Live data from Hacker News

Poll: Do you support software patents?

news.ycombinator.com

11–20 of 319 posts

Re: Poll: Do you support software patents?

#11
post #3

This latest patent that caused an uproar was filed in 1992, when the idea was definitely innovative. It just further goes to show that the 17 year span, at least, is too long for software.

I'm simply not convinced that the concept of an update button acting on the Internet was innovative - not when I could have sat down in high school and worked it out in a day, and when the very concept of an Internet makes it obvious. And yes, I was doing software in 1992 (finished college in '88), so I think I understand pretty damn well what was innovative at the time.

[Edit on re-reading: I couldn't have sat down in high school and done this in a day because I had never heard of the Internet in 1980-1984, but I could have made it work against Compuserve.]

This, like every other "patent on three hours' work", is nothing more than parasitic rent-seeking behavior, and it is anathema to the vitality of our economy. Lodsys should be deeply ashamed of trying to present this as a "solution" when they clearly simply want to profit from the work of others.

Re: Poll: Do you support software patents?

#12
I think the software field is so fast moving that the current time period for a patent is not justifiable. But I do think that someone who invents something should be able to get some advantage out of it, rather than having a Google just waiting to take the invention and make it big.

So I say make the patent expiration period 30 months. This gives you substantial lead time, yet prevents issues like people fearing Apple's multitouch patents years after they've gotten the patent.

Plus it really takes patent trolls off the table, since these patents will have likely expired before they can get them. And it also takes off the table acquiring companies for IP.

Re: Poll: Do you support software patents?

#13

I've spent the last few years working on a new approach to developing software. I've spent a lot of my own money (at least $40,000 USD) on research to perfect it. Currently, from the research I've done, I've found no prior art. Should I be able to patent this technology? Should I just release it? If I just release it, wouldn't a large company patent aspects of it and sue me for infringement (it costs $ to prove prior…

If you just release it, it becomes prior art. And if you're not selling licenses, there's nothing to sue you over.

Re: Poll: Do you support software patents?

#14
My vote would be "Yes, if the Patent office approved only real innovative things". For example if I come up with some revolutionary face recognition technology, that seems novel enough to be patented. But if I just make a form with a submit button (Google's Patented home page), or Amazon's one click purchasing those are just ridiculous as a patents.

Re: Poll: Do you support software patents?

#15
I think the worst practical thing about software patents is that they create uncertainty. If there is one thing that is universally abhorred in the business world, it is uncertainty.

There needs to be a fast, cheap, reliable method by which small businesses can determine whether they are affected by anyone else's patent rights, and if so, by whose and to what extent. This way they can enter useful negotiations and adjust or abandon their plans as appropriate if patents are involved, or they can build their business free of unknown liability otherwise.

Given that I can see no viable means of ever constructing such a system, I don't see how patents or any similar legal tools will ever be anything but a tool for those with vast resources to crush competitors with few resources through means other than fair competition in the market.

Re: Poll: Do you support software patents?

#17

I've spent the last few years working on a new approach to developing software. I've spent a lot of my own money (at least $40,000 USD) on research to perfect it. Currently, from the research I've done, I've found no prior art. Should I be able to patent this technology? Should I just release it? If I just release it, wouldn't a large company patent aspects of it and sue me for infringement (it costs $ to prove prior…

Should I be able to patent this technology? Should I just release it? If I just release it, wouldn't a large company patent aspects of it and sue me for infringement (it costs $ to prove prior art).

If you publish your code, it becomes prior art. It doesn't cost much to stand up in court and say "uh, look guys, this code you're suing me over was published before you applied for your patent", and the case would be thrown out immediately.

In practice I doubt it would ever get that far: If there's one thing patent lawyers pay attention to, it's dates, and lawyers hate looking like idiots.

Re: Poll: Do you support software patents?

#18
Where do you draw the line between a "software patent" and a hardware one? Doesn't make sense to make a distinction based on the method of implementation (especially given that lots of HW is micro-programmable).

Yes - many silly patents are being filed for unreasonable reasons (fight patent wars, get a bonus, etc.) But making a distinction based on only on way of implementation is just as silly.

Re: Poll: Do you support software patents?

#19

I think the worst practical thing about software patents is that they create uncertainty. If there is one thing that is universally abhorred in the business world, it is uncertainty. There needs to be a fast, cheap, reliable method by which small businesses can determine whether they are affected by anyone else's patent rights, and if so, by whose and to what extent. This way they can enter useful negotiations and ad…

That wouldn't stop someone from patenting something that someone else determined was unpatented and then went after them for infringement.

Re: Poll: Do you support software patents?

#20
Since this is specific to software patents, my vote was to abolish them all. Patents on Software make not sense (other protections like Copyright, are essential, though). I have yet to see one "invention" that would not have seen the light of day without software patents.

Patents are a method to protect significant costs cost of an invention (to incentivize research). Having an idea for an algorithm is does not represent a great cost and it is not in the interest of society to grant a monopoly).

Post reply on HN