On Wednesday, 28 January 2026 08:31:30 GMT G. Branden Robinson wrote:
> Hi folks,
> 
> I haven't received feedback on either of my two previous requests[1][2]
> for nominations of "headline" items for the groff 1.24.0 release
> announcement.  I assume that's because people found it daunting to boil
> down ~900 lines of "news".  I know I did!
> 
> Here's my stab at it.
> 
> Let me know of any issues/gripes/suggestions you have.
> 
> ---snip---
> *  groff's "configure" script now supports a "--without-urw-fonts"
>    option to better support systems that don't require full PDF support
>    from groff.

A strange choice for "Top of the Pops". PDF fonts in 1.23.0 were handled 
automatically, if URW fonts were not installed you got the same result as if 
--without-urw-fonts was specified for 1.24. It was a bit "noisy" in that 
numerous warnings would appear during the build, illustrating gropdf's ability 
to produce pdfs would be less than a full service.

Your decision that warnings should be treated as errors [1], even though it 
was pointed out this is not normal behaviour (i.e. compiler warnings do not 
normally abort a build), [2] is the reason rc1 failed for people building 
groff without URW files. 

You changed the warnings to fatal at a time when you were struggling with the 
devpdf makefile [3], and I can understand making changes to your personal copy 
to aid you in your diagnosis but I don't understand why they were committed to
git.  You say [4]:-

"The bad news is that, often enough, a slight (or chunky) thing will go
wrong when the build tries to locate the URW fonts or their metrrics."

Please explain what slight and chunky things go wrong, as far as I know the 
problem is that if gropdf is in basic mode your groff-man-pages.pdf will have 
missing fonts on the gropdf(1) and grops(1) pages, which will emit warnings 
when pdfmom is run, but the pdf will be valid and viewable, similarly mom's 
typesetting.mom. For intermediate (ghostscript installed) and full modes both 
these are perfect. I'm assuming you have kept the chunky "thing" to yourself, 
please elucidate what it is.

In my opinion it was a mistake to promote warnings to errors and a better 
solution would have been to let warnings be warnings, since this worked fine 
for 1.23.0. So I don't understand why --without-urw-fonts deserves such an 
elevated position, given that it is a botch to cover up your decision to make 
normal warnings fatal.

Cheers

Deri

[1]  https://cgit.git.savannah.gnu.org/cgit/groff.git/commit/?
id=b53278be5deb5b6bc4a290ee73ba1e9e1d326134

[2]  https://savannah.gnu.org/bugs/?67612#comment1

[3]  https://savannah.gnu.org/bugs/?67268

[4]  https://lists.gnu.org/archive/html/groff/2026-01/msg00100.html




  • proposed "headline&qu... G. Branden Robinson
    • Re: proposed "he... Deri via GNU roff typesetting system discussion
      • Re: proposed &quo... G. Branden Robinson
        • Re: proposed ... Deri via GNU roff typesetting system discussion
          • Re: propo... G. Branden Robinson
            • Re: ... Peter Schaffter
              • ... G. Branden Robinson
                • ... Deri via GNU roff typesetting system discussion
                • ... G. Branden Robinson
                • ... Deri via GNU roff typesetting system discussion
                • ... Peter Schaffter
                • ... Deri via GNU roff typesetting system discussion

Reply via email to