I think the most common argument I hear against IP protections for software is that so much of it is obvious, or incremental. We have companies patenting and suing over techniques which are considered common knowledge by many programmers. How silly was it that Oracle's recent huge lawsuit ended up hinging on a range check function?<p>But is that the same as a masterful magic trick? And if not, should we treat it the same?<p>I don't think it's the same thing at all. A master magician might easily spend years practicing a trick -- the right mechanics, the right theatricality, anticipating the audience response, coming up with some new illusion -- and then, as the article describes, if that trick is self-contained enough, someone else might see it and work out how it's done and then sell the same trick to people who haven't invested anywhere near as much effort in practicing it.<p>So I don't think that we can say, well, this is just like software, it's an idea, and therefore we shouldn't protect it. I think that the very idea behind IP laws in the first place was to protect <i>effort</i> -- to safeguard someone's livelihood if they spent years developing an idea, so that someone else wouldn't simply come along immediately afterward and usurp the idea and rob that person of all of the effort that they invested in developing it.<p>There's this huge difference in effort between writing simple functions in software and developing masterful magic tricks, and we should recognize that.<p>Nor do I think that Penn & Teller are at all guilty of hypocrisy. I think it's enough to have just seen some episodes of <i>Fool Us</i> to see that they can have huge amounts of respect for magicians who copy their tricks, <i>so long as those magicians add something unique to it</i>. That's all they ask: take what we do, and do it better.<p>I can not imagine defending the reverse engineering and subsequent sale of someone else's unique illusion under the umbrella that "software patents are bad".<p>I am somewhat more dismayed that the article seemed to present a good enough case for this all on its own, but the discussion so far is ignoring it -- almost as if everyone just skimmed it.