Mailinglist Archive: opensuse (2063 mails)

< Previous Next >
[opensuse] MP3/swpatents (was: a user call for attention on SuSE's mp3 support issue)
  • From: Pascal Bleser <pascal.bleser@xxxxxxxxx>
  • Date: Thu, 01 Feb 2007 13:47:15 +0100
  • Message-id: <45C1E153.9060906@xxxxxxxxx>
Juergen Weigert wrote:
> On Feb 01, 07 07:02:00 +0100, Pascal Bleser wrote:
>>>> <http://en.wikipedia.org/wiki/Ogg#Ogg_codecs>
>>> So then those who preach ogg from the hill tops are still
>>> not necessarily any more free of patents until they prove
>>> that speex and vorbis are not violating someone's patent?
>
> It is actually the other way round: Those who claim to have
> patents on speex or vorbis must prove that their claims are valid.
> Unless that is proven a patent isn't worth anything.

Yes and no. In theory, that's obviously true.

But in practice, when patent trolls want to cash in, it doesn't really
matter. Software patents are a ludicrous idea in the first place but
the worst part is that the patent offices (both in the USA and in
Europe) don't make the necessary verifications when they validate a
patent.
It has to be proved (or not) in court, even for prior art.
That's also _almost- acceptable in theory but again, in practice, such
cases can last for months or years, and it gets really, really
expensive (which is why the software patents don't "protect David from
Goliath", unlike what the pro-software-patents lobbying politicians
from UK/Ireland say again and again, thank gawd their EU presidency is
over).

I don't know for the US, but the EPO (European Patent Office) only
makes money from the patents that are filed and accepted, they don't
have any other source of revenue (and no, they're actually not an
official body of European Union). Hence, it's pretty obvious why they
accept anything without properly validating.

>> for (int i = 0; i < 10; i++) {
>> std::cout << i << std::endl;
>> }
>>
>> I'm sure this violates a software patent as well.
>
> not so sure...

Oh really ;)
http://news.com.com/2100-1001-836322.html
http://news.com.com/Kodak+wins+Java+patent+suit/2100-1014_3-5394765.html
http://news.zdnet.co.uk/itmanagement/0,1000000308,39168864,00.htm

"Kodak's case centred on three patents that it bought from Wang
Laboratories in 1997, several years after Java was created. These
patents -- numbers 5,206,951; 5,421,012; and 5,226,161 -- referred to
the integration of data between object managers, and between data
managers, and to the integration of different programs that were
manipulating data of different types.
Kodak argued in court that these patents covered the method where an
application "asked for help" from another application -- such as in
Java's object-oriented programming."

Great example of a patent rip-off.

>> umm... hang on for a minute...
>> /me checks...
>> /me files a patent for the for loop and gets rich !!!
>
> That may get accepted, but it will be most definitly an invalid patent.

Sure, prior art all the way. But still, it could potentially be
accepted (I mean, Amazon's one-click-buy patent [1] isn't any better)
and if I was a business with lots of cash and lawyers, I could sue
<smaller company that eats a little of my market share> and suck out
their cash reserves on law suite costs.
Oh, and as that smaller company is probably filed at some stock
market, it would cause fear and loathing and their shares would drop
really hard.

[1]http://cse.stanford.edu/class/cs201/projects-99-00/software-patents/amazon.html
http://en.wikipedia.org/wiki/Amazon.com#Patent_controversies
http://www.oreilly.com/pub/a/oreilly/ask_tim/2000/amazon_patent.html
And yes, it has been patented in Europe too:
http://swpat.ffii.org/news/03/amaz0818/

>> No, seriously, Vorbis, FLAC and Theora probably violate patents.
>
> Here We can assume that these 'probably violated patents' are invalid.
> Can't we?

Sure, we can ;)
Unless someone proves in court that they do violate some patent, they
don't.
But given that Frauenhofer & friends have filed a lot of patents in
that area, it is quite possible that there is room for a law suit.
From them or from others.
MS has some proprietary codecs too, no doubt they've filed patents on
these.

Note that I'm _not_ saying: don't use Vorbis/FLAC/Theora because they
probably violate patents.
I'm just saying that software patents are an absurd and dangerous
Damocles sword over the head of software development (both closed
source and open source).

> [...]
>> That's exactly why software patents are a ridiculous (but extremely
>> dangerous) idea in the first place.
>
> No doubt about the 'ridiculous', but dangerousness becomes relative, once
> you strip the FUD off.

Not so sure. It only got a little more quiet now because we entered
the "cold war of software patents", where larger businesses file
patents like crazy just to have a retaliation weapon -- IBM (they have
the largest software patent portfolio of them all), Microsoft (I sure
don't like them but they're not patent trolls, fortunately), Oracle,
Amazon (*g*), etc etc...

E.g. if MS or whoever was to file a lawsuit for software patent
violation against Novell or Redhat, there would be a lot of IBM and
Sun and others to step up against MS, waving the threat of a few
lawsuits against them using their own patent portfolio.
That's what is currently sort of balancing the situation out, but for
how long...
And btw, they wouldn't step up because of their love for OpenSource or
Linux, but because they have large investments and partnerships.

>> Still, with MP3, the question isn't even open. It's pretty clear, and
>> it's damn restrictive.
>
> Right. Thomson and Fraunhofer have tried hard to demonstrate the
> validity of their patents...

Indeed
http://bladeenc.mp3.no/
http://www.northcode.com/forums/archive/index.php?t-5393.html
http://news.bbc.co.uk/1/hi/technology/5312696.stm

And J├╝rgen, I know you know all that, it's more meant as a piece of
information for those how haven't had the joy (?) of taking a dive
into the topic of software patents ;)

cheers
--
-o) Pascal Bleser http://linux01.gwdg.de/~pbleser/
/\\ <pascal.bleser@xxxxxxxxx> <guru@xxxxxxxxxxx>
_\_v http://www.fosdem.org http://opensuse.org

< Previous Next >
Follow Ups