Peer Review Starts for Software Patents

perbert writes "As seen in an interview in IEEE Spectrum: Qualcomm v. Broadcom. Amazon v. IBM. Apple v. seemingly everyone. The number of high-profile patent lawsuits in this country has reached a staggering level. Hoping to curtail the orgy of tech-industry litigation, the U.S. Patent and Trademark Office (USPTO) is experimenting with reforming the way patents are applied for and processed. Launched on 18 June 2007 was an Internet-based peer-review program whereby anyone (even you) can help to evaluate a number of software patent applications voluntarily submitted for public evaluation. The one-year pilot Peer-to-Patent program is a collaboration between the USPTO and New York Law School's Institute for Information Law and Policy, in New York City. The program's Web site allows users to weigh in on patent applications by researching, evaluating, submitting, and discussing prior art, which is any existing information, such as articles in technology journals and other patents, relevant to the applicant's claims."

Comments

Looks pretty solid

[Evets]

Nice...

5 applications online, 22 instances of prior art submitted.

While everyone's favorite seems to be Method, apparatus and computer program product for providing status of a process, I had to laugh at Database staging area read-through or forced flush with dirty notification - it's pretty much a description of how every memcached/SQL plugin works. I guess somebody should mention that.

Apparently, if you make the description sound complex enough it will pass initial review.

It's good to see this kind of a process come to light. Three cheers for Beth Noveck.

Re:Looks pretty solid

[Anonymous Coward]

Apparently, if you make the description sound complex enough it will pass initial review.

Yes. I once had a conversation with a lawyer who worked in a patent office about this, and this is very correct. In theory the people reviewing patents and determining whether or not a patent should be granted are technically educated and competent. In fact, many of them are technically literate, but as we all know the world of computer technology is huge and what is often required to intelligently review a patent is not just technical literacy, but plenty of domain knowledge in the specific area of the patent (many people who use computers every day, even people who program computers, know very little about how memcached/SQL plugins work).

So even though some effort at technical literacy is made, the fact is that there simply isn't enough manpower to pull it off. A patent office only has so many people on staff, and so much money that they can spend on expert consultation, and the submitters are always in a hurry. These real-world pressures leave specific clerks holding the bag; they are more-or-less forced to make a decision with too little information, and financially incensed to lean towards approval.

So, yes, techy-sounding complexity will go a long way towards getting bogus patents approved.

Re:Looks pretty solid

[billsoxs]

I think that it has more to do with the time that the Patent officer has to spend on the patent. I am involved in such an issue and the offending patent has a "Korean" patent referenced on the front. That patent was from "1977". Guess what, the new patent was just a rehash of the "77" patent. (There is prior art to ~"1970".) How can I tell, I've read a translation old patent and the new patent. A laywer has said that because it is listed IN the patent - it is unlikely that a court would over turn the new patent.... The court will assume that the USPTO actually looked at the "Korean" document and thinks that it was different. That same laywer is seeking to find out if a translation of the "Korean" patent was looked at by the USPTO. He thinks it is unlikely. This is because of the time it would take to do so.

Note dates and names have been changed to protect the guilty.

Re:Looks pretty solid

[eggnoglatte]

On first sight, the situation may seem positive, but unfortunatly the quality of the "prior art" submissions is dysmal.

For example, if you read even just the intro of "Method, apparatus and computer program product for providing status of a process", it is immediately clear that the authors use the term "process" in a non-software meaning (i.e. a production process or something similar - they use a power plant as an example). Yet two of the supposed prior art submissions refer to operating system process monitors (one wanker actually uploaded a .exe file). Most of the other "prior art" submission have already been rebuked by other participants.

I predict a very fast end to this pilot test unless the quality improves drastically. If you are gonna comment, you should at least bother reading the patent, for christ sake.

No, but it probably aids peer review

[EmbeddedJanitor]

Attaching a bunchof prior art (or supposed prior art) to an application sure helps the peer review process.

Only problem I see with this is that it can easily become deep pockets vs the rest exercise. No doubt MS and some others will have a few people perminantly assigned to tearing apart applications from others.

Re:No, but it probably aids peer review

[LiquidCoooled]

There is hope then.
If one company can do it, then so can others.
At some point they will realise its a waste of all their time, or we will end up with a server room fist fight broadcast on youtube.

win-win I say.