Patents != innovation [LWN.net]

Patents != innovation

Posted Oct 15, 2007 19:44 UTC (Mon) by dwheeler (guest, #1216)
In reply to: respectful naming of your opponent by sepreece
Parent article: A visit from the trolls

That's the theory, but not the practice.

In practice, all a patent troll does is purchase a patent from someone ELSE who didn't invent anything either. In practically all cases, the actual INNOVATION occurred many years before the patent application, and/or was obvious to anyone in the art (and thus should never have been granted a patent). Both the original person applying for the patent, as well as the person who bought the patent, have nothing to do with innovation.

All algorithms and software were originally barred from patent protection. We have an existing legal system for protecting software (copyright), and there is no evidence that there was a lack of innovation requiring the addition of software patents to the set of legal patents. Sure, the idea behind patents is that we have to pay people to innovate or they wouldn't innovate, but the whole thing is upside-down. But it's "patently" untrue; people innovated in software for years before the patent system was enforced on software. The patent system is just another unnecessary government intrusion into the marketplace.

As with most software developers, I think software patents are an egregiously bad idea; they reward people for NOT working (file patents instead of producing useful products). Patents may work in other fields, but they're simply inappropriate for software. Problem is, they'll keep going until a group works to CHANGE THE LAW and GET RID of software patents. There isn't a live group that's working to do that, to my knowledge; much of the anti-patent work recently has been in Europe, not the U.S. The "patent reform" work in the U.S. is primarily funded by large organizations who want to tweak the system, not eliminate it... they tend to be organizations who make money overall from the status quo.

(Log in to post comments)

Patents != innovation

Posted Oct 15, 2007 20:48 UTC (Mon) by and (subscriber, #2883) [Link]

> I generally agree that software patents could be eliminated or radically
> curtailed without affecting innovation.

Their elimination almost certainly would severely affect innovation in the
software industry. But in a _positive_ instead of a negative sense!

Patents on completely abstract concepts like software are simply a stupid
idea. A method convince non-technical people is to use an analogy from
fiction: If plots would be patentable half of the movie directors would
risk being sued for infringing the cinderella patent *g*.

Patents != innovation

Posted Oct 16, 2007 0:34 UTC (Tue) by bojan (subscriber, #14302) [Link]

> I'm reasonably sure that the people who applied for this particular patent were, in fact, doing work that was actual innovation and that they were doing it because the company they worked for (Xerox) was willing to invest a lot in bringing smart people together to generate new ideas.

I think you're making it sound a bit too much like "poor old inventors needed protection badly". This "invention" could have been protected by copyright (if they actually bothered to implement it, which would keep them in the job for a lot longer), which in case of software also gives holders a unique ability to ship binary only to the outside world (therefore further delaying what imitators can do). They could have also used contracts to ensure trade secrecy with the licensees.

So, these poor defenseless folks actually had two powerful mechanism to protect their work, none of which are available to say an inventor of a new auto part. Given that and given that software innovation doesn't require huge investment before beginning (a PC?) and also given that the field's landscape changes on a monthly basis, I think affording these people yet another 20 year monopoly (and on something that requires least effort when it comes to software development) is completely counter productive.

Patents != innovation

Posted Oct 16, 2007 0:54 UTC (Tue) by sepreece (subscriber, #19270) [Link]

I've already said I think software patents could go away without affecting innovation, so you're basically preaching to the choir. However, (a) they did implement it, (b) the problem with copyright is that it protects only the specific implementation, which is obviously much weaker (if the goal is to provide protection that increases the value of innovations) than patents (c) the problem with trade secrets is it's hard to protect a "secret" that's part of the user interface, and (d) in the mid-1980s the barriers to innovation in the field were significantly higher. These guys, and the rest of the PARC crew, came up with central innovations because they were smart, because they were working together, and because innovating was their full-time job, because their employer wanted to drive innovation (and reap the attendant benefits).

Things are very different today; that rate of change is an obvious reason why the 20-year terms was a mistake.

Not sure what the reference to auto parts is meant to suggest - all of these protections are available to people who invent auto parts (copyright only in a limited number of places). [And note that copyrights last a lot longer than patents and that the non-visibility of trade secrets is the key argument in favor of patents - by forcing inventors to trade disclosure for protection, it enables others to build on top of the patented innovations.]

Patents != innovation

Posted Oct 16, 2007 1:25 UTC (Tue) by bojan (subscriber, #14302) [Link]

As for (a), well then obviously they were to reap the benefits of copyright protection, right?

Regarding (b), that's not a problem, IMHO, it's a feature. And it is well suited for software, because software is essentially text.

In terms of (c), that's the whole point. The secret would refer to the implementation bit only, which is also protected by copyright.

As for (d), I started high school (year 9) in 1981 and that's when we received a donation of eight used Cromemco computers from UCLA. So, the doors to cheap stuff were opening big time in 1980's. By the time I started my university studies, PCs were everywhere. And I didn't live in a developed country at the time. I would venture a guess many middle class people in the U.S. could afford a PC of some sort in 1980's.

> These guys, and the rest of the PARC crew, came up with central innovations because they were smart, because they were working together, and because innovating was their full-time job, because their employer wanted to drive innovation (and reap the attendant benefits).

I don't doubt that they were smart. I don't doubt they came up with a lot of new stuff. I just think that double-dipping (or triple if you like) is way, way over the top.

> Not sure what the reference to auto parts is meant to suggest

It is meant to suggest that you cannot copyright an implementation of an auto part (you can the blueprints). And once it's out in a car, anyone can see it, so it's no longer a trade secret either. Hence, it's fair inventors gain patent protection for these things, as otherwise anyone could easily imitate without compensation.

On the other hand, if you're in the software business, you get to copyright both the blueprints (source) and the implementation (shall we say binary here), you also get to keep things secret by binary only distribution, obfuscation and trade secrets using contracts with licensees. And you get patent protection (for mathematical algorithms, no less) on top of all that.

The problem with governments in the U.S., Australia and similar countries is that they started equating progress with economic activity (i.e. such and such number of billions of dollars were made in patent trade) - two things that may be correlated, but are not the same.

Patent = disclosure for protection

Posted Oct 25, 2007 9:58 UTC (Thu) by forthy (guest, #1525) [Link]

Guys, remember that a patent was once granted to exchange disclosure of trade secrets for time-limited protection. This is what's IMHO covered with the original idea to "obvious to one skilled in the art". I.e. something that doesn't need a tantamount of reverse engineering is not worth a patent, because disclosing the blue print is not necessary.

The current legal interpretation of "obviousness" is rather different, but has recently changed a bit. Neither the old nor the new definition really takes the idea of the patent as such into account (an idea which was there before the industrialization took off at around 1860). A patent is not there to protect innovators, it's there to increase the overall amount of knowledge, by giving an incentive to disclose trade secrets. That's the deal, nothing else. For all things which can be copied without much troubles, there's copyright.

The whole ideas of copyright and patents don't fit in well with our modern understanding of state and economy, anyway. A patent is a privilege, granted by the state. Patents as such are a medieval idea; only in the 19th century, this idea was applied to innovations. The whole idea of a privilege is contrary to the idea of a free society.

A more appropriate concept for today would be to generally reward disclosure, and punish trade secrets. Free software culture shows that this can be done: not disclosing the source of a software is considered "evil", opening up is rewarded by community feedback. The state, if we need a leviathan to do this kind of stuff, could encourage disclosure with tax reduction. The open-sourcing of some software could be regarded as "charitable", and as consequence, tax-free allowances could be given to the project members (opposed to taxed salary). Same for innovations in other fields. This takes people right at their greed - even Bill Gates would open-source Windows, if he can save lots of taxes by doing so. Many people invest into tax-saving schemes even if it's a total loss.