Wednesday, June 30, 2010

Versata complains against SAP: patent aggressor demands interoperability, requests antitrust intervention

Today I read about a new EU antitrust complaint in which the important issue of interoperability is raised: a US software company named Versata has lodged a complaint with the European Commission, accusing SAP of "withholding information necessary to interoperate" with its enterprise resource planning (ERP) software. Versata would like to make its pricing software named "Pricer" interoperate with SAP's core product R/3.

This antitrust complaint isn't the first clash between the two companies. Previously, Versata sued SAP for the infringement of five patents. According to TheRegister, Versata "sapped" $139 million out of SAP.

The complaint may be legitimate per se

I don't know the details of Versata's antitrust case against SAP, so I can't (at least not now) offer an opinion on whether its antitrust complaint is well-reasoned.

Even in the event the complaint doesn't have merit in a legal sense, I do believe that all significant market players -- including SAP -- should allow others to interoperate with their products. This means making the necessary information available, offering commercially reasonable licensing terms to everyone, and in some cases that may also have to include the unbundling of components (making components that are marketed together available as separate products).

One of Versata's demands is that SAP be required to unbundle its own pricing software from its ERP product. I don't know yet whether that's a reasonable thing to ask for in this case. But I'm convinced that there are indeed many situations in this industry in which an unbundling is indispensable for interoperability.

On the surface it seems that there may be some parallels between Versata's demands and those of French open source startup TurboHercules, which filed a complaint against IBM with the European Commission in March, asking for an unbundling of IBM's z/OS mainframe operating system from the mainframe hardware.

Versata bases its complaint on the underlying logic of the EU's decision in the spectacular Microsoft case. I don't have enough information to be certain that the related reasoning is applicable in that particular case, but it may be at least in part. What I know more about is the situation in the mainframe market. I took an interest in that issue because of IBM's patent aggression against the 11-year-old Hercules open source project. I'm convinced that TurboHercules's complaint against IBM does have very clear and strong parallels with the original Microsoft case.

Versata's patent infringement suit against SAP

I'm still trying to find out more about the patent dispute between Versata and SAP. I don't like any kind of patent aggression. I just make a distinction between different ways in which patents are used and consider strategic patent holders with exclusionary objectives the worst problem, even worse than non-practicing (often also called "non-producing") entities ("trolls").

The information I've obtained so far doesn't make it clear whether Versata would have been willing to grant a license on commercially reasonable terms to SAP. If they made such an offer and only went to court as a last resort, that's one thing. If all they wanted from the beginning was an injunction, then that would make them fall into the worst category of patent holders.

Glyn Moody saw an SAP position paper arguing for software patents and, understandably, expressed hope that losing to companies like Versata might make SAP learn the hard way that software patents are a bad thing on the bottom line.

I understand Glyn's perspective, and I agree on that particular aspect. Yes, if a company advocating software patents gets sued over such patents, the more they end up having to pay, the better, because that may be the only way they could ever change their political position on the subject. That's my feeling concerning any company: it doesn't matter whether it's Microsoft, Oracle, SAP or IBM, or even smaller ones who support the patentability of software.

However, the best outcome of any software patent infringement suit in my view is that the aggressor ends up losing his patents. Once patents are asserted in court, counterclaims aiming to invalidate the patents in question are routinely a part of the defendant's strategy.

Patent busting (taking out individual patents) isn't good enough to solve the problem as a whole. But if there's litigation, the best news is always if the patents are thrown out.

There are different legal reasons for which a patent can be invalidated. The best outcome from the anti-software-patent point of view is invalidation due to subject matter. This means that a software patent gets thrown out just because it's software (no matter how "good" the patent is in all other respects). Unfortunately, that doesn't happen too often anymore. It hasn't happened in the US in decades (this week's Bilski ruling certainly couldn't be used as a basis to invalidate even one existing US software patent), and European courts increasingly also uphold software patents.

If a court doesn't object to software being patentable, there are still other reasons for invalidation. The most important ones of those are "non-novelty", meaning that by the time the patent application was filed, someone else had already invented the same thing ("prior art") and either published it or filed for a patent, and "lack of inventive step", meaning that the prior art that is presented to the court isn't exactly the same but the difference between the prior art and the new patent application isn't sufficient to justify a (new) patent. That kind of invalidation doesn't address the problem of software patents per se, but it's still much better than no invalidation at all.

Defensive use of patents

There are only two exceptions in terms of scenarios in which I would keep my fingers crossed that a patent does not get invalidated. Both relate to the defensive use of patents. If a company makes all of its patents available to all other defensive parties, such as through the upcoming Defensive Patent License (DPL), then I want it to succeed with lawsuits against companies who don't commit to the same defensive terms because this could over time create a strong incentive for others to join the defensive alliance. Also, if a company gets sued over patent infringement and then countersues, I'd like the aggressor to regret his original action.

In Versata's case, there's no indication that it has a purely defensive patent strategy. That's why I'm torn now when looking at its action against SAP. But I'll watch the case with interest, and even though Versata is not a FOSS company, interoperability is an issue of tremendous importance to FOSS.

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.