Patent litigants target DVRs

(continued from previous page)

for still compression. In Forgent's opinion, the JPEG standard, used to compress images in cameras and on computers, infringed the "672" standard.

In the last three years, the company has struck licensing deals with 38 companies, Forgent's Peterson said, including Sony, Nokia, Sharp, Hitachi, Adobe, Macromedia and NEC. In all, Forgent has received more than $100 million in fees from these deals.

Is this a TiVo?

Forgent Networks owns four patents that it claims give it the right to collect royalties on computers or similar devices that record, store and play back video- and audio-transmitted signals. The abstract, or summary, for one of those patents, U.S. patent No. 6,674,960, reads as follows:

A computer-based television signal recording device is provided. This recording device allows for the recording of input streams onto a mass storage device such as a computer hard drive. The playback of individual programs can be done independently of the recording. A user interface is provided which has a list of the recorded television signals, and the user can select for playback any of the listed recorded television signals.

Generally, the royalty fees on devices such as digital cameras, where still compression is essential, are about 1 percent. A digital camera bought for $500, therefore, brings Forgent $5. The royalty on some software products is less, he said.

A lawsuit pending in the U.S. Northern District of California seeks damages from 40 defendants, including Dell. Four defendants have already settled.

Between 800 and 1,600 other companies could become licensees, Peterson added, depending in part on the outcome of the lawsuit.

"We believe that we have received only about 10 to 20 percent of the value (of the patents) so far," he said, meaning that damages, royalties and licensing fees from the patent could reach $1 billion.

A "Markman" hearing to determine the applicability of the JPEG patents in the pending suit will take place in September or October. While the trial would then be scheduled for nine to 12 months later, suits often settle after the Markman hearing. Markman hearings take place before a judge, rather than a jury, and set technical definitions that will later be used at trial to establish whether infringement took place.

The patents expire in 2006 in the U.S. and 2007 in Europe, but claims filed in the next two years will remain valid.

"The patent, in some respects, is a lottery ticket," Peterson said. "If you told me five years ago that you have the patent for JPEG, I wouldn't have believed it."

But he also asserted that no one should shed tears for some of the licensees. The Joint Photographic Experts Group, which set the JPEG standard, knew of the "672" patent while developing the standard, Forgent says. One bombshell the company will reveal at trial is a written note from a Japanese delegate discussing the patent.

"If you don't pay the inventors to be creating, there is a significant bust in the business model," Peterson said.

JPEG itself has hotly denounced Forgent's efforts to collect royalties.

Ironically, Forgent licensed the 672 patent under the MPEG standards committee. It brings in a much more modest $1 million a year to the company, Peterson said.

CNET News.com's Richard Shim contributed to this report.

[tharcod]

Not more examiners...

What's needed is a public review process so that people of the public can review and contest the patent administratively both before and after it is granted.

TiVo is essentially a file system or database (the two are different ways of representing the same concept) that stores video files. It is obvious that you can compress and store video files on disk, and I believe that MPEG-2, and know for a fact MPEG-1, predate 1991, so there's certainly prior art.

We need an administrative process that doesn't require lawyers, etc. to challenge these sorts of patents.

[Earl Benser]

Patent fights are curious things...

The listed Forgent patent seems to be one more example of
people trying to patent either an obvious concept, or a concept
already implemented. The fundamental criteria for a patent are
'new' and 'unique', and the Forgent patent completely fails both
until you get to the microscopic application claims. and there,
even Forgent noted that the application possibilities are obvious
to generally qualified people.

I think that the Patent Office needs to quit handing out patents
like aspirin and start focusing on serious evaluations of all
applications. So far, they haven't evaluated hardly anything.
Check the drup patents for an oversupply of proof.

[ryan.steele]

You have to be kidding?!?

I can't believe that company's have the gull to file for a patent years after a company has a product out! The Patent says it was filed in 2002 TiVo has been out since like 98-99. Just ticks me off!

[NWLB]

Tis the new coin of the realm.

Its been talked about for some time now. This is going to be how some companies make money, by going after those that did what they couldn't.

Xerox ought to sue Apple for their "obvious" rip-off of their old integrated computer system.

NWLB
****
http://www.nwlbnet.blogspot.com

Forgent Sounds Like a Leech

So let me get this straight: said company, namely Forgent, files patents, does nothing with them, waits until someone comes out with a product that remotely resembles Forgent's "patents", then Forgent strong-arms the companies who are being innovative and productive for royalties?

