Patent Absurdity – A Possible Cure
Over the last decade, a number of absurd patents have been granted for software. Now there may be a cure for that problem less draconian than banning software patents altogether. This post describes the problem, the possible cure, and a live test of the cure with prior art being filed by yours truly in a pending patent application case.
The price paid by legitimate developers in defending against “patent trolls” (companies with no obvious business purpose other than to extort settlements for dubious patent claims) is high even when they prevail in court; even higher when a non-technical jury makes an improper award (happens way too often). Dubious patents are also used by incumbents to stifle disruptive innovation (see almost every telco v. Vonage).
Often the techniques for which the patent is claimed is ridiculously obvious; in other cases it is well-known and in use at the time the patent is filed. The problem is that the patent examiners are often not familiar enough with the field to detect either case. The patents are granted years after they are filed; they are enforced years after that. And, by that time, it is very difficult to identify the prior art which would invalidate the patent.
The US Patent and Trademark Office (USPTO) is doing a test with New York Law School of a public peer review process of some 250 pending patent applications in computer architecture, software, and information security. The public is invited to comment on these pending patents at http://peertopatent.org/. Those seeking patents can also get an expedited review by voluntarily submitting their patents for peer review.
Brad Feld blogged that there were eleven days left (now nine) to comment on an application for a patent on what sounds suspiciously like the ubiquitous progress bar which lengthy applications display. (This patent was filed by Nicholas Kovacs of IBM; IBM is a sponsor of PeerTo Patent.) Here’s the abstract:
“A method for readjusting and interpolating a progress of an execution of multi-step program by a computing device involves a computation of a cumulative point baseline for each step of the multi-step program based on a completion time baseline for each step of the multi-step program, and a regulation of a progress indicator based on the cumulative point baseline and a completion time of each step of the multi-step program as executed by the computing device. The multi-program is an application program or operating program executable in steps by any type of computing device. The progress indicator can have any form, particularly the form of a progress bar.”
Brad challenged those who’ve actually written such code to take action.
Well, back in 1987 I wrote a program called Desktop Express which was published by Dow Jones Software and which was an interface to MCI Mail. At 1200 or 2400 baud downloading mail was a slow process so the program tracked progress to let the users know something was going on. Fortunately, I still have the program and manuals. Would have to go to a museum to find an original Mac to run it on but the manual has a relevant screen shot.
Certainly fits all the elements of the abstract. However, when you contest a patent, you contest specific items called “claims”, not the abstract. You have to demonstrate that the prior art covers each aspect of a claim in order to invalidate that particular claim. Often in a patent review or a patent suit, some more specific claims will stand while general ones will be invalidated. Turns out (in my opinion) that Desktop Express in prior art for five of the twenty claims in this patent application so that’s what I filed.
It looks like the Windows Installer is prior art for at least two more claims relating to use of the “methods” by a program installer and to display as a progress bar.
I didn’t file this as prior art because I can’t vouch for the code in the Installer as I can for Desktop Express; maybe someone else will. You are free to file your own prior art and/or to comment on the quality of the prior art submitted here.
I hope this experiment works but I’m skeptical unless and until the USPTO requires this process for at least all software patents. It’s hard work to match prior art to a claim with the specificity rightly required. People aren’t going to do this, I’m afraid, unless they have a stake in the outcome – ie. don’t want to be attacked under a frivolous patent. The application here has attracted lots of comments, some of which claim prior art or knowledge of the existence of prior art; but, so far no one but me has bothered to actually file the prior art.
If all software patents have to go through this process, than perhaps all developers will track relevant filings and those acting out of healthy self-interest will police the system. I hope so.
I am listed as the inventor on several software patents myself. I have also earned significant consulting income helping to find prior art for patent defenses. And, when I was head of ITXC, we wasted lots of time and money fighting a frivolous claim – we won the case but it cost us plenty. The temptation to settle is always high.