groff
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: First impressions: strange groff default font behaviour after system


From: Oliver Corff
Subject: Re: First impressions: strange groff default font behaviour after system upgrade
Date: Mon, 18 Dec 2023 22:49:55 +0100
User-agent: Mozilla Thunderbird

Hi Deri,

Here we go. In the meantime I came to the conclusion that the dpdviewer
is at fault, or rather the viewer's setup; I forwarded the pdf to a
different system, and there everything was visible as intended.

Best regards, and thank you!

Oliver.


On 18/12/2023 22:34, Deri wrote:
On Monday, 18 December 2023 20:57:07 GMT Oliver Corff wrote:
Dear All,

today I upgraded my system from Fedora 38 to Fedora 39 which has the
appreciated side-effect that groff 1.23.0 is finally installed by default.

I tested the new installation. As was to be expected, groffer was gone
(I liked it for q&d test runs of code snippets, so I actually miss it. I
wrote a small shell script to which calls groff and the pdf viewer, so
that is fair enough as a substitute).

However, I was eager to see whether anything has changed on the font side.

Under an out-of-the-box setup of my previous groff-1.22.4 installation
(delivered with Fedora 38), I could not print the characters  \[<=] and
\[>=]. Little crossed boxes appeared instead. My workaround then was

.char ≤ \o'<_'

which looks close enough to the glyph displayed by my system font for
the terminal, and goes well with a document typeset in Helvetica.

On the other hand, Greek symbols for math and statistics, like Chi
Square, could be entered in the file as χ\~² and appeared quite as
expected. Without the spacing command, the north-east end of χ ends up
in the hook of the ².

Enter groff 1.23.0. I compiled the same file again (a translation which
I had finished just this morning, not knowing what an escape I had*),
and alas! things took an unexpected course. First I looked for the
appearance of  ≤ and was astonished to see that not only was it
invisible, it was truly invisible as no placeholder box appeared; blank
space was there, at least. Then I noticed strange holes in the text ---
the Greek letters did not show up either. Again, no placeholder box,
just white space.

So, this is a brand-new Fedora 39 installation with groff version
1.23.0, the URW fonts being found in /usr/share/fons/urw-base35/.

My first question: Is this new behaviour intended?

If so, what I am I supposed to do?

If not, what kind of tests and diagnostics should I conduct?

If I compile a minimal ms document like

.PP
1≤2

I can copy and paste the white space between 1 and 2 from the resulting
pdf document, and lo and behold, it is a "≤" !

And, as a side-note, there is a typo in refer(1), right in the first
line(2) of ther first contiguous paragraph: "a preprocessor that
prepares bibilographic citations".

As a second side-note, I am willing to learn. I discovered the
BUG-REPORT instruction. So for my next observation, I'll go to Savannah.

Thank you a lot,

Best regards,

Oliver.

* I had to meet a deadline this morning, handing in my work just so-so
in time.

Hi Oliver,

Please could you provide a copy of the minimal ms pdf document run with the
command:-

groff -T pdf -ms -P-d -k

And attach the download file from the font/devpdf directory along with an "ls"
of that directory.

This will let me start to see what is going on.

You can either send them to me or attach it to a savannah bug report.

Cheers

Deri



--
Dr. Oliver Corff
Wittelsbacherstr. 5A
10707 Berlin
GERMANY
Tel.: +49-30-85727260
mailto:oliver.corff@email.de

Attachment: 0_Definitionen.roff
Description: Text Data

Attachment: my_char.ms
Description: Text Data

Attachment: my_char.pdf
Description: Adobe PDF document

Attachment: download
Description: Text document

Attachment: devpdf.txt
Description: Text document

Attachment: urw-base35.txt
Description: Text document


reply via email to

[Prev in Thread] Current Thread [Next in Thread]