Oracle Ordered to Reduce Claims Against Google From 132 to 3

How amazing would it be to see them deliver it on five A1 pages, just to see the look on the judge's face.

Maybe the judge realizes how much bullshit both parties' lawyers are full of?

Judges have a lot of authority, and he is right, after all.

Exactly, especially since Oracle won't be able to use dropped claims in any case against Android in the future.

I was wondering the same thing. Can a judge really order plaintiffs to permanently abandon claims just because they would be too complicated to try? He hasn't even decided which claims can continue, just says Oracle needs to pick 3. Another oddity is that he's limiting the number of prior art claims Google can bring. Can a judge really limit your allowable defenses?

I've always thought there had to be some merit-based or procedural argument to dismiss any part of a case with prejudice. Also, there's cases all the time with more than 3 patent claims at stake, so clearly there's no absolute rule that a case can't go beyond a certain level of complexity.

I hope someone with actual legal knowledge can chime in.

In the order itself, the words "the Court proposes" along with other language used later in the order imply to me that there has been some discussion/agreement among the parties and the judge.

Something has to be done to reduce the scope of the case. So long as software patents are tolerated, there's going to be an overwhelming load on the patent office and on courts. For them to do their jobs well, something has to give.

Generally speaking, lawyers will throw everything they have at the wall and see what sticks. I believe this is basically taught as official strategy.

However, it is pretty obvious that some of what they are throwing has a reasonable chance of sticking, and some things are just being thrown because why not? It isn't that hard for the parties to sort their arguments by likelihood of success and present just the best, and if you think about how to model that, you'll see that for a vanishing fraction of the work you basically end up with a high probability of getting the same result the hypothetical full trial would have.

Now, combine that with the fact that trials aren't free and the taxpayers are a party in the suit in the sense that they are paying the judge, infrastructure, etc, and also the general fact that if there was an injustice committed by Google every day that Oracle goes without renumeration is itself a further injustice, and you can see that there is a compelling interest in completing the trial sooner rather than later, and in the balance of things if that means jettisoning a series of arguments unlikely to succeed anyhow, this is a net win. "It will take too long to try them all" is a valid concern for everybody; what's the point of the trial if the expenses the trial incurs to Oracle, Google, and the taxpayers are larger than any possible judgment's time-value-of-money could possibly be, to take one limit case?

Also, this will apparently go to a jury. You can't stuff arbitrary amounts of information at a jury and expect results. I would argue that the bound on what you can expect a jury to understand is itself a critical (and underappreciated) component of our system. If 12 laymen can't understand it after several months of targeted instruction, the law is too vague to apply in the first place.

This is speculation, but it could be that it is completely pointless to try all the claims. Patent claims fall into two groups: independent claims and dependent claims.

Independent claims stand alone. Dependent claims depend on independent claims--they are basically special cases of the independent claims.

For instance, suppose chairs were new and patentable. An independent claim might be for "a device for sitting comprising a level surface to support the buttocks and 4 legs". Let's call that claim #1. Claim #2 might be "The device of claim #1 where the legs contact the floor at the corners of a rectangle". Claim #2 is a dependent claim of #1. Claim #3 might be "The device of claim #1 where the legs are made of metal". Claim #4 might be "The device of claim #3 where the legs are covered with cloth".

The complete set of claims in data structure terms forms a forrest of trees, one tree per independent claim, with the independent claim at the root, and the dependent claims as children.

Generally, the independent claims are the most general (and the most likely to be invalidated by prior art or failure to be non-obvious). The dependent claims get more and more specific the farther down they are in the tree of claims that springs from an independent claim.

All it takes is one valid infringed claim to stop the defendant. If he's infringing one of your claims, you get to stop him from making and selling his product without your permission, and you get damages for his past sales. If he's infringing 10 of your claims, you get the same thing. If he's infringing 100 claims, you get the same thing.

So what is the point of trying each and every claim? Pick the handful for which you can make your best case, and concentrate on those. If you can't win those, you probably could not win the others.

It sounded to me like the judge doesn't expect this case to go to trial. He sounded like he fully expects them to settle and this was his way of encouraging that.

From the article:

Judge William Alsup described Edge Games as 'trolling' and suggested that it could face criminal charges

In his order, Alsup even created info-graphics comparing Dr. Langdell's USPTO filings with the actual products, describing in detail "evidence of fraud" that forces the court to play "spot the differences:"

---

This judge is brilliant.

May 16th, Groklaw's anniversary.[0] Though apparently it may continue in some form? [1]

[0] http://www.groklaw.net/article.php?story=20110409161444432

[1] http://www.groklaw.net/article.php?story=20110414184224418

Maybe after they get a time machine and go back to around 2006, when plaintiffs were winning more than defendants there, as opposed to now where it isn't even in the top 5.