Kuhn: I Think I Just Got Patented [LWN.net]

Kuhn: I Think I Just Got Patented

Posted Feb 3, 2010 18:49 UTC (Wed) by tzafrir (subscriber, #11501)
Parent article: Kuhn: I Think I Just Got Patented

I'm not sure I get this.

A casual reading of their patent shows that (at least the way I understand it) is about automating the process.

For instance, teaching a human to drive[1] a car is not really novel. Teaching a computer to drive a car would be quite novel.

It's generally simpler to replicate the software Black-Duck produces than replicating the experience of Bradly Kuhn in identifying license violations.

[1] drive a car safely within the streets of a city or on a highway.

to post comments

Kuhn: I Think I Just Got Patented

Posted Feb 3, 2010 20:48 UTC (Wed) by JoeBuck (subscriber, #2330) [Link] (5 responses)

What matters is the claims, and claim 1 would appear to cover any method of automatically finding violations, even when the software packagers have done all the work for you. For example, if you're on a Debian-based system, and you write a quick and dirty Perl script to scan depends fields and license fields of packages to check for violations, you infringe. It doesn't matter if you catch only some cases (e.g. some libraries say GPL2 and some say GPL3).

If someone thinks I'm reading it wrong, please explain why you think so.

Kuhn: I Think I Just Got Patented

Posted Feb 4, 2010 1:49 UTC (Thu) by efexis (guest, #26355) [Link] (3 responses)

No I don't think that's how patents work... patents don't cover an end result, they cover the method to acheive that result. The 'claims' section details what the result is; the method of achieving that result is detailed within the rest of the document. So, if you can achieve the same result, but by using a method different from that which is detailed in the main document, then the patent doesn't apply to you, and unless there is prior art on the method that you've used, you're still free to patent the different method that you came up with.

If this wasn't the case, the remainder of the document would be somewhat redundant.

Kuhn: I Think I Just Got Patented

Posted Feb 4, 2010 4:19 UTC (Thu) by JoeBuck (subscriber, #2330) [Link] (2 responses)

You are incorrect. What matters is the claims. The test for infringement is to compare against the claims.

It is common for claims to be much more general than the actual invention described in the main text of the patent.

Kuhn: I Think I Just Got Patented

Posted Feb 4, 2010 12:09 UTC (Thu) by efexis (guest, #26355) [Link] (1 responses)

A quick lookup says that yep, that is indeed the way patents work in the US (plus also UK 'n Japan), whereas continental Europe is more like how I said, but not identical. That's seems really really dumb huh!

Kuhn: I Think I Just Got Patented

Posted Feb 4, 2010 18:58 UTC (Thu) by hozelda (guest, #19341) [Link]

I was in shock when I realized a couple of years back how the US patent system treats patent claims. The "protection" is over these very broad descriptions.

Those writing patents want broad scope so they push the limits of novel and nonobviousness rather than limit the claims to something noticeably more narrow and specific (comparable to the essentially redundant more detailed description of the actual invention as provided in the patent application).

This is one reason why any fair analysis of almost all existing patent claims (not just software) would conclude that a 20 year monopoly on such broad claims would not promote the progress of science and useful arts (as required by the US Constitution).

Kuhn: I Think I Just Got Patented

Posted Feb 4, 2010 18:59 UTC (Thu) by GreyGeek (guest, #41838) [Link]

"For example, if you're on a Debian-based system, and you write a quick and dirty Perl script to scan depends fields and license fields of packages to check for violations, you infringe."

Assuming that is true, just how does BlackDuck believe they can monitor this kind of situation and enforce license revenues or accrue violation penalties? CSS encryption, anyone?

The Internet is littered with scripts, code snippets and descriptions of such that do this and which have been legally available on line for years. Such code has probably been in textbooks and in online examples of search and sort techniques. People can email these scripts to each other, or print them out and snail mail them to anyone.

Also, other commercial products that monitor license compliance have been around for years, as well. My last employer, before I retired, was using an application that did that in 1997. About the only thing BlackDuck can do is prevent competitors from offering a similar software package, or stop those that already do.

This is just another sordid example of why the USTPO needs an overhaul and why Bilski is needed. But, judging by the behavior of corporate America in the past, you can rest assured that they are spending millions, if not billions, bribing (a.k.a campaign contributions) Congress to pass legislation which would nullify Bilsky regardless of what SCOTUS does.