Video forums for free software [LWN.net]

video codec patent nonsense

Posted Apr 10, 2008 13:56 UTC (Thu) by rfunk (subscriber, #4054) [Link]

You have to admit there's a difference between having a codec that you know is covered by a patent, and having a codec for which you don't know of any current patent issues.

How about this change:

"Patented video codecs are a big part of the problem, though there are ~~free~~ alternatives (Theora and Dirac for example) that are not known to be covered under any patents, they are not widely used."

By the way, that sentence needs one of its commas (preferably the first) to be changed to a semicolon.

video codec patent nonsense

Posted Apr 10, 2008 15:00 UTC (Thu) by DonDiego (subscriber, #24141) [Link]

Of course there is a difference between a codec where the patent situation is an unknown
unchartered territory and another codec where you can go to a patent clearinghouse to buy
licenses.

Some players in the market prefer codecs where you can get licenses from a clearinghouse,
others prefer those where no such entities exist.  This does not, however, constitute a proof
or even a statement that no patents apply to the latter category.

And neither model protects you from trolls coming along later of course.

video codec patent nonsense

Posted Apr 21, 2008 8:47 UTC (Mon) by Ross (subscriber, #4065) [Link]

You assume that for the ones known to be covered by patents that we know all of the patents
involved.

It's almost the same case, really, except you are prevented from applying copyleft terms to
your implementation because of the patent licensing agreement.

video codec patent nonsense

Posted Apr 12, 2008 10:13 UTC (Sat) by man_ls (subscriber, #15091) [Link]

By the way, that sentence needs one of its commas (preferably the first) to be changed to a semicolon.

That is a known bug in our author. Is there any update on its status, or when a fix will be available?

video codec patent nonsense

Posted Apr 10, 2008 14:20 UTC (Thu) by jake (editor, #205) [Link]

> Given the (ever increasing) amount of patents and the way the patent 
> system works, such proof is unlikely to be produced - ever.

It's worse than that, of course, you can't prove a negative.

My point is that there are free alternatives, not (yet, perhaps) known to be using patented
algorithms/encoding schemes/mumble, but that they aren't widely used.

Based on your comment, it seems like you think free software folks should just give up on
using video codecs at all in patent-friendly jurisdictions.  An overthrow of the patent regime
would fix the problem, but is a ways off, if ever.

jake

video codec patent nonsense

Posted Apr 10, 2008 15:17 UTC (Thu) by DonDiego (subscriber, #24141) [Link]

Of course you cannot prove a negative.  Nonetheless, you can attempt to do more or less
exhaustive patent searches and come to the conclusion that none of the patents you checked
apply to your program, e.g. Theora decoder X.  This has been done for many things, but not for
Theora or Dirac.

I don't think free software folks should just give up on video codecs.  Not at all, I work on
FFmpeg and MPlayer every day.  What we should give up on is centering on those small parts of
the world where these encumbrances are a problem (mostly USA) and those people for whom these
encumbrances are a problem (enterprises, not private users).

What we should give up on is buying snake oil.  Statements like "XYZ is not covered by
software patents" - like all sorts of statements really - are only credible when convincing
proof is presented.  No such proof exists.

What irks me is that perfectly reasonable people like you, Jake, repeat this fallacy as if it
were gospel, when they really should know better.

You could avoid falling into all these traps by modifying your statements when talking about
things like Theora and Dirac.  Do not claim that they are "free", the term is much too
ambiguous anyway.  Say that they are free software (this applies to many others as well) or
state that there are no licensing bodies collecting fees for them.  Just don't state things
for which no shred of proof exists.

video codec patent nonsense

Posted Apr 10, 2008 15:28 UTC (Thu) by jake (editor, #205) [Link]

> What irks me is that perfectly reasonable people like you, Jake, repeat 
> this fallacy as if it were gospel, when they really should know better.

I guess I am having trouble seeing what fallacy I am repeating.  I think you read more into my
statement than was there.  I don't believe the article really addressed the patent situation
vis a vis Theora and Dirac, though it certainly implies that they are less patent encumbered
(and they are, at least as far as is known).

Free is ambiguous, but I think most folks understood what was meant.

> No such proof exists.

Nor can it as we have both agreed.

> Statements like "XYZ is not covered by software patents"

Which is not a statement that the article made.

I think you are reacting to what you have heard elsewhere (possibly including on this site and
written by me) as if it is here in this article.

jake

video codec patent nonsense

Posted Apr 10, 2008 15:41 UTC (Thu) by DonDiego (subscriber, #24141) [Link]

Jake, when you write that

  Patented video codecs are a big part of the problem, though there are
  free alternatives (Theora and Dirac for example), they are not widely
  used.

then you are clearly implying that the alternatives (i.e. Theora and Dirac) are not covered by
patents.  I really do not think I am reading things into your article that are not present in
it.

You could avoid this by either replacing "patented video codecs" by something along the lines
of "video codecs with known patent pools" or simply avoid making a statement about the patent
situation at all or by speaking about the licensing fees (protection money rather) that some
entities collect for its use.

As an alternative you could rephrase the above paragraph like this:

  Video codecs requiring licensing fees for certain uses in certain parts
  of the world are a big part of the problem, though there are alternatives
  that do not require licensing fees (Theora and Dirac for example), they
  are not widely used.

This would certainly be a tad more cumbersome, but you would not make statements which are
misleading and for which you have no proof.  You get my drift.

video codec patent nonsense

Posted Apr 12, 2008 14:58 UTC (Sat) by dvdeug (subscriber, #10998) [Link]

Actually, the BBC has checked for patents on Dirac, and found none.

video codec patent nonsense

Posted Apr 12, 2008 15:12 UTC (Sat) by nix (subscriber, #2304) [Link]

Stop injecting *facts* and *logic* into this flamewar. If everyone did 
that, the Internet would collapse, or something.

video codec patent nonsense

Posted Apr 12, 2008 15:29 UTC (Sat) by DonDiego (subscriber, #24141) [Link]

I'm glad to hear that, but I would like to have a bit more than just your word for it. Can you
please provide a link or similar thing to substantiate this?

video codec patent nonsense

Posted Apr 10, 2008 14:32 UTC (Thu) by gmaxwell (subscriber, #30048) [Link]

Ugh. I'm getting fed up seeing this outrageous FUD repeated over and over again. Mostly by
you, Mr. Biurrun.

Although I'm replying to you my message is really intended for others who might be mislead by
your message. It's clear to me that you are far too deep into propritary codec advocacy for
anything I say to leave an impression.

The popular mainstream video codecs, such as H.264, are extensively patented by known and
identified patents. The patents (frequently) cover novel and fairly non-obvious ideas. The
participants in the patent pools for these codecs are raking in enormous amounts of money and
codec patents are some of the few software patents that bring in much money at all. Various
media coding patents have been enforced multiple times in the US, most of Europe, and the
United Kingdom. Many of the most prominent patent holders in this field are European.

These encumbered formats may also be covered by unknown/unidentified submarine patents, or
other patents held by parties which do not participate in the patent pools. Obtaining a
license from the pool grants you only a license to the pool patents and the pool make no
warranty related to the non-existence of other covering patents. Even if some know about them
they aren't obligated to tell you. You're on your own.

Microsoft found this out quite painfully not too long ago after getting a 1.4 billion dollar
judgment against them for some MP3 technology related patents, even though they were a fully
paid of licensee of the patent pool. (The damages are on hold while they appeal...)

Intentionally unencumbered codecs like Dirac and Theora are either free from known patents or
have had the known patents licensed for perpetual royalty free use by everyone. They have
received a reasonable level of scrutiny by their authors and others, since being unencumbered
is a serious part of the purpose of these codecs. If anyone is aware of a specific
likely-valid patent that covers either of these formats please speak up so that it can be
avoided.

Like the known-encumbered formats it is possible that there exists some submarine patents for
Theora and Dirac.

The current patents systems are an utter mess which create terrible problems. Even though it
is easy to demonstrate that some software has better patent positions than others, as has been
done for Theora and Dirac, It is completely impossible to certify that any piece of
non-trivial software is entirely free of possible patent claims. This is by no means limited
to media codec patents.

In comparing two items it is unreasonable to fault one side for a risk which is common to
both.

Some even argue that the unencumbered formats will necessarily be less subject to less risk
from even the unknown patents because patent problem avoidance was an important goal, and
because submarine patents tend to be fairly broad which would result in a Dirac submarine
patent likely also impacting H.264.

In any case, the world is paying many millions of dollars a year to licenses things like AAC,
H.264, and MP3 encoders. These fees have *no equal* for Dirac, Theora, and friends. Perhaps
this doesn't matter to you. It matters to other people.

I think it is very unfortunate that so many people in the various "open media software"
communities are spending so much of their time on known-encumbered formats and ignoring the
problems until the summons shows up on their doorstep. But this apparent lack of foresight is
a natural consequence of the "scratching an itch" nature of most open development.

That you don't care for open formats is your own decision. I'm glad you're happy with it. But
please stop FUDing the work of people trying to improve your freedom just because you don't
care.

Thanks.

video codec patent nonsense

Posted Apr 10, 2008 16:07 UTC (Thu) by DonDiego (subscriber, #24141) [Link]

Ugh. I'm getting fed up seeing this outrageous FUD repeated over and over again. Mostly by you, Mr. Biurrun. It's clear to me that you are far too deep into propritary codec advocacy for anything I say to leave an impression.

I can assure you, Mr Gregory Maxwell, that the feeling is mutual. However, please note that I am not advocating anything, much less proprietary codecs. I am pointing out that the emperor has no clothes.

Intentionally unencumbered codecs like Dirac and Theora are either free from known patents or have had the known patents licensed for perpetual royalty free use by everyone. They have received a reasonable level of scrutiny by their authors and others, since being unencumbered is a serious part of the purpose of these codecs.

Sorry, this is just a blanket statement, not proof. Where is the reasonable level of scrutiny documented? This is nothing personal. I just like getting some backup to hard-to-believe statements.

What makes you so confident about Dirac anyway? It is very recent, so there are a large number of patents that it will not predate. Also, the field of Wavelet coding is heavily patented...

If anyone is aware of a specific likely-valid patent that covers either of these formats please speak up so that it can be avoided.

What makes you so confident that such a hypothetical patent might not be so broad as to be impossible to avoid?

Like the known-encumbered formats it is possible that there exists some submarine patents for Theora and Dirac. In comparing two items it is unreasonable to fault one side for a risk which is common to both.

Sure, but you explicitly claim that no patents on Theora exist when you speak of "patent-free codecs" and similar things.

Some even argue that the unencumbered formats will necessarily be less subject to less risk from even the unknown patents because patent problem avoidance was an important goal, and because submarine patents tend to be fairly broad which would result in a Dirac submarine patent likely also impacting H.264.

I assume you meant "less subject to risk" and not "less subject to less risk" above. Either way, you are in firmly hypothetical territory now. Hearing some people argue for something is quite the opposite of hard facts.

No, this would not help at all. If such a patent surfaces, I'm screwed. Dirac or H.264 would not make a difference. Possibly I would be even more screwed if I was never expecting to run into patent problems.

In any case, the world is paying many millions of dollars a year to licenses things like AAC, H.264, and MP3 encoders. These fees have *no equal* for Dirac, Theora, and friends.

Correct. And this is quite something. You should clearly state this instead of pretending that you have something you do not have. Then we would have no need to argue here.

I think it is very unfortunate that so many people in the various "open media software" communities are spending so much of their time on known-encumbered formats and ignoring the problems until the summons shows up on their doorstep.

What sort of summons are you expecting? So far everybody has been leading quite a quiet life while hacking, reverse engineering, infringing all sorts of patents and doing things that are claimed to get you into jail immediately. The only times when things really got to the courts was CSS decryption. We won fair and square.

video codec patent nonsense

Posted Apr 15, 2008 2:57 UTC (Tue) by gmaxwell (subscriber, #30048) [Link]

I'd hate to flood the threat after already saying so much, so I'll just speak again to make this small point of correction:

The only times when things really got to the courts was CSS decryption. We won fair and square.

