[seqfan] Re: Today's XKCD

[Hugo Pfoertner]

The intent of the comment "Helvetica is a proprietary font ...." was
different. 99.9xx % of the standard system and program installations on
Windows, Linux and perhaps also Mac computers don't provide Helvetica as an
option for printing. To get access to Helvetica, one has to buy a license
from the Linotype-Hell AG. Most Windows users believe that Arial would
deliver  "somehow" similar results to Helvetica, but they cannot verify
this assumption. The same caveat applies to all other substitutions for
Helvetica in Ghostscript, Open/Libre office, ...
The only way for myself to use genuine Helvetica would be to write a
Postscript file, re-activate my first PC and send the file to the legacy HP
LaserJet 4MP postscript printer that has an authorized built-in capability
to interpret Helvetica correctly.

Even the expensive Adobe CS tools like Illustrator, InDesign, etc. didn't
include the Helvetica font in the standard installation. Arial was
available, an I'll use InDesign and its pdf export to create the announced
graphical example. So the task described in the comic " .... printed in
Helvetica" is only solvable for graphics professionals with access to a
licensed copy of the font.

In contrast, all Windows installations -still- include the Arial True Type
font defined in the file
arial.ttf, size 1036584, dated 2018 Apr 12 01:34, last distributed with
Windows 10 Pro Version 1803. What I did, was to extract the Adobe Font
Metrics information from this file  into an .afm file using
https://www.files-conversion.com/font-converter.php
which seems to use FontForge https://fontforge.github.io/en-US/ for the
conversion. I don't think that there is anything violating license rules
with this procedure.

In the meantime I have found an "official" information on the font metrics
of Helvetica at Adobe's Font technical notes webpage
https://www.adobe.com/devnet/font.html. It provides a link to
http://download.macromedia.com/pub/developer/opentype/tech-notes/Core14_AFMs.zip
, which includes helvetica.afm. From its content, it seems that
"integers...  printed in Helvetica" would produce massive differences to
"printed in Arial", due to the absence of the kern pair definition "KPX one
one -74" found in arial.afm, but not in helvetica.afm, that causes the
shrinking of print width for numbers containing consecutive ones. To be
provocative, one can say "Arial intentionally messes up the printing of
numbers".
See the second (Earl Jenkins's) answer in
https://stackoverflow.com/questions/3949422/which-letter-of-the-english-alphabet-takes-up-most-pixels
and David Harralson's answer in
https://www.quora.com/Why-did-Microsoft-change-the-default-font-to-Calibri
for some historical background.

Hugo Pfoertner

On Mon, Jul 9, 2018 at 4:06 AM, William Marshall <w.r.marshall at actrix.co.nz>
wrote:

> On 09/07/18 06:42, Hugo Pfoertner wrote:
>
>> The "print width" example has been too tempting for me to make a "dumb"
>> sequence:
>> https://oeis.org/draft/A316599
>> If it is not rejected, I'll try to insert a graphical example of a
>> correspondingly ordered printout.
>>
>> Hugo Pfoertner
>>
> From the sequence comments:
>
> "since Helvetica is a proprietary font, the almost-identical font metrics
> of the Arial regular font are used here instead"
>
> Arial is also a proprietary font. Perhaps you had meant to use Liberation
> Sans instead.
>
>
> --
> Seqfan Mailing list - http://list.seqfan.eu/
>