Monday, November 15, 2010

Google makes weak showing against Oracle: more than three months later, still no counterthreat

On Wednesday (10 November 2010), Google filed its answer to Oracle's amended complaint dated 27 October 2010.

Five weeks earlier, on 4 October 2010, Google had already responded to Oracle's original complaint dated 12 August 2010. I commented on that answer and wrote that I was "unsurprised and underwhelmed" by Google's defense. Looking at the current state of the case, I think everything is still going according to plan for Oracle and this dispute is going to cost Google dearly.

I'm much less worried about Google than about the Android ecosystem as a whole, which could suffer severely because Google fails to protect it. (By the way, I use a Samsung Galaxy S i9000.)

Besides those who are totally in the tank for Google, only the impressionable can conclude from Google's unsubstantiated and unsurprising defense that Oracle faces a serious obstacle. It doesn't.

Oracle has many strings to its bow: seven patents, some copyrights (a subject on which I commented in my previous posting), and presumably more patents in stock should there be a need to up the ante.

The only thing that could realistically redress or even tip the balance would be a serious countersuit. More than three months after Oracle threw down the gauntlet, Google still hasn't been able to mount a counteroffensive. At this stage it's increasingly likely that Google just can't: other defendants struck back within about a month or two of being sued.

For the Android community as a whole and especially for developers whose applications run on the Dalvik virtual machine, Google's toothlessness is reason for grave concern. There's a saying that "the best defense is a good offense". That's also the name of the game in patent disputes. If Google can't play that game, it may be in for a nightmare.

Google still hasn't managed to do what Apple, Motorola and even HTC pulled off

Google's weak showing is particularly disappointing when compared with how all the other defendants in the major smartphone patent suits fight back:

  • Nokia filed a complaint against Apple in Delaware on 22 October 2009 over 10 patents. Seven weeks later, on 11 December 2009, Apple claimed in the same court that Nokia infringed 13 of its patents. If Google acted like Apple, it would already have countersued Oracle in late September or early October.

  • Apple lodged complaints against HTC with the ITC as well as a Delaware court on 2 March 2010. HTC responded with an ITC complaint against Apple on 12 May 2010 (later it also countersued in Delaware). So there were ten weeks between the original complaint and the first counterstrike. At HTC's pace, Google would already have countersued Oracle in mid October.

  • Microsoft lodged complaints against Motorola with the ITC as well as a court in Western Washington on 1 October 2010. Less than six weeks later, Motorola filed its unsurprising countersuits. If Google were Motorola, it would already have countersued Oracle before the middle of September.

It's not just the time line that's worrying about the absence of a Google countersuit against Oracle. It's also the procedural stage. Apple's counterclaims against Nokia were made as part of its answer to the original complaint. HTC and Motorola instigated their suits even prior to their defensive replies. Google has already responded to Oracle's original complaint and the amended one, but merely defensively.

Google has presented a number of defensive theories but may not be able to support any of them

I can appreciate that Oracle vs. Google is such an important and interesting litigation that several media have already reported on Google's answer to Oracle's amended complaint. But I missed two fundamentally important pieces of information in the reports I saw: a reference to the continued and increasingly worrying absence of a countersuit by Google, and a statement of the fact that a defensive theory presented in such a filing isn't necessarily supported by any evidence.

There's nothing unusual about Google presenting defensive theories at this stage without substantiating them in any reasonably specific way. Other defendants do the same thing all the time. But until Google presents facts that really give meaning to those theories, the mere mentioning of those theories doesn't change a thing. What we see in that court document is a skeleton, but it hasn't really been fleshed out yet.

Google's defensive theories per se are just what every other deep-pocket defendant facing a major threat would do: point to each and every defense possible, including absolute long shots that may completely fail to convince the court when the facts are on the table. There's no penalty for that other than the cost of the lawsuit, which is negligible when you have as much at stake as Google does with Android.

Google obviously wants to keep all options open and clutch at every straw while it can. That's normal and legitimate, but a knowledgeable unbiased observer can't be overwhelmed by that. No one would have expected Google to throw in the towel in the early stages of the game.

Nothing unusual about Google's defense

Against Oracle's patent infringement assertions, Google puts forward all the usual defenses that you can find in any of the answers to the other smartphone patent complaints.

For example, every such document I've recently looked at contained an invalidity defense, and every claim of the patents-in-suit being invalid listed every conceivable theory including a reference to 35 U.S.C. 101, the paragraph on patent-eligible subject matter. Everyone seems to try. Google places particular emphasis on that theory but there isn't any particular reason to assume that those Java patents lend themselves to invalidation on the grounds of being too abstract. Also, the Supreme Court's Bilski ruling didn't expand the scope of what is considered too abstract for patent-eligibility: it merely reaffirmed the standards previously applied.

So my best guess is that Google stresses that part of its invalidity defense to curry favor with software patent critics in the open source community. That approach of submitting to a court of law what is actually directed at the court of public opinion was already displayed by Google's answer to Oracle's original complaint in another context (the Java Community Process).

On the copyright side of the case, it's not a major surprise that Google claims that others contributed the relevant program code. In an open source context, one might try that, but in the posting I just linked to I explained that this would only make a gradual difference. It certainly wouldn't be an excuse for continuing to distribute infringing material.

It's also obvious that Google will try to get some mileage out of the fact that Oracle/Sun published certain program code on open source terms. Google will try to leverage that fact in the patent and the copyright part of the case. However, since the code in question was published under the GPL, it's hard to see how Google's code (under the Apache license) would benefit. Those are incompatible open source licenses.

Again, I don't blame Google for not presenting more evidence at this stage. It's just that if it doesn't put forward some real substance later on, it will lose, and at this juncture I haven't seen Google do anything that would surprise (let alone scare) Oracle. When the decision-makers in Redwood Shores looked at Google's answer, they must have been extremely relaxed and thinking to themselves that so far everything was going according to plan. Whatever that plan is -- it could have major negative consequences for the Android developer community -- remains to be seen. Google hasn't thwarted it, and I increasingly doubt that Google will prevent Oracle from getting its way. Whatever that way may be.

If you'd like to be updated on patent issues affecting free software and open source, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents.