Stating the null hypothesis tends to greatly simplify questions like this.
In this case, the null hypothesis is: "Software patents do not allow the development of technology that would otherwise not be developed." (In other words, "they are not good")
It is important to note that null hypothesis's are never proven, only disproven. That means you won't prove that software patents do not help (failure to show otherwise could mean nothing more than you did not look hard enough), however you can trivially falsify the null hypothesis if you find a single case where software patents helped.
Of course in the event that you falsify the null hypothesis you haven't proved "software patents are good", but rather "software patents can be good". (If you want to go with "they are good", then that becomes your null hypothesis and to disprove it we nearly need to find a single case where they are not. If you want to roll with the 'generally' thing, then you should first start by defining exactly what generally means in this context then go from there.)
As for judging data points to determine what exactly they show? That's rather up in the air. Pick a case, state the facts, and argue why you think those facts indicate what you assert they do. This is really more of a soft science.
I'm not sure how valuable the bias from developer anecdotes is. Naturally patents are going to be annoying for the business that covets the technology they cover. If you have patents at all, you will have friction. Bit like how you can't have parking tickets without having people complain about parking tickets. The majority of users and developers the majority of the time have no trouble with patents. At which point you can argue that it's a hidden cost, something that affects all of us without being immediately obvious. And that's precisely my point... if it's a hidden cost, it could equally be a hidden benefit.
I just don't see much worthwhile dialogue going on in the patent debate, just shouting in both directions. For instance in this article, it's not clear to me what people are objecting to. Are people annoyed because someone is trying to enforce a patent? Or is it just because a group of large companies have acquired the patents? Does that imply that we're happy for small companies to have patents, because they'll probably never enforce them anyway, but unhappy for large companies to have them, because they'll actually put them to use? Or is it the nature of the patents (I don't see much analysis of what the patents really cover... they could conceivably be things which took a lot of R&D)? Or, indeed, is it the concept of patents in general? The article did mention stifling innovation, but that's a bit vague - it stifles business activity but that energy could arguably be used on alternative that would be more innovative.
> "I'm not sure how valuable the bias from developer anecdotes is..."
This is a legitimate concern, and something you have to contend with when you're engaging in the softer sciences. Someone with experience in something like sociology may be in a better position than most of the HN crowd to effectively study this topic in a rigorous fashion.
Anyway, (the way I see it) people are objecting, particularly in this article, because they view patents in software as something that can only do one of two things: nothing (read: sit in a companies portfolio never to be read or cared about, only padding the companies patent wallet), or harm the industry by stifling innovation and feeding the lawyers. Put briefly, they, unlike you it seems, do not accept the idea that patents in the software industry can actually do good.
It's not the concept of patents, it's not that large companies are using them, it's not that people are enforcing them, it's not any of that nonsense. It is more that none of these things have anything to do with improving the industry we have invested our lives in. I think you had a grasp on this concept earlier, but lost sight of it when you tried to analyse it too closely.
Patents in the software industry cannot actually do good - why? Because patents are wrong to start with and the conditions of the software industry (fast-moving, high volume of research/invention) magnify the problem? Because the patent office isn't sufficiently equipped to grant reasonable patents on software (judging 'obviousness' etc.)? Because all these HNers have personal experience of being constrained by patent issues? Because coming up with ideas for new software methodologies/techniques is not comparable with the R&D of other industries?
It's not the concept of patents, it's not that large companies are using them, it's not that people are enforcing them, it's not any of that nonsense...I think you had a grasp on this concept earlier, but lost sight of it when you tried to analyse it too closely.
I'm going to be honest with you. You have me at something of a loss for words.
I think I have given you all the tools you need to understand this issue at this point. Remember that thing I wrote about the null hypothesis? Well people who disagree with you do so because they have yet to see that null hypothesis convincingly falsified.
If you want to disagree with that, then knock yourself out. It requires absolutely zero expended effort on your part to disagree; I disagree with uncountable things in this world every day for a grand total of zero calories burned.
If you want to disagree and convince others of your point of view, then you are going to need to put some back into it. Rationally and methodically state your case for the falsification of the null hypothesis. Write it up on your blog and post it to HN. If the community considers it worthy of discussion, they will address it.
You seem to be attempting some sort of 'know thy enemy' nonsense. Stop. You are clearly unable to or unwilling to wrap your head around concepts you disagree with, so stop wasting your time. Understanding why people think what they do is unimportant, just state your case. I've already told you how.
And do make an effort to post it as a new discussion when you do so, instead of attempting to hijack another discussion looking for a flamewar.
tl;dr: Our case: the null hypothesis has not been falsified. Your case: ???
I eagerly await your blog posting; I think we are done discussing this in this other person's article's comments.
Contradicting the null hypothesis is easy: any time a potential investor has asked a small company with a bright idea 'but what's to stop someone else just copying your idea after i've paid you to develop it?' and the developer replies 'well, we have patent protection' to reassure him/her, patents have been useful and supportive of innovation.
The problem is that the argument will immediately move on to the net benefit/detriment to the industry, at which point we need to establish the principles on which people object to software patents.
You are clearly unable to or unwilling to wrap your head around concepts you disagree with, so stop wasting your time.
