Re: Fonts (was Re: follow up on service workers for publishing platform)

Hello,

Wednesday, December 2, 2015, 1:53:09 PM, Vladimir wrote:

> On Wednesday, December 02, 2015 12:53 PM Leonard Rosenthol wrote:

>> I am not trying to portray doom and gloom, but I am raising a topic that needs 
>> deep investigation both from a technical as well as a legal perspective for not 
>> only ourselves but the users of PWP.  It would be wrong of us developing 
>> a standard to do so known that there are pitfalls for users.

I strongly disagree, based on prior experience.

> I guess the major contention point on my side is that I disagree
> that DPUB (or W3C as an entity) needs to be concerned with the
> licensing / legal aspects of using fonts.

Exactly. Let me amplify that point.

The early history of webfonts, when it was a failed technology,
consistently mixed up the technical and the legal/licensing aspects.

On the one had we had Microsoft's Embedded OpenType (EOT, used by IE4
and up), with a URL embedded in the font for the site that was allowed
to use it; the browser was expected to police font usage and not apply
fonts to the wrong site. Whether a browser which failed to heed this
would be seen as infringing (and liable) under the US DMCA was a
significant chilling effect on browser adoption.

On the other hand we had Bitstream Portable Font Resource (PFR, used
by Netscape 4.x) which was described as being a clean derivation of
font rendering and not an actual font, thus not subject to to the font
license conditions. This was a hotly disputed claim.

There were two attempts to start a W3C working group to standardize
EOT. The patented and proprietary technological aspects were by this
time openly licensed for RF use. Both attempts failed, due to the
legal/licensing aspects. The foundries were still worried about font
theft, the browser vendors were still worried about DRM and liability.

WebFonts only took off when this decade-long mixing up of technical
and legal aspects was replaced by a clean separation. WOFF (and WOFF2)
allows (but does not require) a link to a human-readable license. The
browser is not required (actually, *is forbidden*) to use the metadata
to influence rendering.

As a direct result of this, the vast majority of foundries now offer
clear, dual (desktop/print and web) licensing. All browsers now
support WOFF. WebFont useage on the top 100,000 sites has increased
from 2% in 2009 to 50% in 2015.

This is clear, compelling evidence that we must not dive into those
mistakes of the past. Unlike Leonard, I *am* warning of doom and
gloom here and can confidently predict it if we one again follow this
erroneous path.

The Open Web Platform has eventually suceeded with a WebFonts
solution. Digital Publishing is now following the OWP rather than
trying to re-invent unnecessary wheels. Thus, this is a solved
technical problem for DPub too.

> a publisher needs to get the font license that allows them to do
> what they need to do. If they need to make their content available
> both online and offline they'd need to use WOFF files and they need
> a license that allows WOFF files be embedded in EPUB. (If a
> particular font vendor is not permitting this kind of use - there
> are others that will.) The particular conditions of that license
> (and how much it costs) is something that forms a contractual
> relationship between a client and a vendor - neither W3C nor DPUB /
> EPUB / PWP specs need to be concerned with that part.

Agreed.

If it turns out that WebFont licenses need to be clarified, extended,
or a separate DPub license developed, then foundries will do that.
That need not concern this group; it is a legal, licensing,
market-driven exercise.


-- 
Best regards,
 Chris  Lilley
 Technical Director, W3C Interaction Domain

Received on Wednesday, 2 December 2015 20:54:08 UTC