[OpenFontLibrary] Google Font Directory

[Jonadab the Unsightly One]

Alexandre Prokoudine <alexandre.prokoudine at gmail.com> writes:

> Am I the only person who thinks that having Cyrillics ditched from
> both Droid Sans and Old Standard is a bugger?

For most of the English-speaking world, having that stripped out is a
good thing.  Think of the bandwidth, if nothing else.  Remember, these
are intended for @font-face embedding, which means every single
visitor to the page (whose browser supports it) is retrieving a copy
of the font.  You want that to be relatively small if possible.

Of course, if someone in the Slavic world wants to host some public
web fonts that do have Cyrillic support, that would also be great.
Similarly, if Baidu or somebody wants to host embeddable fonts with
Chinese characters, hey, good for them.

But it doesn't make sense for most English-language websites to embed
the full Unicode versions of web fonts, if there's a version available
that only has characters they might actually use.  For most of the
English-speaking world, all that's really needed is printable ASCII
characters plus about a dozen other characters in widespread use:
cent sign, pound sterling, yen, euro, n with tilde, copyright and
registered marks, degree symbol, lowercase e with accute accent so
people can impress prospective employers with it in the word resume on
their resumes, and maybe c with cedilla and the combining dieresis so
French loan words like facade and naive can be written in full-bore
pretentious mode.  Oh, and those stupid asymetrical "SmartQuotes" that
Microsoft Word always inserts (reusing the regular quotation mark
glyph for both orientations is completely adequate for these).

A few English-language websites (e.g., math-intensive ones, or
anything related to linguistics) need a bunch of additional symbols,
but in the overwhelming majority of cases it's baggage.  Websites that
do need a bunch of additional characters can jolly well go to the
trouble of specifically embedding a font that has them.  Most websites
don't need or want the overwhelming majority of Unicode characters.

Note that I'm only talking here about web fonts hosted for the purpose
of @font-face embedding, like the Google font directory.  For fonts
intended to be downloaded and installed on desktop systems, the
downside of having additional characters is much less and the upsides
of having them available are much more significant.  Consequently, I
*don't* think OFLB should start stripping out characters.  But for the
Google font directory it seems like the right choice.

-- 

v4sw5Phw5ln5pr5FPO/ck2ma9u7FLw2/5l6/7i6e6t2b7/en4a3Xr5g5T
http://hackerkey.com/decrypt.php?hackerkey=v4sw5PprFPOck2ma9uFw2l6i6e6t2b7en4g5T