5 ms·

Or maybe US citizens could vote for someone that invalidates software patents. They're not valid in other parts of the world.

GirlsCanCode 10y ago
That's not completely true.
twblalock 10y ago
That's not true. Lots of countries allow software patents. The rules are sometimes stricter than in the US, but they definitely exist and are considered valid by the courts in the UK, the EU, Canada, Japan, and South Korea, among others.
megablast 10y ago
Sure, that should be the overriding issue that people use to vote for someone?
- NetStrikeForce 10y ago
  It's not a small thing to just brush off, but it depends on your priorities.
jwatte 10y ago
If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for i inventive hardware/chemistry/biology/mechanics. The problem comes when the bar for "inventive" and "novel" and "non-obvious" is too low, and a guy or girl in a garage facing the problem to be solved is just as likely to reinvent the solution independently. Often, it's the problem that shapes the solution, rather than any particular inventiveness. It turns out, other engineering disciplines also have this problem, but because their times to market are much longer and more expensive, and their times of prior art are much longer (and thus the foundations are more well known/documented) it doesn't quite rise to the immediate existential threat that it does for independent software developers.
- nkurz 10y ago
  If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for inventive hardware/chemistry/biology/mechanics. I'm not sure that follows. Do you think the ability to keep and preserve trade secrets is the same for all those fields? Keeping a trade secret in software that runs on end user' computers seems much more difficult than keeping the technique used in a chemical reaction a secret. I tend to think that just about everything software related would otherwise be out there and available. I'm also not sure that coaxing trade secrets into the open is a significant reason that patents are good. At least, I have trouble thinking of cases in any of those fields (much less software) where I think that the overall rate of scientific progress in the world is better as a result of knowledge disclosed in a patent. Do you have examples? In particular, for something software related that is better known today because it was disclosed in a patent which has since expired?
- AnthonyMouse 10y ago
  > If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for i inventive hardware/chemistry/biology/mechanics. There are important ways that software is different. One is that software is inherently abstract. If you're making a battery or a machine or a drug, it matters what materials you use, what the temperature is at different points, how long the product lasts, how much it costs per unit to manufacture, etc. In software there is none of that. But without any of those constraints the inherent result is unreasonably broad patents. This combines on the other end with the fact that software has effectively zero reproduction cost, so there is trivial cost for adding support for things. Which means combinatorial explosion. "You asked for a banana but what you got was a gorilla holding a banana and the entire jungle." If you're making batteries and you use A, B and C then it doesn't affect you if someone else has a patent on the combination of D, E and F. If you're making software there is no unit cost to adding support for everything from A to Z, so you do. But then it doesn't matter which combination of things someone patents, you end up unwittingly infringing it because you support everything and the patent covers some subset of everything. The result is unworkable. The system can't function when everything infringes everything.

GirlsCanCode 10y ago

That's not completely true.

twblalock 10y ago

That's not true. Lots of countries allow software patents. The rules are sometimes stricter than in the US, but they definitely exist and are considered valid by the courts in the UK, the EU, Canada, Japan, and South Korea, among others.

megablast 10y ago

Sure, that should be the overriding issue that people use to vote for someone?

NetStrikeForce 10y ago
It's not a small thing to just brush off, but it depends on your priorities.

jwatte 10y ago

If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for i inventive hardware/chemistry/biology/mechanics. The problem comes when the bar for "inventive" and "novel" and "non-obvious" is too low, and a guy or girl in a garage facing the problem to be solved is just as likely to reinvent the solution independently. Often, it's the problem that shapes the solution, rather than any particular inventiveness. It turns out, other engineering disciplines also have this problem, but because their times to market are much longer and more expensive, and their times of prior art are much longer (and thus the foundations are more well known/documented) it doesn't quite rise to the immediate existential threat that it does for independent software developers.

nkurz 10y ago
If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for inventive hardware/chemistry/biology/mechanics. I'm not sure that follows. Do you think the ability to keep and preserve trade secrets is the same for all those fields? Keeping a trade secret in software that runs on end user' computers seems much more difficult than keeping the technique used in a chemical reaction a secret. I tend to think that just about everything software related would otherwise be out there and available. I'm also not sure that coaxing trade secrets into the open is a significant reason that patents are good. At least, I have trouble thinking of cases in any of those fields (much less software) where I think that the overall rate of scientific progress in the world is better as a result of knowledge disclosed in a patent. Do you have examples? In particular, for something software related that is better known today because it was disclosed in a patent which has since expired?
AnthonyMouse 10y ago
> If you believe that patents, in general, are more good than bad, because they coax trade secrets out into the open after an amount of time, then they are just as valid for inventive software as for i inventive hardware/chemistry/biology/mechanics. There are important ways that software is different. One is that software is inherently abstract. If you're making a battery or a machine or a drug, it matters what materials you use, what the temperature is at different points, how long the product lasts, how much it costs per unit to manufacture, etc. In software there is none of that. But without any of those constraints the inherent result is unreasonably broad patents. This combines on the other end with the fact that software has effectively zero reproduction cost, so there is trivial cost for adding support for things. Which means combinatorial explosion. "You asked for a banana but what you got was a gorilla holding a banana and the entire jungle." If you're making batteries and you use A, B and C then it doesn't affect you if someone else has a patent on the combination of D, E and F. If you're making software there is no unit cost to adding support for everything from A to Z, so you do. But then it doesn't matter which combination of things someone patents, you end up unwittingly infringing it because you support everything and the patent covers some subset of everything. The result is unworkable. The system can't function when everything infringes everything.