CSS wasn't patented. It was a trade secret, precluding it from patent protection. Any litigation related to CSS would be about anti-circumvention or copyright, but certainly not patents, so it's an entirely orthogonal matter.

More interesting to me is that in the US the only court case related to CSS of which I am aware is Universal City Studios, Inc. et al v. Reimerdes, Corley and Kazan, and not only was there no "we won" in that case, legally it was a total and unmitigated failure. Plaintiffs were demanded to not distribute CSS and not link to anyone distributing it and it was upheld by a US federal appellate court, so the ruling is precedential.

Of course, since CSS was a weak protection scheme which ultimately depended on secrecy for its security, once that particular horse was out of the barn attempting to suppress further redistribution was futile. The same situation simply doesn't exist for encumbered codecs, the licensing of which continues to generate millions in income even though their specifications are publicly available. While someone may well get away with distributing a geek-ware movie player including H.264, or bittorrent distributing bunches of illicitly copied movies in a format with per-use fees, plenty of other people won't. Helping these people is not the primary motivation for free codecs. The medium-sized band or independent movie maker trying to make it big shouldn't have to pay what amounts to a codec tax to distribute their works..., but they do because they aren't judgment-proof and the risk of not playing along is too great. If you don't think Theora or Dirac go far enough I invite you to help out.

Cheers.

video codec patent nonsense

Posted Apr 16, 2008 10:30 UTC (Wed) by DonDiego (subscriber, #24141) [Link]

I was making a somewhat more general point, sorry for being unclear.  It is always claimed
that multimedia hackers will get in all sorts of legal troubles due to patent infringement,
reverse engineering and whatnot. Contrary to the FUD, this has never been the case.

The only cases that actually went to court were related to CSS.  However, DVD Jon won fair and
square.  I actually forgot about the case you mentioned, but later there was Bunner and
Pavlovich and this one went well:

http://w2.eff.org/IP/Video/DVDCCA_case/

This is not about "getting away with geek-ware".  VLC is the single most successful free
software application next to Firefox.  I see it on every Windows desktop I encounter.  And it
is so much better than any of its proprietary alternatives that only play DVDs or a small
subset of codecs.

The use cases you describe are valid, but they cover perhaps 1% of the people using multimedia
software.  The vast majority are private people that will never get into trouble for using
libdvdcss to watch their DVDs or for encoding their wedding videos in H.264.

Well said.

Posted Apr 10, 2008 17:44 UTC (Thu) by GreyWizard (guest, #1026) [Link]

Thank you for posting that.  I was considering writing a response much like it, but it would
not have been as informed or articulate.

video codec patent nonsense

Posted Apr 25, 2008 16:26 UTC (Fri) by KotH (subscriber, #4660) [Link]

Sorry if i have to correct you, but Diego is not spreading FUD or anything.
He is also not someone who is promoting anything proprietary.

Diego and I work together on MPlayer and FFmpeg for years now. We've
done more in the field of OSS video and audio codecs than most
you here. We know exactly what the field looks like, how companies
behave and how much FUD is spread by both companies and uneducated
FOSS activists. So, here some facts for you:

1) So many OSS video players violate thousands of patents, not one
has been sued. Threatened yes (libdts and libvp62 are the only
two cases that come to mind, and that in the over 8 years
I closely follow OSS video players), but never actually sued.

2) Patents on video and audio codecs are so plentyfull and broad that
there is no way that Theora and Dirac (or vorbis of that matter) are not
covered by even one of them. Believe us, we've read quite a few patents
on codecs and know how the world out there looks like. So, before you
claim that they are not covered by any non-trivial patent, prove your
claim. And i don't mean by "xy did a patent survey and didn't find
anything" but by "xy did a patent survey, scanned patents based on these
and that criteria and closely looked at those patents". Anything else is
just FUD.

