EVERHART@arisia.dnet.ge.com (05/30/90)
One of the most serious problems with software patents is that a patent examiner has maybe 30 minutes to process an application, and must rely on the often incomplete declarations of prior art on the application. Patent searches made some sense in areas where hardware development meant significant work and likely prior registration of patentable ideas. In the software world, we see a very large amount of work done in public domain or in other areas not readily searchable, and patents granted on quite old idea. (I've noticed a number of these in the last year.) While in principle recognition of the act of invention is defensible, in practice for software, one is often not rewarding invention, but is rewarding applying for a patent. It also is making the writing and sharing of software more difficult and risky, even for the real inventors of many of the ideas. (Example: a patent over a program whose "idea" was to allow one terminal to watch or control another in a multiuser system. This was granted in 1984, notwithstanding the fact that the idea had been in commercial and PD use since the sixties.) I advise GREAT reluctance to support software patents. I also have my doubts about other patents for similar reasons. In cases of two simultaneous inventions of the same idea, why should one inventor be able legally to deny the other the rights to his idea? This has happened quite often, and in a number of quite valuable cases (like the telephone). Glenn Everhart