W3C home > Mailing lists > Public > www-tag@w3.org > July 2011

Re: Revised HTML/XML Task Force Report

From: Eric J. Bowman <eric@bisonsystems.net>
Date: Wed, 13 Jul 2011 14:32:10 -0600
To: Chris Lilley <chris@w3.org>
Cc: Larry Masinter <masinter@adobe.com>, "karld@opera.com" <karld@opera.com>, "nrm@arcanedomain.com" <nrm@arcanedomain.com>, "www-tag@w3.org" <www-tag@w3.org>, "ndw@nwalsh.com" <ndw@nwalsh.com>
Message-Id: <20110713143210.73538769.eric@bisonsystems.net>
Chris Lilley wrote:
> 
> EJB> I don't see image
> EJB> replacement going away any time soon, as WOFF (unlike EOT)
> EJB> requires the user-agent to download the entire font as opposed
> EJB> to the limited number of glyphs required for a site's headings
> EJB> (in many cases).
> 
> Somewhat off topic but I didn't want this misinformation to go
> unaddressed.
> 

In order for SIFR to make a good example, it can't be something which
is obsolete and therefore subject to being hand-waved away.  My point
is that neither the need for image replacement, nor ubiquitous Flash
plugins, are going away any time soon.  If I'm demonstrably wrong, SIFR
is an irrelevant example of a polyglot roadblock (which I wish it wasn't
as I favor Larry's position, I'd just prefer the debate to center around
a real-world problem).

>
> That is an incorrect characterization. Whether an original font has
> been subsetted, lightly or extensively, is orthogonal to whether the
> delivery format is EOT or WOFF.
>

The issue is installable vs. embeddable, not subsetting; I don't know
what I was thinking there so my apologies to all.  Here's why SIFR
isn't obsoleted by WOFF:

I've had a client since 1994 whose branding requires me to use Palatino
Linotype, which their website still doesn't on Linux.  There is an open
Palatino, but it's actually Book Antiqua with hinting stripped, so not
only is the italic-w glyph wrong for my client's logo, but it looks
spectacularly hideous on Linux.

Palatino Linotype's license allows embedded subset distribution, but
disallows installable distribution (without paying a per-user fee,
making it prohibitively expensive for the Web).  So I can make an EOT
of, at the very least, the glyphs used in the logo.  But I can't
distribute this subset via WOFF because it would be installable --  had
EOT been adopted, I'd be able to do cross-platform Palatino Linotype.
Please correct me if I'm wrong, as this licensing problem holds true
for many of the classic fonts website owners have long desired to use
other than by hosting PDFs which stay true to their branding.  

As a developer, this gives me two options:  image replacement, or the
browser-default serif on Linux (I used the latter as this client doesn't
care about Linux).  Making a .gif of the logo fixes it at a set size,
whereas with text the size is proportional to the site text regardless
of settings -- which is why SIFR isn't going away.  Don't get me wrong,
I'm not anti-WOFF; where I have design choice over fonts it's great to
have so many new options to choose from, but it hasn't caused me to
file image replacement away with spacer.gif as I'd hoped.

-Eric
Received on Wednesday, 13 July 2011 20:32:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:48:39 GMT