[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: common configuration for gazillion browsers



on Sun, May 20, 2001 at 01:51:59AM +0300, Tommi Komulainen (Tommi.Komulainen@iki.fi) wrote:
> On Thu, May 17, 2001 at 12:58:36PM -0700, Karsten M. Self wrote:
> > 
> > It's already got a name:  Standard Web Proxy Interface (SWPI):
> [...]
> > ...from Rick Moen Wednesday on linux-elitists.
> 
> ... and all I was thinking about was how to point my browsers to the same
> proxy in the first place.  Sounds cool, though.

It's just a thought, not a project.  I only wanted to note that's it's
been articulated already.

> > I'm arguing fonts in another list right now, the idea of proprietary,
> > controlled-distribution fonts in an environment which parks rendering on
> > desktops, many of which are governed by free software rules, is broken.
> 
> Umm, could you rephrase that so that I could understand it with my limited
> understanding of english?  Something about applications controlling
> themselves what fonts they use is bad...?

Many commonly used fonts are in essence non-free software.

They're software because the sizing, hinting, smoothing, and other
algorithms that are used to produce the shapes are covered by copyright.
In the US, the _shapes_ of characters themselves cannot be covered by
copyright -- it's roughly analogous to the idea of copyrighting the
alphabet -- you couldn't send any written text to anyone without paying
a royalty.

Traditionally -- up until 1995, that is -- the rendering of text could
pretty much be assumed to take place at a limited number of sites.  That
is, a graphic artist would create some work and print it or take it to
be printed.  Word processing documents would be produced and printed, or
occasionally emailed to someone else...who could pretty much be assumed
to be running similar and licensed software capable of reproducing the
fonts.

The history of font creation is rife with stories of creating fonts, the
work involved, and the relatively minimal payoffs.  Fonts are not
themselves copyrightable, so the work of a foundry is difficult to
recoup.  Even prior to the advent of the web, cheap CDROMs with
knockoffs of popular typefaces were rife -- I've still got the legacies
of several of these on my various boxen.

Two things changed all of this:  free software and the World Wide Web.

Free software is largely incompatible with the proprietary licenses
attached to many font faces.  While it isn't a conflict to load a
legitimate copy of the fonts onto a free software based box, it is often
a conflict to distribute the fonts freely with the OS or applications
themselves.  Many font licenses fail to meet the DFSG guidelines.

The problem, of course, is that document rendering no longer occurs in a
tightly controlled, licensable environment.  HTML distributes the task
of interpreting document markup according to the rules of a browser (and
the preferences of its user, if specifiable).  It's been said countless
times:  HTML governs content, not display.  The platform on which a
page is rendered may not have the primary, or fallback, fonts specified,
or even any fonts at all if it's a text-to-speech device.

My suggestion to the webmaster (well, mistress) of a well-regarded and
generally clueful news site was that the ball really is in the web
designers' court.  If the web design community isn't going to
collectively see that site specs are readable across a wide range of
platforms, then client-side tools to fix the problems will emerge.  The
prototypes already exist, including rewriting proxies.

So, rephrasing my earlier comment, a system which is based on the idea
that you can simultaneously restrict (through proprietary licensing)
font distribution, provide highly specified layout directives in
material, and require the rendering of this material to take place in
heterogeneous technical and legally governed environments (different
desktops, OSs, and licensing rules), just don't work no more.  It's
broken.

Initial obfuscatory terseness regretted   ;-)

-- 
Karsten M. Self <kmself@ix.netcom.com>    http://kmself.home.netcom.com/
 What part of "Gestalt" don't you understand?       There is no K5 cabal
  http://gestalt-system.sourceforge.net/         http://www.kuro5hin.org

Attachment: pgpijJBfrKuqo.pgp
Description: PGP signature


Reply to: