PUBPAT Makes Progress Against JPEG Patent

The Data Compression News Blog writes "The US Patent Office has granted the Public Patent Foundation's request for a reexamination of the patent which Forgent Networks is reportedly using to harass anyone that implements the widely used JPEG format. They have already been challenged by many, but PUBPAT had the first concrete case with 'prior art'. In its Order granting PUBPAT's request, the Patent Office found that PUBPAT raised 'a substantial new question of patentability' regarding every claim of the the '672 Patent."

Comments

Could they be sued? succesfully?

[robbak]

Question: If a corporation like this recieves licence fees for an invalid patent, What is preventing the licensees suing them for the money that they have extorted?

Re:Could they be sued? succesfully?

[Anonymous Coward]

The reason the patent holders can keep the money is because the law favors them completely.

Let's say I had a company with reasonable funds (enough to support going to court). I have a patent that looks pretty solid so I ask Sony to pay me a license fee. Sony comes back and offers me a contract that says "we agree to pay this license fee, however, the full amount shall be refunded in the event that the patent is invalidated".

My company would just say "sorry, remove that invalidation clause or we'll sue you for patent infringement and win".

Re:Could they be sued? succesfully?

[Raul654]

"Your question is based on an invalid premise, namely, that there is anything preventing anybody from suing anybody else for any or no reason."
 
There does exist such a mechanism. A court can deem a party to be a "vexatious litigant", at which point they have to get permission of a judge to file suit. On the other hand, these are usually used against people who represent themselves and file many, many frivilous suits.

Re:Could they be sued? succesfully?

[darkmeridian]

It's in the contract. Parties license patents to avoid litigation. They know the patent may be invalid, but it's easier to settle sometimes rather than die trying. They waive their right to recoup the licensing costs if the patents are later invalidated. Exceptions exist if there was inequitable conduct (basically fraud) before the PTO while obtaining the patents. then all bets are off.

The Prior Art:

[TubeSteak]

http://www.pubpat.org/672ReexamOrdered.pdf

Is US Patent No. 4,541,012 to Tescher

Just a short blurb from the reexamination order

"The request details those portions of Tescher as being relevant to claims 1-11 with suggestion that claims 12-46 are either identical to, or obvious implementations of claims 1-11, the detailed claim chart for independant claim 1 is shown on pages 3-4 of the request"

Basically, Tescher preempts claims 1-11 of Chen and claims 12-46 of Chen's patent just repeat 1-11.

More funding for additional work at application

[PornMaster]

I'd really like to see some more resources dedicated in the initial granting process rather than simply cheer the decisions to review. Allowing someone to patent an unoriginal idea contradicts the notion of promoting the useful arts which the Constitution provides for.

Re:More funding for additional work at application

[jonwil]

The answer is to get experts in the field involved in examining patents before they are "rubber stamped".
If people that knew what they were talking about technically (and preferablly leaglly too) were involved in deciding if a patent was valid or not, we wouldnt see so many crappy patents. Enough people would be required to look at it such that people wouldnt be able to say "no, its not valid" because they have a vested interest in being able to use that stuff and not pay for it (also, a simple "no" wouldnt suffice, actual links to prior art or whatever else would need to be presented)

Also, introduce a clause in the rules that says that if a patent is found invalid (either in the initial investigation or later on by a court), the patent holder has to pay up to the PTO.

The idea of "you have to demonstrate your patent somehow" (e.g. for a patent on something like an encryption algorithim, you have to demonstrate working code for it) would also help.

Re:More funding for additional work at application

[mdfst13]

I'm not sure that the problem is even the "smartness" of the people doing the reviews. The problem is that they have about four hours to find a reason to deny a patent.

It's easy enough to sit around /. and say "This patent sucks." It's much harder to encapsulate said suckage into a reason to deny a patent.

In the modern world, I think that it is impractical to expect anyone, no matter how smart, to be able to find the weak spot in a patent in a mere four hours.

IMO, we would be better off going the other way. Make the initial patent application just a listing. Instead of writing Patent Pending, write Patent Requested. Then, when they try to use the patent, the challenge comes from the recipient. Also, patent infringement notices should all be sent *through* the patent office (for another fee). I.e. the workflow would be

File patent application.
Notify others that you feel that they are infringing.
The defendant (who is presumably in the field; otherwise how are they infringing) researches prior art and challenges the patent.
Both sides present their evidence to each other.
If neither has admitted the other is correct, they can then go to a review at the patent office. The loser of the review pays the review fees.

The fundamental problem with the current system is that it is not possible for a single person to perform such a review at the proper level of importance. Some of these patents are requesting multi-million dollar fees. Some would be happy with $10,000. How do you support both uses with a one size fits all fee?

If you are really committed to the idea of an up front application review, perhaps we should reconsider how patents can be used. E.g. what if a patent application also had to include the licensing fees? I.e. I might file a patent and say that users could pay either $10,000 for an unlimited license or $1 per use (with a cap of $10,000; i.e. the 10,001st use would be free). The applicant needs to set

1. The max fee per user ($10,000 in this example).
2. The fee per use ($1).
3. The max to be collected.

Every purchaser of the unlimited license would become a part owner of the remainder of the patent. Once the original patenter gets the max amount to be collected, all the licensers (to that point) now own the patent and may license it to others *up to* the amount that they themselves paid. Once they all have licensed the amount they have paid (or 14 years, whichever comes first), the patent expires.

