“Why patent trolls love east Texas”

by Walter Olson on December 3, 2013

Mike Masnick on a jury verdict against Newegg: “Having Whit Diffie (who invented public key cryptography) and Ron Rivest (who basically made it practical in real life) present on your behalf, showing that they did everything prior to Jones’ patent, while further showing that what Newegg was doing relied on their work, not Jones’, should have ended the case. But…” [TechDirt; Joe Mullin, Ars Technica]

{ 6 comments }

1 Don Miller 12.03.13 at 12:31 pm

basically, the jury in this case apparently failed in the logical fallacy of appeal to authority. The Government issued this patent, therefore the experts who said the Government was wrong, must be mistaken or lying.

Which isn’t true at all. The Patent Office does very little due diligence in determining whether a patent is novel or unique. It is very very possible that the Government issued, especially in software/tech, patents that are invalid on their face.

Inventor (US Patent # 5,212,540)

2 Richard Nieporent 12.03.13 at 2:31 pm

Once again we see the wisdom of the jury at work.

The lawyers from both teams took in the verdict without emotion and shook hands following the verdict.

Of course, in any lawsuit they are always the winners.

3 Bumper 12.03.13 at 3:18 pm

Having had the pleasure of driving through Marshall not once but twice in the last week, I find it difficult to believe that any lawyer looking to try a complex patent case would want to do so in front of a group of the locals.

I suspect the word in out amongst to good people of Marshall that finding for the plaintiff will continue the largesse of legal tourism, and not to worry for truth and justice because our foolish findings will be dealt with on appeal.

4 Bill Poser 12.03.13 at 3:42 pm

It would be most interesting if someone is able to interview the jurors and find out what their thinking was.

5 Richard Falk 12.04.13 at 4:27 am

See the following post I wrote for why the jury may have returned the verdict they did:

http://arstechnica.com/tech-policy/2013/11/jury-newegg-infringes-spangenberg-patent-must-pay-2-3-million/?comments=1&post=25764027

Though in that post I wrote that the patent was likely valid but that the verdict of infringement was wrong, that may not actually be the case. Though the SSL/TLS specification itself does not directly read on the patent, the renegotiation handshake that can be used to generated new keys does have some implementations that work at least in part in the way the patent describes:

http://pic.dhe.ibm.com/infocenter/wmqv7/v7r1/index.jsp?topic=%2Fcom.ibm.mq.doc%2Fsy11000_.htm

“Renegotiate Period” in:
https://devcentral.f5.com/articles/ssl-profiles-part-6-ssl-renegotiation#.Up7kNpH-jzN

Remember that the patent claims are about changing the encryption keys after a predetermined number of blocks (amount of data transmitted/received) and generating such keys using synchronized pseudo-random numbers generators so that one need not transmit such keys. Though the patent was about creating new keys for block ciphers, the claims are broadly written so that they can apply to changing the seed key for stream ciphers as well.

If you actually look at the textbook the defense claimed was prior art (Denning’s “Cryptography and Data Security”):

http://faculty.nps.edu/dedennin/publications/Denning-CryptographyDataSecurity.pdf

you will find that it only references changing keys with regard to stream patents where the keys change for every bit (if XOR is used from bits in the key stream against bits in the plaintext) and where different keys can be used for different records in a database. It does not talk about different keys for different blocks in a block cipher nor about changing the seed key for a stream cipher nor about generating such keys at both ends using synchronized pseudo-random number generators so that transmission of the changed keys is not needed.

Diffie’s testimony made it sound like RC4 was claimed to be prior art at the same time it was claimed to not infringe the patent — you can’t have it both ways and in reality RC4 has nothing to do with the patent so couldn’t be prior art against it.

From what I can tell, secret key renegotiation exchanges (transmits) the generated keys which is something that would not be needed if the keys were generated as proposed by the patent, but if such keys were generated by synchronized pseudo-random number generators and the transmission was superfluous, then that would still be infringement. At a minimum, it sounds like a gray area.

I’m going to be getting a copy of part of the court transcript and do further research on this — what the press and people posting comments about this say is obviously not the real story. There is so much hyperbole from both sides.

6 MattS 12.04.13 at 9:28 am

Bill,

That would be like trying to determine the contents of a vacuum bottle.

Comments on this entry are closed.