I didn't mean for my original post to come across as a challenge that there are no good patents (software or otherwise). I think the examples he gave are quite good.
I think most of his examples are pretty reasonable. I don't have a serious problem with hardware patents (though some economists argue they don't work either).
Software patents fail much more spectacularly. They seem to have little to no benefits even before you weigh the spectacular failures they create for start-ups and small businesses and their ability to mask collusion between large tech vendors.
