Re: New work on fonts at W3C

Aryeh Gregor wrote:
>> Also note that because Gecko (Mozilla Firefox) and WebKit (Apple Safari,
>> Google Chrome) are licensed under GPL those rendering engines cannot
>> implement a patent encumbered font format. And if a license is granted
>> for Gecko and WebKit, such license is automatically applicable to any
>> derivative work (licensed under GPL) of Gecko or WebKit, especially a
>> piece of software which only purpose is to download any EOT wrapped font
>> file and install it as a system font.
> 
> 1) Google's support for H.264 in Chrome suggests they don't buy this
> interpretation to begin with (AFAICT, IANAL).

Mikko is correct if you assume that the web browser vendor wants to 
allow others to redistribute the browser without having to renegotiate 
patent licensing.

Google Chrome explicitly doesn't care about this, based on the recent 
threads about the gpl-related issues here; this is why Chrome is able to 
add support for H.264.  Firefox does, so it's not able to do so.

> 2) Both Gecko and WebKit are licensed under the LGPL (inter alia), not
> only the GPL.  The LGPL permits linking to proprietary code, including
> code implementing patented algorithms, according to what I understand.
>  The portion that actually implements the decoding could be licensed
> under BSD, or LGPL with a special exception, and just linked with the
> LGPL code.

Not all of Webkit is licensed under LGPL, last I checked.  But in any 
case, that doesn't matter; see above.  Mozilla _could_ distribute 
Firefox with H.264 while fulfilling its minimal legal obligations under 
the LGPL.  It's just a matter of whether one wishes to do more than 
that.  As I understand.  IANAL as well.

> 3) But the legalities are moot because Mozilla has said they won't
> ship patented formats regardless of whether they can do so legally,
> and there's no point in going forward with a format without support
> from Mozilla.

That's not what Mozilla has said.  IANAL and I don't officially speak 
for Mozilla on this, but I believe what "they" (we?) have said is that 
for us to ship a patented format it must at a minimum be possible for 
others to then redistribute our code without having to renegotiate the 
patent license.  There may also be something about field-of-use 
restrictions that I'm not qualified to comment on.  But being "patented" 
by no means immediately rules out anything.  What matters is the patent 
license, not the fact of being patented.

-Boris

Received on Monday, 22 June 2009 20:49:42 UTC