Re: Questions re web-fonts

On Wed, 2009-07-15 at 11:40 -0700, John Hudson wrote:
> Thomas Lord wrote:
> > The problem, John, is that TTF/OTF exist as
> > standards precisely to facilitate the exchange
> > of font files between applications.  These
> > are "open standards", intended to be implemented
> > by anyone and everyone who cares to do so.
> TrueType came into being to break PostScript's stranglehold on page 
> layout and printer drivers. OpenType came into being to extend the 
> capabilities of TrueType to handle complex scripts and, as a byproduct, 
> more sophisticated typographic layout. Fundamentally, the formats exist 
> for the purpose of document creation and display, not 'exchange of font 
> files between applications'.

Now, consider what happens if you create a document with
an embedded font using your word processor and then send
me a copy to view with my different word processor.  Or,
more fundamentally, consider the exchange between a font
editor and a window system where a font is installed.
That's "exchange between applications".

>  What is exchanged between applications is 
> typically content, in some form from plain text up to fully styled text 
> and document layout; the circumstances in which a TrueType or OpenType 
> font travels with a document is rare.

Consider the path from a font editor to a desktop 
installation.  Consider the exchange on the Internet 
of permissively licensed fonts.  These are all exchanges.

TTF/OTF became "open" so that applications would
implement it and thereby extend the opportunities for
font file exchange.

>  When a font does travel with a 
> document, e.g. embedded in a PDF, it is most often subsetted and lacking 
> many of the capabilities of the document creation font. Further, such 
> embedding is properly respectful of the embedding bit permissions in the 
> font OS/2 table; whereas the browsers supporting naked font linking are 
> apparently unwilling to be responsible for any such permissions.

We've been over that before.  No standard can
correctly specify and no program can accurately
compute the legal rights of a user with respect 
to a font.  Several important programs already,
today, ignore the permission bits either always or
as an option in order to err on the side of protecting
a user's freedom to exercise legitimate rights.
Browsers are scarcely the first example of such


Received on Wednesday, 15 July 2009 19:14:15 UTC