[rayiner]

Groklaw's editorializing has gone way downhill since the SCO days. E.g.

"I know the USPTO doesn't want to hear that software and patents totally need to get a divorce, but since most software developers believe that, maybe somebody should at least mention it to them, if only as a future topic for discussion. Most developers I know believe software is unpatentable subject matter."

Ask software developers who work in defense, aerospace, medical devices, telecommunications, video/audio processing, and other capital-intensive, R&D-intensive fields. I think you'll get a rather different answer. Indeed, as Google is finding out with the Motorola acquisition and the resulting patent situation, software engineering culture is far from uniform on this issue. The guys who write software that implements radio waveforms don't have the same views on patents as the guys who write software to optimize ad display, or for that matter the guys who write the search engines. Ironically, Google's PageRank is the subject of several patents, and Stanford's interest in the IP netted them $335 million in Google shares when they exclusively licensed it back to Google.

Although, I think this sort of round-table is a pretty good forum for figuring out more about the consensus view of software engineers...

[Silhouette]

Ask software developers who work in defense, aerospace, medical devices, telecommunications, video/audio processing, and other capital-intensive, R&D-intensive fields.

I've worked in several of those fields, and I have yet to meet these proponents of software patents you're talking about. Any advantage to having a patent on the results of your own R&D seems to be far outweighed by the persistent threat of being sued (with or without any actual merit) by someone else who claims you're infringing on the results of their R&D.

If something was useful enough and available enough for multiple parties to invent/discover/exploit it independently, it should be self-evident that patents are not a necessary incentive for that work to take place and can only be a barrier to further innovation for all-but-one of the inventing/discovering parties. This effect is magnified many times over if the patent is not for a specific piece of software but rather for something like a data format or communications mechanism. There is a reason you can't copyright the shape of a font in the US, and I think analogous arguments apply in these cases.

Fortunately, this is less of a problem for us here in the UK where the culture is not as litigious as in the US. Still, if nothing else, the repeated attempts by the US to export its onerous IP regime are a worry for small businesses here that using technologies potentially affected by software patents granted in other jurisdictions. The whole system is just one big barrier to competition for smaller organisations who can't play political/legal games to negate the whole technical issue as the big players do.

[rayiner]

> If something was useful enough and available enough for multiple parties to invent/discover/exploit it independently

What makes you think the technologies in question are invented independently? Take the Motorola Mobility patents, for example. Most of the patent licensees aren't in a position to hire all the experts and do all the R&D it took Motorola to develop that technology. They just use the end result of all that R&D produced by Motorola. If there was no patent system, they would just use the end result without compensating Motorola for all the work of developing the technology.

[Silhouette]

What makes you think the technologies in question are invented independently?

What makes you think they aren't? That's the problem with a lot of software patents: people can just run into them (or be accused of running into them and subject to the resulting legal proceedings) without even knowing they were there. Then, suddenly, possibly after years of their own R&D, they get hit with paying royalties on something that had nothing to do with their own development work.

If there was no patent system, they would just use the end result without compensating Motorola for all the work of developing the technology.

Perhaps, but if they're really only trying to make carbon copies of what Motorola were doing, then Motorola have a huge first mover advantage anyway. Specifically, they have as long as it takes to commercialise their R&D work before they have to disclose it, because the work is a protected trade secret like any other up to that point. So as long as they keep innovating in significant ways that others couldn't, they will remain years ahead of their competitors in the market. That seems like a natural and compelling commercial incentive to me, without any need to add artificial benefits via patents.

[belorn]

There will always be some who want to keep a broken system around if they can earn revenue from it. If money can be extracted, you can never get an complete 100% unanimous vote in fixing the system.

For people who owns land, some do not like regulations that forbid dumping toxins into it. But most societies know that forbidding the dumping of toxins are preferable over allowing it because of the ecological damage the toxins causes to society.

Patents causes an ecological damage to the software community. It might be profitable for a few, but for society as a whole it is a net loss. So long the cost-benefit analyze shows that, patents should have no place in software. Considering that patents are granted monopolies given by the government and enforced by the state, if society can't get more from software patents than it cost, society should not be in the businesses of granting said patents.

[gte910h]

I have worked in defense. Getting a patent was like "Oooh, that's cool". Finding out if there were patents in the way to do anything was weeks of supporting legal's searches.

I'm sure someone, somewhere, at some level, likes patents. I don't think it's engineers though.

[rayiner]

I also worked in defense. The company was run by engineering PhD's and the end result of millions of dollars of R&D and man-years of labor was a set of patents.

Your line engineers aren't going to get super excited about patents, because they're not compensated for their innovation (that's a separate rant), but it wasn't my assertion that they'd come to a rousing defense of software patents. Rather, they're not going to come out to a rousing opposition to software patents like Groklaw claims.

[tracker1]

As far as pure software development (not using at that point integrated custom hardware) as an idea to form, I haven't seen anything in over twenty years that deserves a patent. That includes audio, video and other compression systems that build on algorithms that are over two decades old. Those systems that tightly work with specialized hardware are a bit of a different story.

As far as any system that is defined only in software, and capable of running on a generic computer, with generic input/output devices of today are not, imho, patentable. Anything that involves an obvious idea implemented in an obvious way is not patentable, of which the incredible majority of software is. Again, exception to interfacing specialized hardware...

[gte910h]

You mean "They won't publicly say they dislike something their company likes" That's true, for some of them. Others I can still see them saying "patents are more trouble than they're worth because they're outspoken types.

[noonespecial]

Ask them if they'd be OK with a patent that simply read "using a computer to transmit a radio wave" and see if they'd get behind trolls suing every texi-cab company that uses a dispatch radio tuned by a PLL and a microcontroller.

That might help them understand where this is headed (and already arrived in many cases).

They want to patent that brilliant little algo they spent months on to eek out that seemingly impossible spread spectrum noise piercing scheme that they finally got working. That's not how it happens at the patent office these days. Worse still, after all those months of creating something genuinely usefull, they're likely to be sued by someone who's never touched a radio, or made a product, and has a patent that reads more like "walky-talkies with teh interwebs inside".

[magicalist]

> Ironically, Google's PageRank is the subject of several patents, and Stanford's interest in the IP netted them $335 million in Google shares when they exclusively licensed it back to Google.

That's not irony, as "should not be patentable" and "is patentable" are not mutually exclusive (I've also never heard of anyone being sued over the many, many pagerank-like implementations out there, so really the only thing going on there is University patent policies).

"should not be patentable" and "is patentable" is actually a large part of the problem, because you have to engage in the game to not be sued out of existence, but in doing so, you're making the environment even more toxic.

I always like to quote the Gosling story:

"In Sun's early history, we didn't think much of patents. While there's a kernel of good sense in the reasoning for patents, the system itself has gotten goofy. Sun didn't file many patents initially. But then we got sued by IBM for violating the "RISC patent" - a patent that essentially said "if you make something simpler, it'll go faster". Seemed like a blindingly obvious notion that shouldn't have been patentable, but we got sued, and lost. The penalty was huge. Nearly put us out of business. We survived, but to help protect us from future suits we went on a patenting binge. Even though we had a basic distaste for patents, the game is what it is, and patents are essential in modern corporations, if only as a defensive measure. There was even an unofficial competition to see who could get the goofiest patent through the system. My entry[1] wasn't nearly the goofiest."

You make market incentives to get companies to patent everything they can. Companies incentivize engineers to patent everything they can. Patent lawyers rewrite the software patent applications to take what is already an abstract set of instructions and turn it into the platonic form of that idea, to make sure they cover all implementations of that idea. The patent office accepts it after 7 resubmissions. You're not going to end up with a good result.

[1] http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sec...

[gnud]

> The guys who write software that implements radio waveforms don't have the same views on patents as the guys who write software to optimize ad display

Can you elaborate?

It seems bizarre to me that formulas can not be patented, but telling a general-purpose computer to execute that formula, can.

[rayiner]

It depends on your definition of "general purpose computer." Technically, a signal processor buried in a radio may be a "general purpose computer," but as a practical matter it's no different than an ASIC that implements the same logic in hardware.

[raverbashing]

I'm sure some engineers like patents. It makes them feel important, it's something to show to their bosses (and parents) and keep a score for promotion purposes.

Managers also like patents to schmooze their bosses.

"See, I did a patent, that's like, very important"

Meanwhile the lawyers say "yeah, sure" while collecting heavy fees.

And 90% of R&D is on the finished products (incl. code/blueprints) not on the "invention".

[rayiner]

I don't think that's an accurate statement of how patents function in these industries.

Patents provide a framework for companies to transact in technology while protecting their R&D investment. It enables companies like MIPS and ARM to specialize in producing designs without having to develop finished products, and it allows companies to invest in open standards like LTE or WiMAX without having some Chinese manufacturer just take the end result and produce cheap chips implementing the standard at low cost because they didn't have to undertake any R&D investment.

[raverbashing]

ARM/MIPS are not only about patents, but the actual logic design of the processor

It isn't hard to create an "ARM compatible" processor, but you can certainly not sell it as this without infringing a trademark (and copyright on the instruction set, etc).

With WiMax and similars, again it's not only patents, but also trademarks that are only licensed to companies that pay the fee and usually have to pass a compliance test as well.

Sure, patents are present but they're only part of the story.

Also, patents are not the only way to protect an invention, the two other common methods are secrecy and publishing it (which prevents anyone else from getting a patent on it)

[gte910h]

MIPS and ARM are hardware, not software.