Let's add business methods and existing genetic sequences (thankfully that one has been partially corrected) to the list.
At the very least, one should have to demonstrate economic damages arising from the violation.
If I'm a glass company that spent time and money developing a revolutionary new compound to strengthen glass and my sales are being impacted by someone who copied the compound, then it seems reasonable that I should have some protection.
But if I developed it and then sat on it, I should have zero ability to claim anything.
Patent trolls and lawyers.
The lawyers. Hence...
And the crybabies that think that all they do is a unique invention and that it costs millions of dollars to write a couple of lines of code
I think that is a great point. I believe that many people think that you should get a patent for something if you thought of something first. It certainly has an appeal towards the common man if you can get rewarded for just an idea. As IT professionals we all know someone who has an idea and just needs someone to execute.
Companies that have spent billions on defensive patent portfolios, like Google.
Though, ultimately they might still come out ahead, depending on their ongoing patent litigation costs.
Why limit it to software?
Different fields of human invention have different economic characteristics. For a field where patents may be more economically useful than they are in software, consider the pharmaceutical industry. Developing a new drug costs a decade and hundreds of millions, and remains useful for a long time, but once developed is easily copied. None of those conditions are true for software (and there is no equivalent of copyright protection for drugs).
For more, try http://www.becker-posner-blog.com/2012/09/do-patent-and-copy...
If you think that it makes sense to give the first person to come up with an idea exclusive rights over it for 20 years, compare the amount of innovation supplied to the world in having the idea of putting something interactive on the Web to the amount of innovation that has been built on interactivity on the Web (and yes, the patent is very close to that broad). The former is hardly trivial - it was a novel idea and would have been a very cool demo at the time - but the latter is half of the technology industry. Consider the tens of millions of dollars Eolas was able to extract from random companies while doing nothing, and how much more they might have gained if they had happened to have been a few months earlier. Consider the stupid hack required for years to work around the patent in IE - it could have been worse, since this patent was broad enough that nobody was going to stop innovating because of it yet not broad enough to lack a simple workaround, but it was still stupid.
I have a hard time seeing who would be harmed if software patents were categorically invalid.