Software, in particular, typically is equally amenable to licensing or sales models. In a word, control is typically the most compelling reason to license your IP rather than selling it. Freedom from ongoing duties to buyers is often a compelling reason to sell it. Video game makers, particularly those who make games for play on the Internet or otherwise among gaming “communities” seem increasingly have settled on the licensing model because they think it enables them to insure a level playing field for all of their customers. Inevitably, however, their customers feel, or at least want to feel that they own the software they purchased.

So what are the principal factors courts use to differentiate a license from a sale? Case law seems to be coalescing around three. Considering how one wants to address each may help individual IP owners decide whether licensing or selling works better for them.

The first seems obvious, what the agreement says. If it says it is a license, it more likely is.

The second is whether the IP owner significantly restricts users’ ability to transfer the software. In a recent Ninth Circuit case concerning the World of Warcraft game, the court held that restrictions prohibiting re-sale, restricting transfer to those who agreed to abide by the end user license agreement and requirements that the transferor delete all of its copies, evidenced a license rather than a sale.

The third is whether the IP owner imposes significant restrictions on the manner in which the software may be used. The court in the World of Warcraft case held that restrictions on the locations in which the software could be used , prohibitions on use with third-party software and reservation of a right to alter the game or suspend users evidenced a license.

Faster Patent Processing Still Won't Be Fast Enough

Interesting article by Jennifer Martinez on Politico last week: Tech investors call for patent reform, followed by an even more interesting response from a tech VC, Gary Lauder of Lauder Partners, the day after. The essence of the Martinez article is that certain VCs suggest that we need a separate patent regime for software because our existing system is unable to keep up with the rapid pace of software development. Mr. Lauder's response recalls my post from last month in which I questioned whether even 20 months would be fast enough.  Politico quotes him as saying: "The patent backlog is the primary problem, not laws providing fair protections to inventors. The patent backlog for software and Internet patents is now at least 40 months in an industry that re-invents itself several times a year. This is caused by the patent office's revenue being diverted away by Congress, so it can't hire sufficient examiners."

Mr. Lauder went on to highlight one obstacle to patent reform.  Politico quotes him as suggesting: "it may make sense to adopt different rules for software patents (e.g. forced Peer-to-Patent), but I would be concerned that the body that comes up with such rules would not adequately represent the ones who need patents most: start-ups."

Bilski Decision Could Have Serious Consequences for Software Patents

The current case of Bilski v. Doll, U.S., No. 08-964, cert. granted 6/1/09, pending before the Supreme Court, may have broad implications for software patents. While not directly dealing with software, the Bilski patent application deals with a process that some are calling too “abstract” to be patentable. The Bilski invention involves a method of managing risk in a commodities market. The steps of the process do not necessarily require software, a computer, or any other physical implementation, but are merely a series of steps leading to the desired result.

How might this impact software? When the Bilski case was heard by the Court of Appeals for the Federal Circuit, a test that is now commonly referred to as the “machine or transformation” test for patentability was resurrected from a previous U.S. Supreme Court patent case, Diamond v. Diehr, 450 U.S. 175 (1981),  See In re Bilski, 545 F.3d 943, 961-62 (Fed. Cir. 2008) (en banc). This test says that, to be patentable, a process must either be tied to a particular machine or must transform an article to a different state or thing. The Bilski process was deemed to be patent ineligible because it was not tied to a particular machine, nor did it transform matter in any way.

Now, the Patent Office and the courts are using Bilski to invalidate software patents that appear to be “tied to a particular machine,” namely, a computer. It has been a common practice to tie software process steps to a computer to render them patentable, especially since the Federal Circuit’s decision in State Street Bank & Trust Co. v. Signature Financial Group, Inc., 149 F.3d 1368 (Fed. Cir. 1998), which involved utilizing software to perform a calculation on a computer. The patentability of Bilski was argued to the Patent Office based on State Street, but the argument was rejected because the State Street process was performed using a machine (the computer), but the Bilski process required no computer.

However, the Patent Office’s Board of Patent Appeals and Interferences has upheld rejections of applications that specifically recite a processor for computer-implemented processes that may not necessarily be considered “business methods.”  For instance, the case of Ex parte Cornea-Hasegan, 89 USPQ2d 1557 (B.P.A.I. Jan. 13, 2009), decided earlier this year, recites steps performed by a processor. Nevertheless, the Board upheld the rejection based on Bilski, stating that “the recitation of a processor in itself . . . does not tie the process steps to a particular machine. In other words, the recitation of a processor does not limit the process steps to any specific machine or apparatus [and] fails to impose any meaningful limits on the claim’s scope. The recitation of a processor performing various functions is nothing more than a general purpose computer that has been programmed in an unspecified manner to implement the functional steps recited in the claims.” Id. at 1560.

This is scary stuff for software developers who have applied for patents and spent thousands on patent preparation and processing fees, not to mention those already having issued software patents who are afraid to litigate them for fear of invalidation. However, there have also been cases where the Patent Office has reversed rejections of similar claims, so there is a definite lack of consistency, at least at the Patent Office.

Where is the Supreme Court likely to lead us? Bilski is a big wake-up call to practitioners who should have realized that State Street never opened the floodgates to patenting processes that apply abstract principles even if a tangible result is obtained through the implementation of the principle. We can expect the Supreme Court to delineate how tangible that result must be for a process to adequately limit a claim that embodies an abstract principle to qualify as patent-eligible subject matter by not pre-empting the principle per se. The key to drafting a claim to a machine-based process that will survive future challenges is keeping in mind that the claim must not pre-empt a basic principle. In the interim there are still plenty of opportunities left on the table by the Federal Circuit to obtain business method patents, provided the applications and claims are carefully crafted to carve out a specific application of a basic principle without pre-empting the entire principle.