Friday, October 8, 2010

Google answer to Oracle patent suit possibly wrong on Java history

Earlier this week, Google filed its answer to the patent infringement suit Oracle started in August. At the time, I looked into certain licensing issues, explained why this represents nothing short of a debacle for the Open Invention Network, analyzed a recent Google submission to the US Supreme Court with a view to Google's position on software patents, and last month I criticized the Free Software Foundation's belated reaction for its grossly misleading content and spam tactics.

When I read Google's response, I was unsurprised and underwhelmed. I tweeted this assessment:

#Google response to #Oracle #Java lawsuit over #Android denies and disputes everything. No surprises at all in the document.

When asked me for comment on Google's defense strategy, I said that Google can obviously afford to deny and dispute everything and stressed that Google "knowingly and willing did what it did", while I didn't see any indication of Oracle having lured Google into a trap. That was my way to comment on one of Google's arguments, the so-called "unclean hands" defense, meaning that Google blamed Oracle/Sun for reproachable behavior.

A short and simple description of Google's defense strategy as a whole is that Google would prefer the suit to be dismissed before it even begins; it tries to have all seven patents declared invalid; if they are valid, Google denies there's an actual infringement; if there's an infringement, Google claims it had the right to use those patents anyway; and if all else fails, then Oracle isn't entitled to anything because it's kind of evil.

David "Lefty" Schlesinger, an open source licensing expert with significant knowledge of patent matters, tweeted his amusement at Google's wholesale denial, which he paraphrased like this:

"Yup, that looks like a patent. Beyond that, we know, much less admit, absolutely nothing."

Again, it's understandable -- and had to be expected -- that Google exercises its rights of defense to the fullest. Others in Google's situation would do the same. But it's unrealistic to assume that Oracle's suit would be dismissed. This case won't go away anytime soon unless Google makes concessions that Oracle regards as a satisfactory outcome. I'm afraid that won't happen quickly, so the two companies will have to fight this out in court. If Google wants to get away unscathed, even a large number of unconvincing arguments won't help. It will need to make some really compelling points, and at least for the time being I don't see those.

I'll keep following those developments, of course. I didn't see any other observers being impressed by Google's defense either. The engadget website published an analysis that accurately points out Google will be in major trouble if only one of Oracle's seven patents is deemed both valid and infringed. I would liken this to a tennis match where your opponent has seven match points at a time. In tennis, the theoretical limit is actually six, and in this legal process, one "killer" argument could theoretically do away with everything, but there's still no reason to assume Google really has one. Therefore, I tend to agree with engadget's prediction that "when all's said and done we'd guess the state of Java on mobile will be very, very different."

Besides a legal defense on all counts and all fronts, Google also raises points in its response that appear to be written for the court of public opinion more so than the court of law. Google talks about how Oracle/Sun never really made good on the promise to open up Java and about what Oracle demanded back in 2007, more than two years before it actually acquired Java as part of Sun.

Open source ethics and expectations aren't a legal concept. Someone's voting record in a standards-setting body isn't tantamount to granting a patent license, especially not a license to patents they get to own only a couple of years later.

My personal opinion is that a filing with a court should respectfully focus on what the court needs to know rather than place a great deal of emphasis on what would more appropriately form part of a position paper to be shown to the open source community.

So far it seems that Google's open-source-specific arguments don't convince this community either. Simon Phipps, a member of the board of the Open Source Initiative (OSI) and the former chief open source executive at Sun, tweeted that paragraph 7 of the "Factual Background" section of Google's filing "also suggests their lawyer is being badly advised" and described it as "not actually factual."

That paragraph 7 mentions the Apache Software Foundation's interest in ensuring compatibility with Sun's Java software. Since Google incorporated some of the code of an Apache project (Harmony) into its Dalvik virtual machine for Android, the Apache-Java connection could be interesting. However, what Google quotes relates to compatibility testing, not patent licensing. At any rate, I am sure that Simon was the Apache Software Foundation's liaison at Sun, so if he (who certainly isn't an Oracle apologist) has a different recollection and/or different assessment of the situation, something may be flawed. Maybe we'll find out more about that some other time. However, like I said before, the legal case will hinge on different issues, particularly on patent validity and patent infringement.

Even if Google ultimately convinced the community that Oracle committed some wrongdoing that's outrageous from an open source perspective (which so far doesn't seem to be the community view, apart from a general dislike for software patents and litigation), they'd be barking up the wrong tree, I guess. Oracle isn't going to be susceptible to community pressure. Oracle is the epitome of toughness in this industry.

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.