Now, here's the kicker: base the patent application fee on the max to be collected amount. E.g. 5%. That will determine how much effort the patent office puts into researching the patent. As a result, people who can realistically expect their patent to be the next big thing will pay huge fees up front, will get lots of patent review, and will presumably make back the money in licensing. People who put in $5000 worth of work and are hoping for $10,000 in fees will pay much smaller application fees, get little patent review, but will not cost that much.

I would still prefer the delayed review. I think that it is fairer to all concerned and puts much less burden on the patent office. This is an example of an alternative.

Now Linux can support JPEG!

[Anonymous Coward]

Oh wait, it already does. What's the difference between MP3s and JPEGs, such that we'll violate patents for JPEGs but not for MP3s?

Re:Now Linux can support JPEG!

[jonwil]

The difference is that people who know what they are talking about both legally and techically have said that the MPEG patents are valid (no idea if they have been tested in court or not though) but those same people (including the pubpat people) say that the JPEG patent in question is not valid.

Also, a lot more people have been sued for violating (or made to pay licence fees for) the MPEG patents than have been for this JPEG patent

The difference

[ad0gg]

Mp3s make you deaf
Jpegs make you blind

Prior Art demonstrated on ASCII

[Anonymous Coward]

I'd just like everyone to know, I've filed a claim of prior art in the name of a long deceased relative on ASCII, and every poster here may be in violation of my patent. Please remove your posts at once or I will be disposed to take legal action. Your IP has been logged.

Knowledge of patent encumberance.

[jbn-o]

Neither of the commenters to date take your question seriously, so I'll make a guess: at the time software was developed to encode and decode JPEG, it was not common knowledge that JPEGs were ostensibly patent-encumbered. As far as I know, no unencumbered alternative to JPEG was developed. But it is widely understood among those who deal with these matters that MP3 is patent-encumbered and that we should use and encourage others to use the apparently unencumbered (and higher quality, besides) Ogg Vorbis instead.

Pondering...

[jmorris42]

> But it is widely understood among those who deal with these matters that MP3 is
> patent-encumbered and that we should use and encourage others to use the apparently
> unencumbered (and higher quality, besides) Ogg Vorbis instead.

Yes, MPEG was always upfront that they were pooling patents and doing the RAND thing. But I have a question. When do they start expiring? I remember a VCD like tech (OS9-68K based, Phillips, brain cramp on the name now.... CDI?) in the late 1980's and VCD (MPEG1 video, MPEG1 layer 1 audio) itself not much later. MPEG1 layer 2 was the failed Phillips Compact Digital Cassette in what, 1992? Question is what is the date on the patents, especially of course on MPEG 1 layer 3 audio and MPEG2 video. AC3 audio is probably several years newer so the last part of DVD and HD-TV won't be public for a bit.

I'm thinking we need to find out and start a countdown, much like everyone did for RSA and the GIF patents.

Re:You ins3nsitive cloD?!

[werewolf1031]

what are the differences between png and jpg?

JPEG images are "lossy" meaning that some data (image quality) is sacrificed for the sake of higher compression (smaller file size and thus higher download speeds). PNG's are "lossless", so while the integrity of the image is preserved perfectly, this results in a larger file size. It's a trade-off. For the sake of not alienating dial-up users, I personally prefer JPEG for displaying my artwork online, even though image quality takes a (moderate) hit in the end result compared to the original image.

One other advantage of JPEG over PNG is the near-uniform rendering of JPEG files across all known graphics-capable browsers, unlike the PNG format which renders quite differently from one browser to another, especially with regards to transparency alpha channels. For ex., Firefox renders transparency (alpha) data in a PNG correctly, while Internet Explorer completely ignores the alpha channel, and thus erroneously renders the image with no transparency at all. While the 8-bit (256-grayscale) transparency and transluscency of PNG files are far superior to the all-or-nothing per-pixel transparency of GIF images, the latter has far less universal support than the former, which unfortunately negates many of the PNG format's inherant advantages.

Re:The prior art is a patent from one year earlier

[crazyjimmy]

As near as I can tell, the reason this patent (Tescher 4,541,012) is important is that it has EXPIRED (as of September, 2005... nice timing that). By invalidating the later patent, it basically drops the JPEG format into open domain.

However, let me make this perfectly clear: IANAL (Especially not a patent lawyer).

--Jimmy

Re:offtopic, but you know what I'd like?

[alx5000]

Just sci-fi, fantasy, DYI tech, and gadgets.

Do Yourself In?

You're not one of... those guys , are you??

JPEG expiring this year.

[tbird81]

Well patent 4698672 was first filed in October 1986. So add twenty years, and that's 2006.

The date on the patent document says October 6, 1987. But I'm sure it's 20 years after filing, not processing.

It seems like a last ditch effort to scrounge for money. Because in 9 months in won't matter anymore.

If you get someone pregnant tonight, you could celebrate the birth of your child on the same day as JPEG becomes free (as in no-threats-of-lawsuits).