It is funny that the Unix team at Bell Labs ended up realizing the setuid was not such a good idea after all, and replaced it with a much better design in Plan 9 (meanwhile this has been pretty much ignored back in *nix land, where many limitations, like the lack of usable private namespaces, and security issues are due to setuid):

http://doc.cat-v.org/plan_9/4th_edition/papers/auth

Somewhere there is a comment by Rob Pike (which I can't find now) about how it was ironic that the only thing patented in Unix early on turned out to be such a bad idea after all.

This historical notes aside, that we allow this software patent insanity to go on is scary and depressing. The billions of dollars being wasted (not to mention the amount of time and other precious resources) because of patents is staggering.

  I think it was 1992 or so before it was really clear
  that patents were the true and final nemesis of my
  chosen profession.

The real issue is that patents favor large players (even if they are 'branching out' and aren't an incumbent in a particular area) to the detriment of smaller players.

I think the REAL issue, is that they lock up IDEAS. Ideas by themselves are inherently worthless. It's the EXECUTION that matters. Everyone has ideas. Some are great, some suck. The idea by itself is really meaningless until someone does something with it. Ideas, to me, are like knowledge. They should never be locked up, and should be freely shared.

Patents aren't supposed to lock up abstract ideas, they are supposed to unlock concrete implementations of ideas (unlock by giving details of the implementation to the public in exchange for a short-term monopoly).

There are good arguments either way for whether this is a good idea, but make the ideas far more abstract (like software) and add incredibly short technology cycles (like software), and there is no doubt it's a bad idea.

Patents have never actually worked that way though. For instance, take a look at the history of the early US automobile industry sometime.

That is an excellent point! The way they are used vs. the way they are MEANT to be used, seem to be worlds apart.

Patent trolling as a business, which IV has proven is viable, relies on a precondition: patent litigation must remain expensive. In specific, it must be prohibitively expensive to determine whether each and every asserted claim in each and every patent asserted against an "infringer" is valid. This determination can only be done via patent litigation.

What happens if this precondition does not exist?

The patent trolling business will fail.

This is because everyone knows most, maybe even all, of the claims in the patent portfolios the patent troll has amassed are not valid. They are potentially worthless. Companies simply do not know which ones are valid and which ones are not. And it's too expensive to find out. So companies are willing to negotiate instead of engaging in patent litigation.

If we were to make patent litigation so inexpensive that anyone could afford to "call the patent troll's bluff", these ridiculous patents would never be filed for much less asserted. In other words, if we could have an inexpensive determination of what claims are valid and which ones are not valid, we could separate the wheat from the chaff. And these enormous patent portfolios would shrink down to size. Just the wheat. If there is any. It would be difficult to make hollow threats and engage in IV-style extortion.

It is only the expense of finding out whether a claim is valid (through litigation) that makes patent trolling a viable business.

Maybe patent lawyers managed to get the best of Myrvold while he was a CTO, and he believes by opening the gates for a vibrant patent trolling industry, he will make life easier for every technology company in the future. All they have to do is pay some protection money and their worries will be gone.

"the belief that just because people could be evil, doesn't mean they will be evil"

You might attribute the problem to immaturity. Behold the attitude in this thread: finger-pointing.

A patent is not a right or an obligation to produce anything.

In truth, it is a right to sue others.

But surely, that is not how most patent applicants think about patents when they apply for them. We expect they are have intentions to produce something. We expect they will have products or services to sell or license.

People on the web discussing patents are apt to mention government-granted monopolies and pull out quotes from the US Constitution about promoting the progress of useful science and the arts.

Clearly, these people are not just thinking about patents as rights to sue: "If I obtain a lot of patents I can start threatening to sue other companies." They are thinking about companies that are planning to produce something and bring it to the marketplace, and the competition those companies might face.

But the IT industry is showing us that indeed there are people who are thinking of patents as rights to sue. Thye even think this somehow constitutes a legtimate "business". Because that is the ONLY way they are using patents. "Patent troll" is a term coined by someone in the IT industry. That's where it was born. It's proliferation as a "business model" is being led by a former Microsoft employee.

As is true of technology, a patent portfolio is not inherently "evil". It depends on how it is used. The IT industry is showing us how to use patents in the very worst way.

I call it immaturity. Steve Jobs throwing a hissy fit and calling Eric Schmidt from Burning Man about Android.

The problem is not patents. It is how the IT industry is using them.