— Leland Teschler, Editor

The event made headlines, but the issue hit home for me when I helped judge a design contest hosted by Lantronix Inc. Entrants had to devise something that incorporated wireless communications in an innovative way. One eager designer mentioned he was in the process of patenting his entry.

Another judge scoffed. The designer's idea was just a wireless version of a product that had been on the market for a while. My colleague couldn't see how such a thing would pass the test of being nonobvious to someone familiar with the field.

Not so fast. It's not entirely clear what isn't patentable, and recent court decisions don't help matters. One problem is the current federal code defining patentability does not contain the word "physical." The Court of Appeals for the Federal Circuit has used this fact to conclude that Congress wanted nonphysical designs of all types to be patentable. But that runs counter to a decision made by the Supreme Court back in 1981 that said something based on a mathematical principal couldn't be patented, even if there was a trivial physical step involved somewhere.

Legal scholars have pointed out that recent court rulings make the line between true innovation and a trivial physical step increasingly hard to discern. The upshot is that patents have ever-broader bounds.

Cynical observers have noted that the budget of the U.S. Patent and Trademark Office is tied to how many applications it receives, so it has a vested interest in seeing as many applications as it possibly can. The Patent Office, they've suggested, has become a cheerleader for the broadest interpretation possible as to what ought to be patentable.

That definition should include ideas that are outside "technological" fields, insists the Patent Office. But with such a broad definition, almost anything is patentable, including mathematical algorithms.

The situation has gotten out of hand. If you have devised a macro in Excel, for example, there is a chance someone has already patented your technique. Post your spreadsheet on a Web page and you are distributing your infringement. Professional programmers now just assume the code they write violates someone's patents, and they just shrug it off. The alternative is to do a patent search on every line of code written.

If there is no clear definition of what is patentable then anything is patentable, including baseball batting orders and the route you drive to work in the morning. Clearly there's a need for reform in how patents are granted.

There's got to be a better way. One suggestion I've seen it to post all patent applications on the Web and let the public make comments. If nothing else, you'll get a much better feel for prior art. The approach is reminiscent of an old Web site called BountyQuest.comwhich, a few years ago, offered a bounty to anyone who could give evidence of prior art invalidating selected patents. One of its success stories was that of an engineer who remembered an article from an old Byte magazine that laid out the basics of an idea long before the patent had been issued.

BountyQuest.comfizzled out, but figure out a way to implement what it was trying to do on a worldwide scale and you have a system for recognizing intellectual efforts that is both fair and equitable.