[byoung2]

There is much debate over whether software patents encourage or discourage innovation. There are a lot of arguably obvious software patents (e.g. "Method to retrieve backups from the internet"). Since the USPTO only started issuing software patents in the late 1990's, some or most of these patents are predated by plenty of examples of prior art. Imagine if a movie screenwriter were suddenly able to patent the romantic comedy, or the buddy-cop movie and then use these patents to sue anyone who tried to make a similar film? How many movies would not be made just because of the fear of litigation?

[hugh3]

If the best thing you can say is that the benefits of software patents are debatable, then.... well, you haven't yet convinced me that they're bad.

[CamperBob]

It's not very debatable at all, actually. Plot two curves on the same graph, one showing the rate of fundamental innovations in software and computer science in general and the other showing the number of software patents issued by the USPTO. My guess is that those curves will cross sometime in the 1990s, and never approach each other again.

(The trouble is that the politicians, not knowing any better, will go to the USPTO to get both sets of figures. "See? All of these new software patents must mean that innovation has accelerated at a tremendous rate!"))

[hugh3]

Now let's plot the same curve "rate of fundamental innovations in software and computer science" against the popularity of the name "Tyler" for young boys.

Correlation, causation, yada yada yada.

[CamperBob]

Yadda, yadda, indeed. The point is, there is no real correlation to begin with, except for what a naive politician will see when he looks at the number of patent grants.

[teaspoon]

In case you're not aware, a patent's title isn't an indication of what is actually claimed, nor is it a unique identifier. There could be several patents titled "Method to retrieve backups from the internet", each claiming very domain-specific and non-obvious techniques.