Slashdot Mirror
New Prior Art Cited In 2nd Eolas Patent Rejection
theodp writes "To be able to reject the Eolas browser plug-in patent a second time, the USPTO had to add the teachings of G.Toye after Eolas' response prompted the examiner to withdraw his previous finding that was based solely on the teachings of the W3C's Dave Raggett and Tim Berners-Lee. It's unclear where the Toye prior art came from, since the W3C didn't offer it when it asked the PTO to overturn the patent. Also, a newly available document reveals that the W3C's widely-publicized prior art filing, which was hastily made without community input, differed little from an unpublicized filing that was made weeks earlier by attorneys from Microsoft and AOL."
The patent specifically mentiones hypertext, so its not really a patent on "plugins" but instead on the EMBED or OBJECT tags.
Ray Ozzie did demonstrate something using OLE and Lotus Notes, but I dont think it is being used as prior art in the lawsuit.
You might change your opinion when The Evil Empire will be the one using patents against Linux.
It would be ever worse of algorithms were copyrightable!
No, it wouldn't.
Patents at least expire in 20 years, but copyright is forever (less one day).
Cripes, people, learn the difference between the two. Copyrights and patents aren't the same thing with different terms of expiration. They're related, but conceptually different. Copyright is what prevents me from cutting and pasting the contents of a romance novel (or copying the executable file of MS Word) and selling it as my own work. I am free to write my own romance novel (or word processor) because I'm not copying someone else's work. It's called "copyright" for a reason, i.e. the right to make copies.
Patents, on the other hand, are short term monopolies on methods and processes. You would not (to construct a bizarre example) be able to patent MS Word; rather, you'd patent the concept of a word processor itself. The problem with algoriths and source code is that it straddles the line between machines and written work: it is the code that makes the machine perform the process. Because of this, the USPTO has been instructed to treat an algorithm as a machine that performs a process. This is a bad move, in my opinion. Code has more in common with mathematical formulas (non patentable) than machines (patentable). I think what the original poster meant was that specific algorithm code should be copyright protected, and the algorithms themselves should be unpatentable.
If a job's not worth doing, it's not worth doing right.