[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#42240: Interpretation of 'size' Property in Fontconfig Pattern Strin
From: |
Sebastian Miele |
Subject: |
bug#42240: Interpretation of 'size' Property in Fontconfig Pattern Strings |
Date: |
Thu, 16 Jul 2020 20:10:30 +0200 |
User-agent: |
mu4e 1.4.10; emacs 27.0.91 |
Robert Pluim <rpluim@gmail.com>:
> Cairo implies --without-xft in any case, so no harm done.
Good to know. Thank you.
> Looks like Iʼll have to boot up my GNU/Linux box (although Iʼm hazy on
> the difference between 'size' and 'pixelsize')
The main Fontconfig manual page describing all properties is called 'man
5 fonts-conf' (which can be hard to find, because neither 'fc' nor
'fontconfig' are substrings of 'fonts-conf').
The 'size' is the point size. In general, there are many different
definitions of point size. But in practice only one is used nowadays
(except by TeX which uses a slightly different but very close definition
that works better with its fixed point arithmetics). Usually a point is
1/72 inch (https://en.wikipedia.org/wiki/Point_(typography)).
Together with 'dpi' (dots/pixels per inch) that defines the relation
between point size and pixel size.
X has at least two different places where a dpi may be defined. One for
the entire X display, although that does not really make sense with an X
display having monitors of different pixel densities. Apart from that at
least the Xrandr extension has a notion of distinct dpis per monitor.
In general, Fontconfig has an even separate dpi. I do not really know
what dpi Fontconfig uses, if it is not explicitly set in the Fontconfig
configuration, but, e.g.,
fc-match :family=Hack:size=11.1 size pixelsize dpi
or
fc-match -v :family=Hack:size=11.
display the dpi used.