Re: MP3 decoder packaged with XMMS
> I have copied the Executive Contact and the Legal Counsel for Xiph.org
> on this message. Please drop them on follow-ups that are not relevant
> to Ogg/Vorbis. Mr. Rosedale and Mr. Moffitt: the topic of MP3 patents
> arose on debian-legal (thread at
> http://lists.debian.org/debian-legal/2005/07/msg00081.html ) and we
> could all use some competent advice.
Just a quick note on this thread. Time seems to have erased the memory
of Thompson going after everyone. 8hz-enc, bladeenc, lame, and many
other projects have shut down (from cease and desist letters) or refuse
to distribute binaries because the MP3 suite of patents _is_ actively
enforced. Try going out and finding unlicensed implementations outside
the Free Software and Open Source worlds. That out of the way, I will
address the issues raised below.
> > > If I were defending, say, an Ogg/Vorbis implementation [...] I
> > > would argue that a wavelet transform is sufficiently different [...]
> > Wavelet transforms are not the only thing the format supports, but it
> > may be usable to defend a particular encoder.
I don't believe there are wavelet transforms in Ogg Vorbis. These are
planned for some future incompatible update.
> Do you happen to know whether the Xiph.org team has retained competent
> counsel to evaluate the possible impact of the Fraunhofer and Sisvel
> patent suites on Ogg/Vorbis? (They claim that Ogg/Vorbis is
> "patent-and-royalty-free" at http://www.xiph.org/ogg/vorbis/ , which
> is pretty strong language.) If not, maybe Fluendo would fund the
> legal fees -- they seem willing to pay money to random lawyers for
> (IMHO, IANAL) dubious opinions and to post the result publicly
> (Google: gstreamer Moglen).
Before we released Ogg Vorbis beta 1, we did indeed hire a patent
specializing attorney to go over the MP3 suite of patents. He only
thought it necessary to issue a formal opinion on a single one of these
patents. We were advised by him, and other attorney's, that the
specifics of this opinion could not be divulged publically. Since that
time (around 2000,2001 I believe), I believe several companies have also
had lawyers look over this issue. RedHat ships Ogg Vorbis, and they are
obviously aware of these patent problems due to their removal of MP3
related software, so I assume they made this decision based on sound
legal advice. I don't believe anyone is going to publically share their
findings any more than we have for the same reasons our lawyer original
New patents come up all the time. No one can afford to keep track of
them all, or to have attorney's issue legal opinions on anything
related. We have done informal, but educated, analysis on many patents
that others have brought to our attention, and never found anything
worth troubling a lawyer over. Also, we originally intended the
patent-free part of our software, so we based many algorithms on old,
widely published results, and avoided many methods that would lead to
Many large corporations ship Ogg Vorbis with their products, including
Microsoft, RealNetworks, EA Games, and many more. There are plenty of
billion dollar companies to go after for infringement, should
infringement actually be occuring due to Ogg Vorbis. The fact the none
of these companies has been, to anyone's knowledge, threatened with
litigation over related patents speaks volumes. We've been around for 5
years, and we've taken this issue seriously the entire time.
> Personally, I would be little more inclined to rely on the continued
> availability of royalty-free open-source Ogg/Vorbis encoders than
> their MP3 equivalents without some indication that someone competent
> is on record as to the basis for a reasonable belief that they do not
> infringe the Fraunhofer suite.
What I have said above we thought was common knowledge. There are
probably very few Free Software projects that have dealt with this issue
as seriously as Xiph.org.
One last note: I am still on the board of the Xiph.org Foundation, but
Monty <email@example.com> is currently the Executive Director. Tom
Rosedale is our current legal counsel, but was not the attorney who did
the patent review, nor was he actively involved with us at the time the
patent review was done.
Feel free to continue copying me on the discussion. As a fellow debian
developer, I'm quite interested in this issue from both sides.