Patent Claimed on System-Level Encryption

nattt writes "The Register is reporting that a Californian firm, Maz Technologies has been granted a patent for application independant file encryption, and is now going after other companies with its lawyers to press its claims. It seems that the US patent office doesn't check very well for prior art, and their laxity is causing small firms that get attacked on infringing these bad patents a lot of money to defend themselves."

Comments

Prior-art

[eddy]

The nice folks over at sci.crypt seems to have listed quite a few cases of possible prior art.

Not that that makes it much better on the whole, but...

probably wont happen

[Edmund Blackadder]

secret use is not prior art.

But if the above mentioned agency put their computer on sale, or used it publicly then it would be prior art.

a fine example of patent problems..

[thesupraman]

This would seem to be a good example of how the patent system is being mis-used at present. Apparently this patent is very widely defined and not backed up by much 'implementation'. this would generally not be considered a very 'defendable' patent, yet the owners are trying it on with a bunch of middle level software vendors, trying to strong arm some cash from them.

The difficulty with this is that the patent gives the owner a degree of 'high-ground', and defending against this from the point of view of the apparent patent violator, can be VERY expensive, so often just coughing up is the cheapest option, which then lends weight to the defendability of the patent.

Certainly, being filed in 1998 makes this particular patent laughable.

for exmaple, have a look at:
http://www.cs.auckland.ac.nz/~pgut001/sfs/
w here sfs (Secure File System) exists, and this page was LAST updates in september 1996, and covers just about every possible level of eccryption in a general file system, it is also not unique.

Re:a fine example of patent problems..

[sparcv9]

crypt(1) only offered trivial protection, but it *was* an application-independent system resource that could be used to encrypt and decrypt messages with a classic algorithm.

Not quite. The UNIX crypt() function is a one-way algorithm. A character string is run through crypt() along with a 2-character salt, which is used to mutate the encryption algorithm. The result is a 13-character string, the first 2 characters being the salt. To check, for example, a password against a crypt()'ed password, you take the user-supplied password string and the first 2 characters of the encrypted password, run them through crypt() and compare the output with the previously encrypted string. If both encrypted strings match, then the user-supplied string matched the string that was previously crypt()'ed.

By the way, since crypt() is a system library function , it is in Section 3 of the manual, and is denoted as crypt(3), not crypt(1).

Note: I'm not sayng that this isn't a dumb patent, but your example of prior art is in error.

Clipper

[wiredlogic]

We need look no further than our good friend the Clipper chip for a good example of government created prior art that is covered by this patent.

Re:Here's a thought...

[Frank T. Lofaro Jr.]

Since when does the patent office deny a patent?

They allow patents on things for which prior art exists, even things for which prior PATENTS exist. IBM got a patent on LZW and then Unisys was granted their patent on it - the patent office either didn't check for a prior patent or didn't realize the 2 patents described the same thing.
And Unisys wants us to pay THEM? IBM should be more pissed than they are.

Oh, yeah, they will deny a patent for a "perpetual motion device", unless you are clueful enough to call it something different, like an "overunity device". 1/2 ;)

Here is a GNU page describing the GIF situation and mentioning the 2 LZW patents.

Also, even if someone can afford to fight a lawsuit - the odds are stacked against them. A patent is presumed valid and the defendant must rebut that with a preponderance of the evidence, or lose. This is a reversal of the normal situation in civil law, which is where the plaintiff needs to have preponderance of the evidence. The plaintiff needs to have a preponderance of the evidence that the action alleged to be infrigement occurred - but the defendant has to prove the patent is invalid.

So if the court can't tell who is right on the latter issue, they'll presume the patent valid and you guilty.

Ask a lawyer for legal advice. (if you got tons of cash lying around to pay for one)

References ahoy

[isaac]

Check out The Economist's excellent article, "Patent Nonsense" for a good primer. Sorry, you'll have to pay to view it, but here's a germane excerpt:

Todd Dickinson, the PTO's director, admits that there is a problem here. "In software, in particular, we need to develop our sources of prior art. We've been talking about reaching out to the software industry to get access to more databases," he says. Still, he argues, the problem should self-correct as more software is patented.

What will not correct itself, though, is the bias in favour of issuing patents. This arises partly because America does not have the "opposition" system that Europe has, where the competition can put its case against a proposed patent.

What is more, the way patent examiners are paid encourages the issue of patents. They are paid partly through bonuses for "disposals" of patent cases. But as Robert Merges, professor of law at the University of California at Berkeley, points out in a paper called "Six Impossible Patents Before Breakfast", while a patent issued is always a case disposed of, a rejection may not be, because the inventor can amend his application and try again. Quality, sticking-power and morale among patent examiners is also a problem. The PTO's intake is largely made up of law graduates. Starting salaries are around Dollars 40,000. In east-coast law firms, they are commonly Dollars 140,000.

Here is an anonymous patent examiner on Mr Aharonian's site:

You know what? I'm sick of finding ridiculous patents every time I look [in my files]. Part of the blame goes to the patent corps. We don't fight hard enough against the bull - being shovelled by upper management. And of course, that is where the rest of the blame goes. It's a system that's burning up, and management just keeps adding fuel to the fire. And why should you care? Hey, management pays you for good patents or bad, right? In fact, why should you fight with management? Why reject?

Also check out this article on patent strategies of japanese companies. I do not claim to be a scholar here and freely concede that mmy posts are not scholarly works. That said, they do derive from the cumulative impact of articles I have read in a variety of sources over the years. I invite you to do some more research yourself to dispute my view - indeed, a quick google search reveals plenty of B-school theses fawning over the Japanese tendency towards cross-licensing to the extent that it permits more than one company to extract monopoly rents from a market, though I would dispute (again, admittedly without my own body of scholarly work) the assumption that this is a good thing from a macroeconomic perspective. If anything, I think this would reinforce my point, but since nobody (that I've found) has directly studied the impact of cross-licensing on companies excluded from the "club," as it were, I cannot provide you with an online reference directly supporting my proposition wrt stifling of innovation.

-Isaac