[huhtenberg]

I frankly do not understand such black and white take on software patents. I don't think that all of them are evil. If one invests his time and money in developing a software algorithm - something non-trivial, say, an IFS compressor for binary data - why should such invention not be entitled to the same level of exclusive use protection that a mechanical design receives?

[JoshTriplett]

The previous story on this provided the perfect response: "If millions of people carried machine shops around in their backpacks, mechanical engineering would be incompatible with the patent system too." (http://news.ycombinator.com/item?id=3182882)

[anamax]

> The previous story on this provided the perfect response

Not really - it was wrong then/there and it's wrong here.

[eps]

I fail to see why you think it's a perfect response. Patents exists to stimulate inventors by protecting fruits of their labour from trivial copying. What the ease of copying has got anything to do with the patent system being broken?

[JoshTriplett]

What patents supposedly exist for does not match the net effect they have. (And remember, we only care about "stimulating inventors" because it gives the general public a net benefit, just like copyright theoretically exists to encourage the writing of more creative works.) People have this theoretical vision of patents protecting small inventors/businesses from some other company building the same product better/cheaper and putting "the little guy"/"the underdog" out of business. That romantic notion needs revision, because it doesn't work anymore, if it ever did. Patents today (and software patents in particular) tend to fall in three major cases:

- Small company gets a patent, thinks it will protect them, and either never uses it or quickly learns that their one patent will not protect them from the thousands held by whoever they want protection from. They also don't have the time or money to waste on litigation rather than just building a product.

- Large company gets a patent, adds it to the war chest, uses it as a giant stick to force anyone working in the same general area to pay up. Small companies have to pay up, and Free and Open Source Software gets excluded entirely (or just publishes their code anyway and says "screw patents", which works nicely for people not worth suing but makes anyone else using their code a target).

- Patent trolls, who write or more commonly acquire patents, wave them around like a stick, and extort money from people actually being productive.

One of these cases derives no benefit from software patents, and the other two turn software patents into net losses for the public. So, tell me again how software patents represent a useful tradeoff for the general public, and why we should grant an artificial monopoly that would not otherwise exist?

[hugh3]

But millions of people have workbenches and toolboxes in their garages, and are quite capable of building physical objects and devices.

[marshray]

But physical manufacturing outside of freestanding factories has never been a significant part of an industrial economy. It more or less amounts to "arts and crafts".

This changed with computers. You could never start a car factory in a garage, but there was a golden age where you could start a high-tech electronics company like HP or Apple.

Today you can start a web-based or software company on a laptop in a coffee shop or library. This is a big deal, let's not kill it.

Software patents have to go away or we're going to be stuck in the industrial age, which would be bad because we've already sold off most of our factories.

[anamax]

> But physical manufacturing outside of freestanding factories has never been a significant part of an industrial economy.

Actually, it has. Large scale production may have usually been in a factory, but often that factory was small production multiplied. Ford and the like were anomalies, even in the car biz.

> You could never start a car factory in a garage

I don't know about you, but literally thousands of folks did exactly that.

[marshray]

This is an interesting claim, do you have any links or citations?

[anamax]

http://en.wikipedia.org/wiki/List_of_U.S._cars

See also http://en.wikipedia.org/wiki/List_of_car_brands . Australia is sort of interesting as is Britain.

[JoshTriplett]

s/machine shops/nanoassemblers/ then, if you like; software still has a much smaller barrier to entry. And for the record, I don't necessarily believe in mechanical engineering patents either, but as a software professional I primarily care about abolishing software patents, because they hurt me personally.

[gruseom]

Why should it not be treated the way a mathematical result is? People invest time and effort into proving theorems too. It seems to me that software is ideas, like math is.

[huhtenberg]

Perhaps because math is a fundamental science and therefore funded in a completely different way?

[gruseom]

Yet your own algorithm example is a lot closer to math than to, say, a typical web app.