Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Zing]

If I were to draft a software patent app I would do so as Turing machine that performs xyz functions. That way the patent office wouldn't complain as much on 101 grounds.

Let us not forget that rejections are based on the content of the claims.  Simply adding a description of a Turing machine as an embodiment in the specification would not magically prevent a 101 rejection.  To claim software (e.g., a processor and memory comprising program instructions executable by the processor to yada yada yada), you would still need to include the requisite structure of a computer system in the specification, and you would still need to write the claim from that perspective as well.  You could claim a Turing machine, but then your claim is basically worthless as 99.99% of competitors aren't going to use a Turing machine to implement their products.

[JimIvey]

I didn't mean to suggest claiming a Turing machine, but rather to describe one.  I would expect to claim a method and perhaps a machine performing the method (in addition to the Beauregard type claim-set).  The claims would be infringeable by software and Turing machines alike.

The whole point is to illustrate the futility of excluding "software" from patentable subject matter as the designation of "software" is largely arbitrary and meaningless since software and the functionally equivalent fixed digital hardware logic are technically interchangeable.

Regards.

[dablueman]

You could claim a Turing machine, but then your claim is basically worthless as 99.99% of competitors aren't going to use a Turing machine to implement their products.

Jim may not have been talking about claiming an actual turing machine, but I was and I refer you to the 1L concept of doctrine of equivalents. It was just a thought that came to me when I went through my undergrad C.S. books, but I believe it to be a valid argument that if docketed by the S.Court could override in re bilski and maybe elevate C.S. in the eyes of the PTO. Right now C.S. is treated as a second class citizen.

[Isaac]

IThe whole point is to illustrate the futility of excluding "software" from patentable subject matter as the designation of "software" is largely arbitrary and meaningless since software and the functionally equivalent fixed digital hardware logic are technically interchangeable.

If you describe a process as operating on data using circuitry, but draft claims that don't recite any hardware, the underlying method will be considered an algorithm, and you're going to get a Bilksi type 101 rejection.  It might be that claims that won't pass transformation/machine test are more prevalent in "software patents" particularly after State Street encouraged people to write such claims, but I don't believe the rules discriminate against "software patents"  Many practitioners were in the habit of scrubbing recitations of hardware out of method claims, and in many cases, those claims are not going to be in an acceptable form anymore.

I'd be very surprised if the Supreme Court did anything helpful in Bilski.  I expect that the Solictor General will recommend that cert be denied if asked to opine on the issue.   Further there appear to be at least 3 Justices that feel that State Street was wrongly decided by the federal circuit.

[Zing]

Claiming a Turing machine seems imprudent.  Relying on such a claim effectively ensures you cannot prove literal infringement, which leaves you with the DOE.  Relying on the DOE to prove infringement is murky territory at best: (1) you will have the burden of proving that estoppel does not negate your DOE claim and (2) determining whether DOE applies will most likely fall into the hands of a jury (do you trust them?).  On top of all that, don't forget about dedication to the public, see e.g., Johnson and Johnson.

[JimIvey]

If you describe a process as operating on data using circuitry, but draft claims that don't recite any hardware, the underlying method will be considered an algorithm, and you're going to get a Bilksi type 101 rejection. 

I'll respectfully disagree with that.  Method steps (at least the ones I write) tend to include steps that implicitly require a machine of some sort -- steps like receiving signals generated by a user, storing data, retrieving data, sending/receiving data, etc.  I believe implicitly requiring a machine is no less sufficient than explicitly reciting machine hardware, although I suspect the latter would be helpful so long as Bilski is good law.

Note that Bilski tossed out most of the algorithmic bases for rejection under Section 101.  A machine that does no more than solve a mathematical equation is now statutory subject matter under Bilski.

FWIW, I believe that result is correct -- a machine is a machine is a machine... and is statutory.  Period.  I believe prior caselaw to the contrary was also contrary to Section 101 and quite clearly so.

Of course, why the same isn't true for a process is beyond me.  Section 101 could easily have listed machines, compositions of matter, and processes that involve machines or compositions of matter, but it doesn't.  It just says "process" without qualification.  Seems pretty clear on its face, if you ask me.

Regards.

[ababab]

The Supreme Court has granted certiorari in the Bilski case.

Thoughts?

Predictions?

Prosecution strategies for now?

[JimIvey]

I don't expect much of Bilski to survive the Supreme Court.  I think the whole of the rationale upon which the Bilski opinion relies is just bad interpretation of prevailing Supreme Court precedent.  Here's what I said about that before:

At the foundation of the opinion is that the machine-or-transformation test is the test for 101, relying on (i) an interpretation of "clue" as meaning "dispositive test" and (ii) failure of the Supreme Court to repeat an earlier clarification when repeating the "clue" of machine-or-transformation.

In particular, once the Supreme Court added that "We do not hold that no process patent could ever qualify if it did not meet the requirements of our prior precedents."  In a subsequent opinion reaffirming the use of a machine-or-transformation test for 101 purposes, the Supreme Court did not repeat that caveat. 

That's it?!  That's all the Supreme Court has to do to reverse itself?!

Let me ask you this:  I loan you my car and warn you not to scratch it or harm it in any way.  You comply.  Next time, I loan you my car and don't warn you.  Are you now free to scratch and bang it up to your heart's content?  The Federal Circuit appears to think you are.

I'm trying to figure out when the case might be heard without much luck.  It's interesting that Sonia Sotomayor has some IP experience, although I admit I don't know much about that yet.  From my quick, sketchy research, it looks like the case will be decided before she's on the Court.

Bilski hasn't affected my practice much, except for perhaps trying to make links between the processes and machines and/or material transformations more clear and explicit. 

Overturning Bilski wouldn't change my practice much either.  Who knows what the law will be 20 years from now?  If you can prepare for contingencies such as Bilski, you just do it.

Regards.

[Isaac]

I'm trying to figure out when the case might be heard without much luck.  It's interesting that Sonia Sotomayor has some IP experience, although I admit I don't know much about that yet.  From my quick, sketchy research, it looks like the case will be decided before she's on the Court.

The plan is for the new Justice to be seated by October. 

My own sketchy research.... For KSR, cert was granted in late June 2006 and the case was heard in November 2006.  Perhaps that means there is some chance the case she'll get to hear Bilski.