Fluendo MP3 GStreamer Plugin in Main for Dapper?

Eric Dunbar eric.dunbar at gmail.com
Sat Dec 31 17:46:08 GMT 2005


No desire to start a flame war...

On 12/30/05, Stephan Hermann <sh at sourcecode.de> wrote:
> Hi Eric,
>
> On Friday 30 December 2005 15:10, Eric Dunbar wrote:
> > On 12/29/05, Stephan Hermann <sh at sourcecode.de> wrote:
> > > [...]
> > > It's all about patents. Patents are evil in this case, that we're not
> > > allowed to freely use them for non-commercial purposes. It would be wise
> > > for Frauenhofer, to make a commitment to the opensource software world,
> > > and have a special treatment for the MP3 patent for (F)OSS.
> >
> > Are the patents evil simply because... people can't do whatever they
> > feel like doing because someone else was ingenious enough (and brave
> > enough to devote resources to the project) to go out and develop
> > something new, or are these patents evil because they were improperly
> > granted?
>
> No, but software patents are useless. Take for example this "one click" patent
> stuff from amazon. It's a simple matter of fact, that others can have the
> same idea, a different implementation, but even then they would have problems
> with this.

It is a bad patent. Plain, and simple. Patents can be valid and be
bad, but that doesn't make all patents bad. AFAIK, the MP3 patents are
_valid_ and they are not "bad".

> > People want want want but these patents are valid*. Just because
> > something is popular doesn't mean that patents should be violated
> > willy-nilly (of course, many MP3 fans have no qualms about stealing
> > <ahem, borrowing> music (simply because they can... we don't steal a
> > TV simply because we can) so it's not surprising that the this
> > attitude goes hand-in-hand with the idea that it's right to violate
> > these patents).
>
> Well, I don't read now your statement, that you think that all people who are
> using MP3 for their music are thieves. I for myself am using mp3 since ages
> to convert it to listen to my legally bought music, which is more then 600
> audio CDs now and over 300 vinyl discs, on my legally bought mp3 player.

As you could have VERY EASILY read, I did NOT state that _all_ (or
necessarily most) MP3 users are thieves (just a very significant
portion).

> > Plus, keep in mind that patents expire. IIRC the MP3 patents won't be
> > around for that much longer, so, like JPEG and GIF before that,
> > eventually the algorithms will fall into the public domain!
>
> Well, same here...PNG is a result of a poorly implemented standard like JPEG
> or GIF. If there was a possibility to improve JPEG or GIF (without any
> software patents), I don't think we would have ever seen PNG in this form.
> But JPEG and GIF came first, and now they are, so to say, the industry
> standard.

This is merely an argument used to demonstrate why "open" formats are
effective in that they allow for extensibility and refinement. What
that fails to acknowledge, however, is that R&D is necessary to
develop these ideas in the _first_ place (and, patents are a necessary
outgrowth of R&D to protect the investment that goes into it).

Yes, patents DO stifle innovation at times, but, they also help it.
It's a trade off. Throw out the patent system. You will be able to
bring the few new ideas that are developed to the market much more
rapidly than at present, and see a more rapid development rate on
older, recent technologies. However, you lose out on big ticket items
-- no protection = no incentive to spend lots of money on R&D.  If you
want to ensure major returns on investment for innovators, and stifle
innovation make the patent system stronger. This'll allow big money to
spend big on R&D (with the expectation of winning big on the balance
sheet), but, it'll prevent new technologies from being widely adopted,
especially in novel ways.

It's all a trade off. You can't have your cake and eat it too. No
patents are bad*. Too many patents are bad.

*Though, given that software is being turned into simple commodity
status (as the start-up, commercially successful Linux vendors/support
cos demonstrate), the issue of patenting in software is becoming less
important.

> And about the time: http://www.mp3licensing.com/patents/index.html, when I
> remeber a patent is valid for 20 years? So some of them could expire in 2010,
> e.g. the digital encoding patent.

It depends on the jurisdiction, the type of patent, from when the
patent is valid and when the patent was filed. Some are 17 years, some
20 (possibly from date granted or date applied for patent, etc.).
Also, MP3 (and the other MP*s) are complex beasts encompassing a
number of patents so different parts will expire at different times
(thus, encoding and decoding will become universally available at
different times).

> That's why I don't think that Vorbis is completly free of the software patent
> issue of MP3. Because somehow the patents are not only descibing a way to
> implement the format, but as well give an explanation what should it do and
> for what reason. So, if someone comes up with a similar idea, but with a
> different implementation, it would fall under the same patent. We see it
> everyday in our usage of the computer.

It's been a long time since I've read up on this stuff, but IIRC there
are markedly different tests for determining when and why something is
patentable and when something is merely copyrightable.

> The idea behind patents were, that if someone created a tool or machine or
> some technical new gadget, and another one wanted to produce this for his own
> commercial use case, the original inventor can give those people the right to
> implement it the way he created it.
> But in terms of software, this is not easily possible, because there are
> always different ways of implementation. And an idea is something which
> shouldn't be patented at all.

Software is tough, but, arguable, there must be a certain level of
protection given to R&D to ensure that real R&D happens. Most
important patents are expiring soon anyway (there has been little
_real_ progress in what users can do with computers since 1984 (except
for fancy media and web)).

Eric.



More information about the ubuntu-devel mailing list