to documents meant to
be read (on paper or otherwise) optimal line length and the
syntax of a proper URL are just barely compatible.
It's well established that a comfortable line length for reading
is somewhere between about 40 and 70 characters. Most printed
books have line lengths of about 24 to 26 picas (27 max).
Anything with a longer line was likely designed by someone who
rarely reads (unfortunately I have found in my publishing career
that most people doing page design or actual layout for books
these days don't read much). Anything longer usually adds to eye
fatigue because of the distance the eye needs to travel back to
the beginning of the next line without losing vertical position.
Obviously vertical line space (or "leading") affects this as
well. Since most of us who aren't typesetting books are probably
typesetting for a letter-sized or A4 sheet of paper, we should
be setting type on two columns per page, which usually means a
line length of 20 to 22 picas. Most of my typsetting these days
is in that format. Way too many URLs don't fit on that size of a
line, so chopping off the protocol is entirely practical in order
to get consistent word spacing -- which is essential for a
smooth, rhythmic (Robert Bringhurst's concept) reading
experience.
An alternative rule followed by the publishing company I mostly
work with these days is to leave the protocol off if the address
begins with "www". I think that's a bit of a hack compromise,
since it assumes people will only quickly recognize a string of
characters as a URL if it begins with either the protocol or
something as familiar as "www". I also use colour to indicate
URLs in text as an additional aid to recognition, but that's not
always practical.
I think it's an abomination that a man page extends it's line
length to fit the width of the terminal; built into the macros
should be a 65- or 70 character maximum width. It's interesting
that the Python Style Guide insists on a maximum line length of
79 characters and recommends 72. A basic premise of python design
is *readability of code*. The main source of authors for man
pages is, I assume, programmers.
-- Steve
--
Steve Izma
-
Home: 35 Locust St., Kitchener, Ontario, Canada N2H 1W6
E-mail: si...@golden.net phone: 519-745-1313
cell (text only; not frequently checked): 519-998-2684
==
I have always felt the necessity to verify what to many seemed a
simple multiplication table.
-- Ilya Ehrenburg (Soviet author and critic; he's not
talking about mathematics)