Forgent sounds like nothing more than a leech. Must be a great business: let's go into the office today and see who we can sue.

Creeps.

[Homer J. Simpson]

What goes around.....

Although I can't stand the patent process and feel that the abuse of it is stifling innovation, I can't really feel sorry for Tivo.

It looks like they are about to get a taste of their own medicine.

Full disclosure: I own a Tivo.

[My-Self]

Fix the patent system ?

Easy :
- When a patent attack is lost, the patent owner should pay the defendant fees (would stop most frivolous patent lawsuits while not hurting valid ones)
- Patent examiners should be real professionnals in the field, able to distinguish the gems among trivial claims (for that to happen, they need to be paid more, not from public funds, but from those who apply for patent protection).
- the USPTO (and similar offices all around the world) should be legally accountable when they accept frivolous patents
- When a patent lawsuit arises, a patent review by the issuing body should be mandatory, before the court can proceed.
- Patent application should not be kept secret until accepted
- Participation in a technical commitee should automatically void patents issued later on the standards created by said commitee (prevent "midnight patent")
- Hiding relevant prior patent application while participating in a technical commitee should automatically void said patents (prevent "Rambus DDR" patents)
- Just like trademarks, patents should become unenforceable if the holder fail to warn known potential infringers (prevent GIF style late claims)
- A patent not used by it's inventor should loose it's offensive power after a short period of time (can't attack others for using something you patented but you're not using).
- More difficult : create different classes of patents depending on their merit & the applicable field (different protection length)
- When a patent holder have claims on more than one patent against a defendant, he should be forced to merge all the claims about all of his patents at once, instead of being allowed to launch lawsuits one after another
- outlaw "evergreening" (last minute "new uses" for patents about to expire)
- Software patents should be either eliminated or severely restricted since the appropriate protection for software exists in copyrights laws.
- "composite" patents that simply put two already known techniques together should not be accepted. Only the way to "glue" them should be acceptable, if non trivial / not obvious / not overly broad.

If only half of these proposals was implemented, the system would return to sanity, but it's unlikely since patents are an anticompetitive tool used mostly to ensure the issuing country's domination in the field rather than the neutral inventor rewarding system it's supposed to be.

Your solutions.

- "When a patent attack is lost, the patent owner should pay the defendant fees (would stop most frivolous patent lawsuits while not hurting valid ones)"

In many cases, they probably do. In virtually every civil case, there exists the possibility of one side being ordered to pay the other's legal fees. It simply isn't a sufficient deterrent, however, and the risk-return formula still favors trying, even on frivolous cases, since there seem to be enough wins.

- "Patent examiners should be real professionnals in the field, able to distinguish the gems among trivial claims (for that to happen, they need to be paid more, not from public funds, but from those who apply for patent protection)."

This *is* how it works, or how it is already supposed to work. The patent office has lots of engineers and other highly qualified people, but they lack enough software experts, are understaffed, and plagued by an old IT infrastructure that makes it difficult for examiners to find prior art. This desperately needs to be fixed.

- "the USPTO (and similar offices all around the world) should be legally accountable when they accept frivolous patents"

Not a bad idea, actually. Except, of course, that we pay the taxes that would fund their defense and any rewards they would pay, and it would diminish their budget and ability to do things right. . .I suppose it really isn't that good an idea, after all.

- "When a patent lawsuit arises, a patent review by the issuing body should be mandatory, before the court can proceed."

This one makes sense. The review should be done better the first time, however--the current process relies too much on the applicant both being honest and doing his own homework finding the prior art that could invalidate his own patent--now, why would someone do that?

- "Patent application should not be kept secret until accepted"

Agreed on this point. Companies should be given a chance to submit possible prior art. Still, it doesn't matter unless they are released to the public in a way that makes them easily accessible.

- "Participation in a technical commitee should automatically void patents issued later on the standards created by said commitee (prevent 'midnight patent')"

Fully agreed, unless, of course, the committee itself applied for the patent--this option should remain open for committees to file patents as a defense against possible lawsuits.

- "Hiding relevant prior patent application while participating in a technical commitee should automatically void said patents (prevent "Rambus DDR" patents)"

Agreed. That's just a dirty trick.

