Re: [gentoo-user] Fonts and bad eyes

[Frank Steinmetzger]

On Sat, May 17, 2014 at 02:58:26PM +0100, Peter Humphrey wrote:
> On Saturday 17 May 2014 02:17:17 Dale wrote:
> > Howdy,
> >
> > I'm curious.  I'm sure there are some older folks on here that have eyes
> > that are not in the best of shape.  Mine are not real good even with
> > glasses.  My question is, what font is the easiest to read for folks
> > with bad eyes?  In other words, for you folks who can't see good, what
> > font do you use?
> > […]
> So far I've found these to be acceptable:
>
> 	Liberation Sans
> 	Bitstream Vera Sans
> 	Clockopia
> 	DejaVu Sans
> […]
> 	Verdana

For the record: DejaVu is practically the same as Bitstream Vera, but
with a much wider range of supported characters.
I suppose one of the reasons for their wide-spread use is (apart from
them being free) their high readability. Sans and serif are very similar
to another and do look nice.

> That last one, I believe, was designed by M$ for use in web pages.
>
> I'll spend some time with each of them and find which I like best. You'll
> notice that they're all sans-serif. That's because I believe serif fonts need
> a higher pixel density than most screens have, and that's why they work well
> when printed on paper but not here.

Serifs help the eye at staying on the line while perusing. We as Linux
users have the big advantage of the great font rendering engine (that
actually brought me to Linux in the first place many moons ago) which
can render such details beautifully, so we would notice them, but
without them distracting, even *if* they are a bit pixelated. (I switch
off the RGB subpixels rendering though because I don’t like the apparent
colour bleed.)
Serif fonts are designed to be used in longer texts. Thus they are a
suboptimal choice for UI elements, because those are usually rather
short.

Using hinting at full level might actually be a not-so-good idea,
because while it makes smaller fonts really crisp (filled pixel or no
filled pixel), it may lessen readability because
1) lines thickness can only vary by full pixels, making lines thinner
   than they actually are, especially on low-DPI screens.
2) the inter-letter spacing must be quantified in full pixels also.

So using half strength hinting might make the font look fuzzier on first
sight, but will improve reading flow because spacings are more even and
details can be perceived without poking out. And if you look from
farther away, it will look more natural.
The MS fonts have very detailed hinting information because they were
designed for screen use. That’s why they still look quite good with full
hinting on.


Another big advantage that we as Linuxlers have is that most GUIs will
scale nicely if we crank up the font size, as opposed to some commercial
OS I could mention. So Dale, since you are on KDE, use the freedom it
gives you (unlike some other DEs *cough*) and just crank up the sizes. ;)

Twiddling with DPI settings OTOH may be counterproductive. If you visit
a website that says: "font-size: 10pt", then the font will look the same
on *all* screens if their DPI is set to the actual value. If the DPI is
set to the same value for all, but they have *physically* different
pixel pitches, then the font will look different on each screen.


I’m not an old fart[TM] yet, so I can afford running a tiny terminus on
this 136 DPI laptop. ^^ However, when working in vim, I do use a colour
scheme similar to what wabenau describes in his mail: dark (but not
black) background with light (but not bright) text colours. For the
interested: that scheme is called Wombat:
http://www.vim.org/scripts/script.php?script_id=2465
-- 
Gruß | Greetings | Qapla’
Please do not share anything from, with or about me on any social network.

It is in human’s nature to think reasonably and act illogically.