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

Re: MP3 decoder packaged with XMMS

On 7/16/05, Arnoud Engelfriet <galactus@stack.nl> wrote:
>    (5) An act which, apart from this subsection, would constitute an
>    infringement of a patent for an invention shall not do so if -
>    (a) it is done privately and for purposes which are not commercial;

OK, this makes sense.  Probably means I can bake cookies from a
patented recipe without infringing the patent as long as I don't turn
around and hold a bake sale.  My execrable Deutsch is still enough to
tell me that German law is similar.

> > At least in US copyright law as I understand it (which borrowed the
> > term from patent law), "contributory infringement" can't be found
> > where there is no direct infringement to be contributed to, nor even
> > where there are substantial non-infringing uses of the product and the
> > distributor makes a good-faith attempt to discourage, and avoid
> > deriving substantial revenues from, infringing uses.
> Right. In most patent laws, the act of distribution of a product
> is by itself a direct infringement. Even when the distribution
> is intended for individuals that want to privately use the product.
> So I don't think this is going to help you much.

Right, that's the same as US law AIUI.

> This of course presumes that a sequence of bits is a "product" in
> patent law. I'm not aware of any caselaw either way. But it does
> seem likey that this would be covered by the "essential element"
> clause in patent law. See e.g. article 60(2) and (3) UK Patent Act:

It would clearly be a misreading of the legislative intent to say that
neither manufacturer nor distributor nor end user needs to obtain a
patent license just because the end user isn't making "commercial" use
of the product.

> One could argue that "other person entitled to work the
> invention" can mean "a person performing the acts privately
> and for purposes which are not commercial". I'd have to check
> with a UK colleague whether that's accurate.

Probably not, if the only basis for that entitlement is the statutory
out provided in 60 (5) (a).  But if Thomson actually were to express
the intention of exempting private users of an open source
implementation from the need for a patent license ever to have been
obtained for _their_copy_ -- i. e., if in Thomson's official view
their usage (transcoding other audio formats to and from MP3s for
personal use) is such that neither upstream, nor packager, nor end
user needs to pay a per-unit fee for that copy -- that's a very
different ballgame.  (For DFSG purposes, it would of course also be
necessary to verify that no per-implementation fee was demanded if one
acts in good faith to communicate the limits of this "safety zone" to
recipients, lest downstream developers should be obligated to pay a
lump sum when renaming the package or merging it into something

I'm not saying that I think that Thomson is offering such a deal, but
I'm not saying they aren't (or haven't already) either.  My impression
is that they are focusing on extracting revenues in return for the
mastering of MP3s for purposes of commercial distribution.  It's a
fine point and I know there are some DDs who won't look at it this
way, but if the software authors aren't attempting to discriminate
among fields of endeavor -- only disclaiming responsibility for
obtaining patent rights from a third party on anyone's behalf, and
advising recipients that upstream focuses on the "substantial
non-infringing uses" (personal use and use by patent licensees) that
protect upstream and distributor from allegations of infringement --
then I think there's at least a case for calling it DFSG-free.

I care about this principally because I'd like to see a model "square
deal" that qualifies as free speech whether or not it's free beer. 
Commercial, end-user-oriented MP3 encoders, hardware and software,
compete with one another and with other formats anyway; will they
really be all that pissed at Thomson for giving the green light to
LAME and its like?  As psycho-acoustic research tools, toys for the
ABXers at Hydrogenaudio, and building blocks for product mock-ups
under the umbrella of a patent sharing agreement, LAME and libmad and
ffmpeg and so forth are interesting and useful without cutting unduly
into the revenue stream that amortizes off the cost of research.

If Debian and Thomson knock this domino over, someday even DeCSS may
be blessed by the powers that be.  (Though IMHO the only
non-infringing use it really has is the product mock-up scenario.)

- Michael

Reply to: