"And was he [RMS] not right about the Java Trap? How many times must he be right before developers listen?"

He who has ears...

The real question is: how broken are they, and what level (and form) of reform/repeal is necessary?

The standard case for patents is to compare them to business secrets. In a world without patents, innovators may choose to keep their methods secret, and not share them with the world, in order to protect them. A patent gives the inventor a legally enforceable monopoly for a limited period of time, in exchange for publication of the details of the method such that others can reproduce it when the term expires.

The key practical difference in terms of software patents is that reverse engineering makes it relatively impossible to actually keep an innovation secret-- but, as I said, that's a practical issue, and doesn't effect the philosophical justification of the patent system.

In other words, young grad students Sergey and Larry are able to slave away on their super-secret algorithm, secure in the knowledge that no better-funded evil empire can steal it and implement it and compete with them.

Uh-huh. Just like groklaw always is, right? You keep telling yourself that.

The blindest man in the world is the one who thinks he can see.

Admittedly the SCO reporting may have sounded one-sided, but in that case the facts were legitimately also rather one-sided.

But were the facts as one-sided as groklaw suggested? The judge's opinions were nowhere near as one-sided as groklaw's opinions; given a choice between believing that groklaw is biased and believing that a US district court judge is biased, I know where I stand.

As far as I can tell, they were. SCO made a lot of strong claims, and not one panned out.

I personally try to go out of my way to see the other point of view. And usually I can see things the other side says that leave me thinking, "OK, I can understand their point of view." I might not agree with it, but I can understand where they are coming from. But somehow I never saw Darl McBride say anything with that result.

For instance take a look at http://www.sco.com/copyright/ for SCO's position that the GPL is unconstitutional. I've read that many times, and every time I come away wondering what planet the author was on. The laws on copyright are very clear, have been ruled constitutional, and creates a notion of property that is the foundation of the GPL. If you don't like the license, don't use it. If you don't want to encumber your software with that license, don't use it. If other people wish to use that license, that is their prerogative.

Now possibly I'm missing something. Because a similar argument was advanced in Jacobsen v Katzer that you can't sue for a copyright interest which you weren't commercializing. So obviously a trained lawyer thought that claim was worth advancing. Yet the judge came down heavily against Katzer on that one, which suggests that I'm not missing too much there.

The judge's opinions were nowhere near as one-sided as groklaw's opinions; given a choice between believing that groklaw is biased and believing that a US district court judge is biased, I know where I stand.

Your theory is that the judge's opinions were not one-sided because he thought SCO had a good case. My theory is that he bent over backwards to be fair because he expected it to wind up being appealed, and he wanted there to be as few grounds for appeal as possible.

If you look at the decisions, the judge gave SCO every procedural technicality, and every opportunity to present their case in full. Yet in the end he came down against them on every single fact of consequence. And the case went to appeal, and most of his decision stood. Which fits perfectly with my understanding.

A patent issue on an app I write can be easily resolved (pay the fine, royalties, etc.) and re-write that code. But when the underlying bits that I've built on are in question, I'm screwed.

I think that sums it up for Google. They picked the wrong platform for the right reason (lot's of Java devs) and they hoped Sun would continue to be nice about it. Sun is gone now and all bets are off.

As developers, if we want our apps to have futures or to at least have some control over our own future, then we must use entirely free foundational tools (ISO standard programming languages with public domain/GPL/BSD/ licensed compilers and libraries) or be willing to face the consequences.

Nothing compares to the rug being jerked out from under you. Except for when they jerk the foundation that the rug sits on out from under you. Google will learn this lesson the hard way. I hope they have a serious plan B.

The blog is written for and from within the FOSS community; what kind of disinterested lawyer do you imagine blogging about legal matters related to free software, and what kind of programmer would want to read a bunch of dry legal analysis?

I have no idea what this is supposed to mean.

The best thing that happened to Linux was the BSD fight.

It's not exactly clear that we're overall better off than we would be if everyone used BSD instead.

The best thing that happened to the web was the Microsoft antitrust fight.

Because it undid part of the damage that Microsoft had spent the last 15 years causing.

Whether or not these are good for everyone or bad for everyone is hard to judge. But as stated, software patents are a spur to innovation (Linux isn't necessarily better than BSD, but it is its own innovative phenomenon -- the same with the web) by being an antispur (causing stagnation) to greed-touching systems (Oracle, Google, Microsoft, etc., all have parts of their business that are motivated by capitalistic desires or what Gordon Gecko, etc., called greed).

I probably shouldn't have said these were the 'best' things to happen to those respective technologies. But who knows how important they are? Certainly it's easy to overlook them -- as the larger frame of reference in which technology develops is often overlooked.

So that was my point in responding to the last sentence in the article proper:

If you still think software patents are a spur to innovation, you're not paying attention.

But, I understand the signal/noise ratio is a big concern, so somewhat argumentative claims are usually discounted. But whatever -- I think it's important not to forget the larger contexts if we can...

Your overall thesis relies on innovative work being immune from patent holder attacks. I think that might be true in a system that functioned correctly, but not in ours. Patents work if you're forced to claim a specific high-effort solution rather than the entire problem space.

This is a beautiful, concise encapsulation of the concerns that some have with how software patents have worked in practice.

How I read it, fwiw, is that kernel development involves a lot of drivers, and benefits heavily from a leaf-heavy committer structure. Which is what Linux is built on (even if it's perhaps less elegant at the top, with one or two people arbitrating everything instead of the FreeBSD teams, etc.). But I imagine BSD might've adapted / grown a more driver-amenable tree structure if there weren't legal pressures in their inception.

Perhaps I'm naive, but I don't know how important the GPL was to Linux's success. I think it's a hugely important piece of history -- and a very active one. But the resulting forces from legal outcomes in technology -- they're just so momentuous. Think of other industries. What other industries cycle through major corporations every decade? I mean, that's 4-5 major shifts since personal computing was even thought of. Is it just the rapid speed of technology driving things?

Your overall thesis relies on innovative work being immune from patent holder attacks.

That's a good point. But my point is that the day you sign up for a patent is the day you've tapped someone else to beat you. And you could go even further and say, the day that happens, you've spurred the next great technology.

The day you go anywhere near the law is the day you've crested around exponential growth. The law is too slow. Technology is too fast. But Google would have to detach themselves totally from any form of greed to realize that. I doubt they will, even though I can't think of a single technology firm (in the past 60 years) that has grown exponentially after engaging in a legal fight voluntarily.

That said, the points you make are interesting and perhaps more pertinent than the kind of crazy long-term theorizing that I am rather fond of.

ETA. One more point. BSD, by actively avoiding any type of legal engagement (i.e., allowing proprietary use via their licensing) -- which decision arguably was a direct result of their involvement in the 90s legal battles -- may prove in the long-run to be something that pushes its adoption further than Linux. BSD ironically forms the basis for iPhones and iPads and all that -- but is it still BSD? It's curious though because whatever it is -- it seems like technology swiftly moves around whatever legal engagements are in place. Whether or not this is a good thing, and whether we should protect free software for the long run, I don't know.