Archive of UserLand's first discussion group, started October 5, 1998.

Re: solved before (Patents)

Author:David McCusker
Posted:9/22/2000; 7:45:34 AM
Topic:Patents
Msg #:21631 (In response to 21625)
Prev/Next:21630 / 21632

Josh Allen: And the idea that you could set out to solve a problem without bothering to see if it has been solved before is horrifying.

Ravi Nanavati:I'll also add that finding out whether or not what you're doing has been done before is much harder than it looks, even in the age of the Internet.

It rarely seems feasible to research patents for answers to problems. I have never done so, and I have never met a single person who has said or hinted they have done so. Even Ravi does not say he searched patents for answer -- he implies he was looking for new ground. This is what I hear folks say about patents. One looks for conflicts.

Patents intend to encourage creation by rewarding disclosure. I'm also under the impression that patents are intended to actually share the information with other folks who might use the inventions, and that this is the purpose of disclosure. But when no one actually reads the disclosures, something must be drastically wrong with the process.

Folks don't read patents for many reasons. The dominant reason varies. My main reason for not reading patents is because they tend to be idiotic. They disclose obvious things clothed in absurdly obfuscated language one can only assume helps befuddle poor examiners who don't understand software in sufficient depth.

But the secondary reasons are nearly as important. Patents are both too narrow and too numerous to be useful. Software solutions tend to narrowly address problems in a brittle way. When solving a similar problem, it's often easier to start from scratch, because adapting an old solution is often not productive. Patents often describe solutions with no more insight than first blush reactions to a stated problem. The only novelty is that a patent even addresses a problem.

As mentioned, there are just too many patents. Even with search engines in software, it's not feasible to find ones related to an issue one is considering. Language based searching for words in common is not quite successful enough to be that useful. And as Ravi said, once one is working on a problem, one cannot continuously research side techniques that ought to be trivial or inconsequential.

A patent library would be useful if small enough in scale that a practitioner could expect to grasp the entire contents affecting practice without years of research. A huge patent library acts merely as a property registry, where folks can put up fences to keep other folks out. But since most folks can't waste the time necessary to find the patents, they must risk being unpleasantly surprised.

The main problem with patents is they fail to share useful information with practitioners, so the intended purpose of sharing does not work. If we assumed sharing worked (which it doesn't), then the main problem would be that the give rights for too long a period of time.

There are responses to this message:

This page was archived on 6/13/2001; 4:56:49 PM.

© Copyright 1998-2001 UserLand Software, Inc.