3) Theora (i haven't extensively tested Dirac yet) is so inferior to
anything out there (PSNR vs compression rate), that you can just simply
forget it. Even MPEG4 Simple Profile beats it. Not to talk about the
commonly used Advanced Simple Profile. Which is even more embarrassing
if you think that snow (a prototype wavelet based OSS codec) was written
in a few months by a single guy years ago, and not touched since and still 
beats Theora. This alone makes Theora obsolete as the market will not pick
anything vastly inferior if there is a better solution available at a
similar price (eg MPEG4 ASP or h.264), especially not if there are
OSS implementations available. Just have a look at VC-1, microsofts
attempt to gain a foothold in the video codec business. They couldn't
deliver a much better performance than h.264 and hence are not much
considered, even though they are very comparable in performance and
under some circumstances even better.

4) MPEG4 & Co are _not_ proprietary codecs. They might be patent
encumbered, but the standards are freely available for anyone.
You can get them "legally" from ISO and/or MPEG (and not too
expensive) or "illegally" by searching for them on google/yahoo.

5) Beside what most people think, for non-commercial use, you
don't have to pay any patent fees at all. Patents cover only
the commercial world. Anything you do at home or in your free
time is not covered by patent law and hence you don't have
to pay anything.

If you still think that you have to prove us that we are spreading FUD,
i personally invite you to visit us at LinuxTag end of May in Berlin.
We can have a long and nice chat there.

umm, when did FUD become all-encompassing?

Posted Apr 25, 2008 18:36 UTC (Fri) by jake (editor, #205) [Link]

> And i don't mean by "xy did a patent survey and didn't find
> anything" but by "xy did a patent survey, scanned patents based on these
> and that criteria and closely looked at those patents". Anything else is 
> just FUD.

I have a hard time seeing how "xy did a patent survey and didn't find anything" can be
considered FUD.  It sounds like it is insufficiently rigorous for your taste, but it is hardly
spreading "fear, uncertainty, and doubt".  It may well be that there was either no patent
survey done or it was done in a cursory fashion, I don't know, but characterizing it as FUD
is, at best, counterproductive.

FUD is not (as far as I know) another name for "incorrect" or "insufficiently rigorous", which
is what seems to be implied here.

jake

Video forums for free software

Posted Apr 10, 2008 17:46 UTC (Thu) by graydon (subscriber, #5009) [Link]

I quite agree. What amazes me further is the habit of posting *forum-like conversations* via
exchanges of video streams on youtube. Not only is it absurdly wasteful -- what in the modern
world isn't? -- but it's much less convenient for both sender and recipient than text. Why
anyone bothers is beyond me.

Video forums for free software

Posted Apr 11, 2008 19:19 UTC (Fri) by bronson (subscriber, #4806) [Link]

How is it absurdly wasteful?  Are you against phone conversations because they're vastly more
bandwidth-intensive than just mailing a letter?

Personally, I'm really happy to see new modes of communication cropping up organically.  It
gives me hope for the future.

Video forums for free software

Posted Apr 11, 2008 19:39 UTC (Fri) by graydon (subscriber, #5009) [Link]

"Against"? Well, hardly. As I said, the modern industrial condition is nothing if not a
massive party of waste. Almost everything we do recreationally, by volume, is voluntary waste.
What sense would it make to worry about wasted bits when we're drowning in wasted physical
goods?

Comparison between phones and letters is not apt as phone calls are just transient energy
signals, not physical objects. Phone calls vs. the same message expressed as dense text codes?
Sure, absolutely wasteful. Why do you think phone networks are so much more happy to carry SMS
over voice streams? They charge you similar "get the message through" rates (say, a dime)
while saving almost all of the cost of transmission. When videophones are the default, heaven
forbid, "voice only" calls will be cheaper too, and text will carry even fatter margins.

Anyways, if it gives you hope for the future, don't let me interfere. I'll stick with text as
long as I can, but I'm a cave-person.

Accessiblity too

Posted Apr 10, 2008 17:51 UTC (Thu) by GreyWizard (guest, #1026) [Link]

Agreed, but text is also much easier to access for those with visual or auditory disabilities,
including many elderly people.  There's nothing wrong with video or podcasts but offering a
separate transcript is important.

