[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: mplayer / divx

On Tue, Aug 28, 2001 at 09:32:57PM -0700, Aaron Lehmann wrote:
> I would like to package Mplayer with the ffmpeg modifications that
> allow it to play DivX videos without non-free code. The upstream code
> of Mplayer includes OpenDivx which is non-free and I am willing to
> strip this code from it in favor of the ffmpeg code if the OpenDivx
> license is not changed soon.

Good idea. 

> I think that takes care of licensing, but I know that patents are a
> problem when it comes to DivX. Unfortunately I don't know much detail
> beyond that. Rules of my home country governing me beside, what
> patents are problematic and how would it affect my hope to package?

Like the holders of many perceptual audio/video encoding patents, the
holders of the mpeg-4 video patents may claim that decoders are covered.
Indeed, the holders of the various patents surrounding mp3 encoding claim
that their patents cover decoders. Of course, their claim is probably false
and they know it - which is why they haven't pushed the issue. But they
still manage to sucker a good number of hardware/software makers into paying
their fees for decoders.

Additionally, it's still unclear if instructions (computer or not)
explaining how to perform a patented process are capable of being restricted
by a patent. Traditionally, patents have applied only to building/performing
the invention, not to writing code or books about how to make/do it. In the
past, Debian has allowed source code which explains how to perform patented
processes such as LZW encoding to be distributed from non-US. We also
distribute binaries which are not only very detailed instructions about how
to perform such code, but are also instructions which can be followed
automatically by a computer.

Of course, there is some disparity in our policy with regard to patent
encumbered software. For example, we allow LZW, idea, and other patented
algorithms, but we fail to allow layer-3 encoders (although one was in the
archive unnoticed for some time - see bug #97198). Nobody has come up with a
compelling argument in favor of the current disparity.

Brian Ristuccia

Reply to: