Posted by Andy Dingley on 03/30/06 14:52

charles cashion wrote:

> A search of my entire hard drive does not find
> arial or helvetica or courier. the closest thing
> is Courier.afm.
>
> Can somebody help me understand this?

The mappings of typefaces names onto actual fonts is an arcane process
that's not understandable by mere mortals, like web designers.

There are three levels of mapping

- "serif", "sans-serif" and "monospace" are about all you can
practically request on the web. You don't know what the users actually
have installed and the browser just maps it for you. This isn't
flexible, but it works. You can't (practically) achieve much more than
this on the public web. If you assume "Most people run Windows, most
recent Windows ships with Comic Sans" then you can do better - but
still not usefully (the commonplace fonts that aren't already the
serif/sans defaults aren't particularly attractive)

- For copyright and trademark reasons, it's possible to make a
typeface called "Swiss" but not one called "Helvetica" - even though
"Swiss" might be an indistinguishable copy of it. So real-world
computer desktops and printers already understand how to map these
various names around (there are embedded lists of close synonyms),
because that's a significant part of making any sort of installable
font system work.

- There are "font metrics", describing a typeface as hard numbers in
terms of its quantified appearance. "Similar" fonts can now be
identifed automatically. No-one outside the ivory towers of Gutenberg
really understands this.

[Back to original message]