- "Just like trademarks, patents should become unenforceable if the holder fail to warn known potential infringers (prevent GIF style late claims)"

I don't know about this one--maybe if it can be proven that they did it willfully, but I don't think this is necessary.

- "A patent not used by it's inventor should loose it's offensive power after a short period of time (can't attack others for using something you patented but you're not using)."

I would offer this alternate suggestion, similar to a good idea posed for copyrights; use a renewal system. For patents, I would make it yearly. If the patent holder doesn't care enough to pay a $5-10 fee to renew it yearly, then they shouldn't be holding it. For copyrights, I say twenty years is better because of their longer durations and the different nature of what they protect.

- "More difficult : create different classes of patents depending on their merit & the applicable field (different protection length)"

Not a bad idea, but I think your later suggestion is better. It may simply have been a bad idea to make software patents legal in the first place, and eliminating them may be the better option. We need to consider the benefit to society for these patents outweighs the cost to society; the benefit to the holders shouldn't be the concern. I think the cost has long past outweighed the benefit.

- "When a patent holder have claims on more than one patent against a defendant, he should be forced to merge all the claims about all of his patents at once, instead of being allowed to launch lawsuits one after another"

Kind of like protection against double jeopardy? Maybe, but this would only be good where the patents were highly similar in the first place, and the fact that many patents are so similar is a symptom of the broken process. Patents are supposed to be novel, unique, and non-obvious--the fact that four separate patents are involved in this suit tells me that isn't happening. There should only be *one* related to JPEG-style image compression, if that. If that gets fixed, this step wouldn't be necessary.

- "outlaw 'evergreening' (last minute 'new uses' for patents about to expire)"

Wow. I hadn't heard about this. It seems to be the spirit of patent law should already prohibit this--again, the inventions must be novel, unique, and non-obvious. New uses for existing inventions seems to violate most of these. I also remember reading that obvious combinations of existing inventions are not supposed to be patentable, so it seems like this should prevent this practice as well. I would say it isn't because the patent office is sloppy, especially on software.

- "Software patents should be either eliminated or severely restricted since the appropriate protection for software exists in copyrights laws."

I've long since agreed on this one. Software has become too much of a legal quagmire--a cesspool of legality, if you ask me. I don't see how creating the risk of blinding stumbling over someone else's patent, or look-and-feel, or whatever else the courts and Congress have messed up for us programmers is supposed to aid innovation. Wouldn't a feeling of freedom and security be more productive to developers than this mess? I certainly think so.

- "'composite' patents that simply put two already known techniques together should not be accepted. Only the way to "glue" them should be acceptable, if non trivial / not obvious / not overly broad."

These already shouldn't be accepted, if I understand the law properly. Clearly, unqualified examiners are letting these slide through, at least in the software arena.

"If only half of these proposals was implemented, the system would return to sanity, but it's unlikely since patents are an anticompetitive tool used mostly to ensure the issuing country's domination in the field rather than the neutral inventor rewarding system it's supposed to be."

I'm beginning to think you're right, but even with this, they seem to do more harm than good, at least in quickly developing technologies.

At the very least, I think their length is too long. The same can be said for copyrights. Patent terms should only be about 5 to 10 years; perhaps patent examiners should be given the power to set the patent's length based on the state of the particular industry, but obviously the problems of understaffing and inadequate expertise would have to be fixed first, or this would probably only worsen the problem.

At any rate, I can say this much: either fix software patents, or make software unpatentable. One of these solutions needs to be done soon.

[System Tyrant]

Patent

I think that if a patent is filed incorrectly the patent should stand. Then those who are wronged should sue the patent office. Oh, I know it won't work, but it sure would be funny to see the patent office get hit with thousands of lawsuits because they can't do their job right.

On the other hand the patent office is probably understaffed and overpaid. Maybe it's time to do away with patents all together. So I suppose solution #2 would be to pester, er I mean lobby you local congress person till they do away with patents.

I don't think that the fair market place has ever been fair, but know the way to make a buck is to sue someone over a vague idea. I think patents are like reading nostradamus. It only really make sense after something has happened.

Here are three things our government needs to get rid of...
1. Patriot Act.
2. DMCA
3. Patents

This list could go on for days, but those to me are the most problematic for the tech industry.

Unfortunatly I don't have the answer, but I do have ideas (unless somebody has patented it). It ain't much, but when you put together my ideas with yours we might just start to see a picture. We are a country of the many so why are we led by the few?