Codec Buddy in Fedora 8 [LWN.net]

Codec Buddy in Fedora 8

Posted Nov 3, 2007 13:30 UTC (Sat) by freemars (subscriber, #4235)
Parent article: Codec Buddy in Fedora 8

Codec Buddy doesn't offer a legal route to encode MP3s or other proprietary formats.

From the size of the scroll bar, it looks like the EULA is long and restrictive. Does it forbid licensing the Fluendo implementation but using FLOSS code? (I suspect so.)

Sad to say, codecs are one of the best arguments in favor of software patents. Making a good codec is work; it's not obvious to the skilled coder. And there's more than one way to get the desired end result -- witness the number of proprietary codecs out there.

That said, it's still just a bunch of math. I shall now turn my laptop off and go flog myself.

(Log in to post comments)

Codec Buddy in Fedora 8

Posted Nov 3, 2007 13:41 UTC (Sat) by spot (subscriber, #15640) [Link]

AFAIK, there is no legal route (in the US) to encode MP3s (on Linux).

Codec Buddy in Fedora 8

Posted Nov 4, 2007 5:07 UTC (Sun) by beoba (guest, #16942) [Link]

Playing devils advocate:

Could you do it through Wine with a licensed Win/Mac application?

Codec Buddy in Fedora 8

Posted Nov 4, 2007 22:42 UTC (Sun) by drag (subscriber, #31333) [Link]

Depends on the EULA of the application, and weither or not the EULA is actually legal. 

Most of Microsoft's tools say things like 'any application you make requires a microsoft
windows license to run' or something like that. Basicly they don't want you to use anything
you make using Micrsoft stuff to run on anything other then Microsoft Windows.

For example:
Years ago there was a fall out with FoxPro developers and Microsoft over this. Microsoft
bought out FoxPro basicly to get FoxPro's customer base. Some of those developers disliked
Windows and did their development on Linux using Wine.. which Microsoft then came back and
told them this was completely illegal. After some fighting the end people figured that while
it was ok to develop their applications in Linux the as long as the end users could only run
Windows to run their applications. (or something like that. It's been a while)

http://www.theregister.co.uk/2003/04/17/ms_legal_threat_d...

That sort of thing is probably happenning with Apple stuff also. I would expect that Apple has
similar things about applications targetting their platform running on Linux.

It's all pretty stupid and is why you either want to be fully legal by licensing stuff from
people like Fluendo..  or just ignore the laws completely and do what you want (like the vast
majority of people)

What you choose to do depends on the sort of liability you would be faced with. :)

Codec Buddy in Fedora 8

Posted Nov 3, 2007 14:57 UTC (Sat) by proski (subscriber, #104) [Link]

Actually, codecs are good arguments against software patents. The encoded data is distributed worldwide, to users of different software. Some users receive both the software and the data on physical media and cannot download additional codecs. The benefit for the society is not to have interoperability issues.

For software patents to be justified, you would need to show that nobody would make a comparably good codec in absence of software patents, and that the advantage of the codecs developed under the software patent regime would offset the disadvantages for the society, such as compatibility issues and inability of others to improve existing codecs.

Mind you, once the patented codecs are popular, it's hard to replace them with non-patented ones. The new codec would have to be substantially better, but it may be mathematically impossible.

Simply put, is mp3 so much better than ogg that all Fedora users have to pay price for that?

Codec Buddy in Fedora 8

Posted Nov 3, 2007 15:33 UTC (Sat) by k8to (subscriber, #15413) [Link]

I may miss your point, but mp3 is not better than ogg.  ogg with an order of magnitude less
development effort has managed to exceed the quality of mp3 within a very impressive bitrate
range.  The nature of encoder tuning is such that quality is a laborious process occurring
over months and years, not just a brief stroke of innovation.  That ogg has managed to do
better than mp3 in approximately the 40kbps to 180kbps range with about 3 developers speaks
volumes.


In its category, mp3 is not an impressive codec.  It is about the worst of its kind.  However
it was the first, and by dint of effort it is about the most polished (in terms of available
encoder quality), and it is pseudo-open due to these factors, open source implementations, and
reasonably good published specifications.

Then there are all the fringe benefits:  efficient silence encode, sane metadata, no onerous
requirements for detection of bit-shift errors by decoders, built-in multi-channel.

So I suppose a support for your footnote about being substantially better.  It is better, but
not enough.

Codec Buddy in Fedora 8

Posted Nov 3, 2007 17:03 UTC (Sat) by proski (subscriber, #104) [Link]

OK, "better" includes "first to come". Suppose the software patents didn't exist. When would a first mp3-quality codec appear? Maybe a year later (just guessing). So, the early adopters would not have a good sound encoder for one year. But, chances are the emerging format would be better supported by the software. Chances are the whole number of incompatible codecs (think wma, aac, ogg) would not exist. Chances are that free software would have good support for all major codes out-of-the-box. And those trying to improve sound encoding would not try to avoid the ideas found in the existing popular codecs, thus making it easier to make a significant improvement that could succeed commercially.

Thus, in the long (5-10 years) term, the public would be better served without software patents, even if the original developers would not be as motivated to come with the initial implementation.

Codec Buddy in Fedora 8

Posted Nov 4, 2007 21:49 UTC (Sun) by man_ls (subscriber, #15091) [Link]

Your example is not valid, but just because the MP3 innovation did happen without software patents. In fact the codec was invented on this side of the Atlantic. Revenue has come largely from hardware implementations (which can be patented), I believe. I don't see great innovation coming from Fraunhofer nowadays, as a result of all that money, or from other licensors.

Codec Buddy in Fedora 8

Posted Nov 4, 2007 23:52 UTC (Sun) by job (subscriber, #670) [Link]

Well, patents or no patents juridically, lots of things surrounding the MP3 encoding algorithm
was somehow patented anyway even in Europe. So it's not a clear cut case of innovation
"without software patents".

That said, I personally firmly believe the world would be much better off without them. The
number one problem right now is the lack of standards, everywhere in the IT sector. And
patents are what slows them and encourages competing standards. Even if we would have had to
wait an extra year for compressed music to appear, it would have been _so_ worth it.

software patents and data formats

Posted Nov 4, 2007 17:22 UTC (Sun) by coriordan (guest, #7544) [Link]

Codecs are the perfect example of why there should be NO patents on software ideas.

Patents are supposed to encourage "lateral innovation".  I.e. by blocking off one way of doing
something, they encourage others to find new ways of doing that same way.  This should lead to
improved ways of doing things.

This is worth zero for data formats.  If the job is to display a certain format, lateral
innovation is impossible.  Writing software that would display some other format simply won't
get the job done and isn't worth working on.

Your argument seems to be based on an idea that patents exist to help people profit from
research.  This isn't the case, but even if it was, it fails for data formats for another
reason: Imagine I invest ten million Euro into developing a data format that's better than
Microsoft's.  If I can patent it, Microsoft will simply not use it.  If no one uses my format,
it and its patent are worthless.

Data formats aren't made valuable by of their features.  They're made valuable by having a
relationship with (or being) those who own the channels of communication.

More info on software patents:
http://ciaran.compsoc.com/software-patents.html