I was on the road in Boston late last week, and thus was unable to easily write something up on the April 15 jury finding in the case of Bedrock Computer Technologies, LLC v. Softlayer Technologies, Inc. et al.
That's the catchy name for the patent infringement lawsuit launched in 2009 by Tyler, Texas-based Bedrock against Softlayer and CitiWare Technology Solutions, LLC, two Texas-based software companies, and a few firms that are decidedly not from Texas: Google Inc., Yahoo! Inc., MySpace Inc., Amazon.com Inc., PayPal Inc., Match.com, Inc., AOL LLC and CME Group Inc. The suit alleges that a patent that Bedrock owns, US 5,893,120, is infringed by the defendants in the suit, because such a method is employed by the Linux operating system and as major users of Linux, the defendants are liable for damages.
Back on April 15, after a five-day jury trial, the Federal jury in Tyler, Texas indeed found in favor of Bedrock and specified that Google owed the company a huge, staggering amount of $5 million in damages. (Yes, that was sarcasm.)
Patent law followers will note the location of the trial venue. The United States District Court Eastern District of Texas is well-known as a favored district for patent infringement suits. It is no coincidence, surely, that Bedrock's founder David Garrod opted to start his company, which exists only as a holder of patents like 5,893,120, in such a patent-friendly location.
If you read the coverage and the various blogs regarding the patent decision, it is clear that 5,893,120, methods and apparatus for information storage and retrieval using a hashing technique with external chaining and on-the-fly removal of expired data, is something that's actually been around for some time, despite being patented by Pace University Associate Professor Richard Nemes in 1999. At least, that's what most people with any level of programming knowledge claim--I have to go with that as a claim, since I don't know programming well enough to say otherwise.
What is interesting to me is that despite a lot of people screaming that 5,893,120 is the weakest piece of IP to come down the pike, Google still lost. Now, that says one or more of three things is going on here: East Texas juries really are the easiest to convince that patent infringement exists, Bedrock has some really good lawyers, and/or Google needs some new lawyers. It still does not convince me that actual infringement exists, mind you.
Nor will it convince many others, despite the cackles of glee from some quarters who are predicting a whole new industry of Linux licensing is about to start again. For one thing, Google will appeal and the rest of the plaintiffs haven't had their turns yet. (Well, nearly: Match.com and Bedrock apparently settled their part of the case on March 28 for an undisclosed amount.)
For another, Red Hat filed its own declaratory judgment (DJ) suit against Bedrock in December 2009 to get 5,893,120 declared invalid. In June 2010, Red Hat tried to unsuccessfully intervene in this case, too, which Bedrock fought, claiming "this case isn't about Red Hat" and besides, Red Hat had already filed the Dec. 2009 DJ suit. But while Bedrock was successful in its attempts to stave off Red Hat, that DJ suit against 5,893,120's validity could blow this whole thing out of the water.
Observers of the East Texas court have noted the low amount of damages as another potential weakness in Bedrock's case. Getting only $5 million when asking for much more could be a sign that the jury, while agreeing with Bedrock, didn't think they had much of a case.
So, while notable as yet-another-reason for patent reform, I wouldn't set too much on this jury decision yet. And if I were a Linux detractor, I definitely wouldn't get too excited: if somehow proven valid, the methods described in patent 5,893,120 are found in nearly every operating system. Look out Windows and OS X users, you'll be next.