W3C home > Mailing lists > Public > www-style@w3.org > June 2009

Re: RE: New work on fonts at W3C

From: Dave Crossland <dave@lab6.com>
Date: Tue, 23 Jun 2009 19:20:25 +0100
Message-ID: <2285a9d20906231120g25dd071boa7c9f3e8ce150559@mail.gmail.com>
To: Chris Wilson <Chris.Wilson@microsoft.com>
Cc: "www-style@w3.org" <www-style@w3.org>
2009/6/23 Chris Wilson <Chris.Wilson@microsoft.com>:
> Dave Crossland wrote:
>>2009/6/23 Chris Wilson <Chris.Wilson@microsoft.com>:
>>> I don't speak for IE.
>>
>>Do you personally think IE would implement a non-DRM web font format?
> >From what I have heard from Microsoft representatives, I don't.
>
> If "non-DRM web font format" means "TTF/OTF direct linking", then no, I don't.

I didn't mean to imply TTF/OTF, but a new web format that isn't DRM, unlike EOT.

> If non-DRM web font format means "web font format that specifies what
> license and rights the font carries, font vendors define those rights and
> the UAs are expected to honor those, but there's no strong crypto or
> single-vendor-controlled system 'protecting' the font," then yes - in
> fact, that's pretty much what EOT does, despite being demonized
> as "DRM".

It becomes DRM when you say, "UAs are expected to honor those [restrictions]"

A non-DRM web font format to me means "web font format that specifies
what license and restrictions the font carries, font vendors define
those restrictions, and the UAs are expected to display them to users,
but there's no strong crypto or  single-vendor-controlled system
'protecting' the font or UAs enforcing restrictions on users."

> The real problem here is workflow.  You likely have a
> couple of hundred TTF/OTF files on your system (I have
> 236 on my Windows machine, and 241 on my Mac).  They
> contain copyright information, but likely NOT licensing
> information, other than the embedding bit.  Fundamentally,
> you don't know if you can use them with any solution that
> requires redistribution of the original file, unless you track
> this (e.g. keep a spreadsheet or attach information somehow
> to each font, describing its license).  There's not anything
> you can do to change this; the direct linking solution is
> going to be a mess.

I agree that scarcity of licensing information in file metadata is the
central problem, and I think Lord's wrapper proposal is a fine
solution to it, and as I understand it, Tom's proposal means wrapped
fonts would not be directly drag and dropped into the OS font system
as a side-effect of supplying this metadata.

This seems to be the central desire of proprietary font developers for
a new web format.

> I've been very clear in other forums on direct, unadorned
> placing of TTF/OTF files (e.g. with no machine-readable
> licensing information that is checked by the UA) on a web
> server - I think it is a drastic mistake,

Checking the information is okay, it depends what UAs are required by
a W3C Rec to do with it; are they required to use the machine-readable
licensing information to _assist_ people and display that information,
or to _manage_ people and enforce restrictions?

>>But trying to make read-only fonts is like make water not wet?
>
> No.  Trying to put up rigid defenses around font software
> is likely not a winning proposal, because as you imply, to use
> fonts is (technologically) to use fonts - rendering and editing
> look about the same.  Trying to carry licensing information to instruct
> proper usage is like putting up signs saying whether the water is
> potable or not - it doesn't stop people from drinking it, but it does
> encourage doing the right thing.

Here you seem to be talking about assisting people and not expecting
UAs to enforce restrictions.


P.S. Thanks for a thoughtful discussion, I'm enjoying this :-)
Received on Tuesday, 23 June 2009 18:21:24 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:19 GMT