Video forums for free software

Posted Apr 11, 2008 0:17 UTC (Fri) by nix (subscriber, #2304) [Link]

Also it eliminates the substantial number of people without broadband or 
sufficiently capable computers (such as, oh, everyone in my extended 
family three links in any direction, bar only me and a cousin: if this 
sort of ratio is common that's a huge numer of people being excluded). 
It's annoying for those of us with slow broadband links (why pay lots 
extra for fast broadband? I don't need it. And lots of people can't get it 
at all.)

Video forums for free software

Posted Apr 11, 2008 7:37 UTC (Fri) by angelortega (guest, #1306) [Link]

And don't forget people which native language is not english. You don't 
need to have many skills in english to (at least partially) understand 
written text (you can pick a dictionary for words not known to you), but 
understanding someone speaking in 'real time' is a very different thing.

Also, you can't search for words in audio.

Video forums for free software

Posted Apr 11, 2008 19:27 UTC (Fri) by bronson (subscriber, #4806) [Link]

The graybeards are out in force today.  :)

With multi-tens-of-MB links commonplace today (more in Europe and Japan than the US, alas),
and >1TB on a 3.5" spindle, who cares how big a video file is?

And, I'm afraid you guys have been watching lots of conference proceedings or something.  Very
few videos are quite so easy to boil down into an easy-to-read transcript.

It wouldn't surprise me if very similar arguments were used against the telephone 100 years
ago!

Video forums for free software

Posted Apr 11, 2008 19:51 UTC (Fri) by graydon (subscriber, #5009) [Link]

"It wouldn't surprise me if very similar arguments were used against the telephone 100 years
ago!"

Oh please. They're not "arguments against" video, rather qualifications on exactly how much we
should expect video to "displace" other things just because it can. And subsequently how much
we need to worry about it as the only way things will be done in the future.

Here we are 100 years later still reading books, writing papers and "talking" by exchanging
text fragments. Why? It's a *good technology* that's much more appropriate than a string of
recorded telephone conversations.

Video forums for free software

Posted Apr 11, 2008 20:19 UTC (Fri) by vmole (guest, #111) [Link]

I can download the transcript for a 30 minute interview in a few seconds. The equivalent video is several minutes. And the problem is not so much download time or storage, but *my* time.

I'm not against video per-se; there are clearly things for which video is the preferred medium. My objection is the trend for projects to use it for things that are just as easily communicated via text.

Video forums for free software

Posted Apr 15, 2008 8:09 UTC (Tue) by appie (subscriber, #34002) [Link]

Indeed, however much important it is to not exclude people lacking the required bandwidth
and/or cpu power, it's the whole issue of efficiency and wasting _my_ time. Talking about
time, it's more circumvent to time the speed in which to take in the information comparing
video to text+pictures.
I think it already has been pointed out, most people are better writers than actors/teachers.
Jotting something down also forces you to rethink and reconsider. Guess it's the 'gray beards'
all over again when I point out that educational institutions and students alike all over the
world still use... books. Yes, video shows up here and there, but mostly when they add actual
value to the process.

I never watch tutorials, howtos, articles in video format. It's much easier to soak up the
information in a decently written article. If helps a lot of people if the article consists of
text plus pictures (screenshots or 'CLI shots'). Still one can time, copy/paste, forward,
condense, (dead tree alert) print & read somewhere else - no batteries needed, comment on,
search, etc. a text article in a much much more efficient manner than a
video/vodcast/screencast/webinar.

And that's not even considering the impact it has on indexing in search engines. Not an
unimportant part of our online lives.

And the person mentioning the telephone: how come we have video telephony for ages now and
almost no-one owns/uses one ? For the majority of people it plain just doesn't add much to the
communication.

I kind of suspect the whole issue of screencasts is technology for the sake of technology.
There IS a place for video, but looking at the majority of the screencasts out there i'd say:
not the place, nor the time.

Video forums for free software

Posted Apr 21, 2008 10:33 UTC (Mon) by robbe (guest, #16131) [Link]

> What's wrong with text?

I like audio features because I can listen to them wherever I am (e.g. in the garden) and
while doing a lot of other things (e.g. household chores). I don't have much use for video,
though.