|
|
Author |
Topic: algorithms and Ex Parte Lundgren (Read 11757 times) |
|
smgsmc
Full Member
Posts: 269
|
 |
Re: algorithms and Ex Parte Lundgren
« Reply #5 on: Jun 30th, 2007, 9:23am » |
 Quote  Modify
|
on Aug 4th, 2006, 2:32pm, JimIvey wrote:Lundgren states, citing another case as evidence of a pre-existing rule (i.e., didn't make new law here), that a process producing a concrete, tangible, and useful result comports with Section 101. Novelty and non-obvious are additional hurdles beyond Section 101 -- Sections 102 and 103, respectively.
In short, the word you use ("patentable") is a loaded term and must be qualified.
Another loaded term, "algorithm", should also be qualified. Algorithms are patentable (if novel and non-obvious) in a tangible context. For example, suppose I just invented the Pythagorean Theorem and that it's novel and non-obvious. I couldn't patent it by itself, but I could patent technical application of the theorem. For example, the ancient Egyptians used the theorem to make "squares" -- to check a corner for 90-degrees. They'd make a triangle with lengths of 3, 4, and 5 of whatever length units. Of course, that's a right triangle and they'd used the square corner as a 90-degree angle reference. I could patent that (assuming the ancient Egyptians didn't already do that and that it was still novel and non-obvious).
As for "in business", using something "in business" neither paves the way for getting a patent nor excludes patent protection. In short, the fact that something is used "in business" is irrelevant to it's being patentable.
Regards. |
|
I'm getting heavily involved with "algorithms" and related material. I have a generic question. Let's put aside the issue of patentability. From a $$$ perspective (which is what patents are all about), does it make sense to shell out $$$ to get a patent. I mean, are you able to sue someone for infringement? In most instances, I can dissect a piece of hardware to determine whether there are any infringing components. If I have a signal processing method like MP3, I can hookup a signal analyzer. But you normally can’t reverse engineer an application level piece of software (which is what’s publically visible) to get the underlying code (which has the algorithm). So if I’m a landscaper, and have a nifty piece of software to compute how many curbstones I should buy when laying down a border in the shape of a right triangle, how will anyone prove that I’m infringing on a Pythagorean patent? Or as a more realistic example, someone develops an RF engineering simulation tool for laying down cell sites. How is anyone going to prove that my code uses the same numerical integration algorithm?
|
|
 IP Logged |
|
|
|
JimIvey
Moderator
Senior Member
 
Posts: 2584
|
|
Re: algorithms and Ex Parte Lundgren
« Reply #6 on: Jun 30th, 2007, 9:38am » |
Quote Modify
|
First, money and value are business issues, not legal issues. I can generally tell you how much, but whether it's worthwhile is a business decision each businessperson must make for herself.
Second, you raise a point that's a good one and one we haven't touched on here for a while -- what do you do about inventions for which infringement is difficult to detect? Back in the day, you could sue someone you suspected of infringement and then do discovery (exchange documents/evidence in preparation of trial -- sometimes under a secrecy order to protect trade secrets) and determine infringement that way. That door was shut about 10 years ago and I haven't revisited the issue since then, so it may be different now.
Back then, a patent holder claimed infringement of a patent that was the subject of a declaratory action (someone sued to have the patent declared invalid, believing an infringement suit was imminent). It used to be quite natural that, if someone sued to have your patent declared invalid (rather than not infringed), you could infer that they probably infringed and sue on that inference. In that particular court case, the court said you must have some overt reason to believe they infringe (not merely the inference).
So, what's the result? It's possible to have a patent that's infringed that you can't sue on because the infringement is undetectable.
So, what's the solution? First, just about anything is reverse engineerable -- people decompile code, shave chip layers and view under a microscope, yadda yadda yadda. Sure, it's expensive, but it's possible. With enough money at stake, just about anything is possible.
Second, any patent practitioner aware of this case (or its effect) should be careful to draft at least some claims that should be detectable. In my technical areas, you focus on user interfaces ("front of screen" technology), observable network traffic, APIs, etc.
So, yes, it's a problem, but the risk can be managed.
Regards.
|
|
IP Logged |
--
James D. Ivey
Law Offices of James D. Ivey
http://www.iveylaw.com
|
|
|
|
|
Forums
Help
Search
Members
Calendar
Login
Register
Reply
Send Topic
Print