Re: grohtml: confused about constant-width font handling

"G. Branden Robinson" <[email protected]> Tue, 20 Jan 2026 11:49:09 -0600

Newsgroups	gmane.comp.printing.groff.general
Message-ID	<20260120174909.etetlemzfnfxglnk@illithid>

At 2026-01-20T16:18:11+0000, Colin Watson wrote:
> Further confusion with groff 1.23.0.  ".nf" then ".ft C" results in
> <pre>.

Taken together, that almost makes sense.  But only almost.  This was a
bad idea because disabling filling does not have the same semantics as
HTML's <pre>, even if you ignore font selection issues.

(At which point someone says "what? there are semantics other than the
font selection?"  Whereupon grizzled old text formatting system veterans
smile wanly and nod.)

When filling is disabled, all *roff formatters still:

* apply inter-sentence spacing (groff: if configured);
* interpret control lines; and
* interpret escape sequences.

They also still honor character translations, which isn't very "<pre>"
or "\verb".  Fortunately, only a madman would use character translations
in a man page.

People need to not get carried away with what "no-fill" or "disabling
filling" means.  It *roff systems, it is limited to that and its logical
consequences (no automatic breaking, no hyphenation, no adjustment).

> ".ft C" on its own does not result in <tt> even though you might think
> it would based on the behaviour in combination with ".nf".  ".ft CR"
> has no effect with or without ".nf".
> 
> I can see why the author of this page decided to use ".ft C" for the
> HTML device even though it doesn't seem to actually make any sense,
> since it _works_ in 1.23.0

As far as I know, that idiom wasn't _documented_ anywhere.

> - but doesn't any more.  What's going on?

The font name "C" is not portable and is now deprecated.

https://cgit.git.savannah.gnu.org/cgit/groff.git/tree/NEWS?h=1.24.0.rc1#n366

Regards,
Branden

signature.asc (application/pgp-signature, 833 B)