ietf
[Top] [All Lists]

Re: Alternative decision process in RTCWeb

2013-11-28 10:38:27
On Thu, Nov 28, 2013 at 3:57 PM, Eric Rescorla <ekr(_at_)rtfm(_dot_)com> wrote:

It's actually not clear to me why IceWeasel and Chromium can't use OpenH264
in exactly the same way that Firefox can, i.e., by doing download from
Cisco's
site. I appreciate that people may have philosophical objections to doing
this, but I haven't actually heard any reason why this is a practical
problem
as opposed to a philosophical one. This isn't to say that there aren't
situations
in which a downloaded module doesn't work, but I don't see that those are
any different for Chromium or IceWeasel then they are for Firefox.


I think there's various issues of greater or lesser degree; logistics,
licensing compatibility, IPR concerns, and concerns about monoculture. Some
issues could be pigeon-holed as more than one; I don't know if the lack of
support on iOS counts as logistical, licensing, or IPR, for example.

But yes, I think the most vocal objections are from those who follow the
general spirit of the GPL, in as much as having a strong desire to ship
source. But I hesitate to brush that off as purely a philosophical debate
on how many codecs can fit on a pin - eliminating open source
implementations across many jurisdictions has a practical basis as well as
a philosophical one. Equally, eliminating barriers to open source
essentially eliminates all the other objections.

Don't take this as being a heavy bias against H.264 from me - I entirely
understand the reasoning behind wanting H.264 too. But it's easier to
express - deployed silicon is a simple, but powerful, argument.


(And

As a side note, the relationship between the Firefox binaries distributed
by
Mozilla and the Firefox source isn't quite the same as between Chromium
and Chrome. The Firefox that is distributed by Mozilla is basically the
same
as that you would get if you compiled the source yourself. By contrast,
Chrome
contains a bunch of extra bits that are not in Chromium (e.g., Flash).


Sorry, I'm aware of this but was simplifying somewhat too liberally. I also
simplified in implying there weren't those who were perfectly happy with
both H.264 and VP8 - but I do think there are significant chunks of the
market unhappy with one or the other.

Dave.