The Software IP Report
Software Claims Held Patent Eligible
By Charles Bieneman
Categories: Miscellaneous, Patent Eligibility, Software Patents, The Software IP Report
A Northern District of California court has rejected an argument that “a method of executing an instruction” was not patent eligible subject matter. Nazomi Communications, Inc. v. Samsung Telecommunications, Inc., No. C-10-05545 (N.D. Cal. March 21, 2012). The representative claim, reproduced below, recited a method by which a Java interpreter could more efficiently access byte codes.
Analogizing to Gottschalk v. Benson, 409 U.S. 63 (1972), which held that a method of converting binary-coded decimal (BCD) numbers to pure binary numbers was not patent eligible, the defendant argued that the claims covered “abstract steps of obtaining a piece of data, using said data to determine a next step to be taken, and then potentially modifying said piece of data.” The patent owner countered that “when read in light of the specification, [the claims] are clearly directed toward execution of computer instructions and thus are not abstract.”
The court found that the claims were “clearly more specific than [the defendant’s] generalized description” because “the claims specify that the data obtained must include data from a resolution data field, that the next step to be taken is a resolving step, and that the modification is to indicate that a reference is resolved.” Further, the claims did not apply to compiled languages such as C, but only to interpreted languages such as Java, and recited “specific implementation details.”
Then the court’s reasoning took an interesting turn. If you are a patent prosecutor, have you ever added a “computing device” to a claim to get an examiner to allow it? Have you then wondered whether this language would be enough to make the claim patent eligible in a court proceeding? Then consider this. Although claim 1 did not recite a “computer” as such, the court rejected the defendant’s argument “that simply requiring the use of a general purpose computer does not impose a meaningful limitation on claim scope” because, unlike the BCD-to-binary method of Benson, the defendant “made no showing that the ‘160 Patent claims involve ideas that have no substantial practical application except in connection with computer instructions.”
The implication seems to be that if your method can only be executed in a computer, then it is patent eligible. Another way of stating this might be that if your claim passes the machine-or-transformation test, it is patent eligible. But as I wrote recently regarding the Prometheus case, which the court here distinguished, it has certainly seemed that failing the machine-or-transformation test guaranteed patent-ineligibility, although the converse is certainly not true. Yet this case seems to have been decided on the basis of claims’ limitation to a machine.
As an interesting side note, the court mentioned, but only in passing and without comment, that the patent owner had argued that the case was not ripe for a determination of patent eligibility because claim construction had not yet occurred. Unlike other courts, this court was not reluctant to apply the test of Section 101 early. That said, this case was decided on an early summary judgment motion. One wonders whether the fact that this defendant’s motion was denied means that the defendants cannot nonetheless raise Section 101 questions again later, e.g., after claim construction.
Perhaps Nazomi illustrates nothing so much as that Section 101 jurisprudence remains of, let us say, uncertain consistency. As illustrated at this site’s Patentability page, lower court decisions following Bilski v. Kappos have gone in various directions, and, in the software realm, at least, many more cases like Nazomi will need to be decided before any clear trends emerge.
Claim 1 of U.S. 6,338,160
A method of executing an instruction comprising:
obtaining from an instruction storage location, an instruction that references a data structure, the data structure storing an indication of a reference that may need resolution;
obtaining data from the data structure including data from a resolution data field;
using data from resolution data field as an index to a jump table to determine whether to do a resolving step; and
thereafter, if the data in the data resolution field indicates that the reference was not resolved, resolving the reference and, thereafter, modifying the data in the data structure including modifying the data in the resolution data field to indicate that the reference is resolved, wherein the data in the instruction storage location is not modified.