Ask HN: Why is there no "execution" rule for patents?

[frisco]

At my company, in addition to working on our core competency, we occasionally look at cool, compelling side projects to branch out to in order to grow the company. Recently, though, I've been consistently frustrated by existing patents blocking us from executing on really compelling, valuable concepts owned by people who aren't doing anything with it! At the end of the day, a lot of these things we want to use ourselves and would pay for them I'd they were offered. Are there subtle (or not subtle) arguments against a "to qualify for patent protection you actually need to build this" clause? Ideamen are usually annoying but harmless, but more and more i'm seeing them actually impede development and innovation.

[arockwell]

This allows a lone inventor to patent a device and then license plans to a large company to manufacture and sell the product.

I agree that this might help curtail patent trolls, but I am not sure this will help that much since the way patents are granted seems fundamentally broken. I think requiring someone to actually produce the invention is treating the symptom of patent trolls and not the cause. If the USPTO stopped granting patents when there is prior art or stopped granting patents that are "obvious" would help more.

[frisco → arockwell]

My experience has been with the inverse of this, though. I've been seeing Microsoft and friends reserving literally thousands of patents on things that aren't even tangentially related to anything they're working on, in language so broad it rules out huge areas of development.

[arockwell → frisco]

This is true, however for these large companies many of the patents are filed as a defensive measure. Its fairly uncommon to hear them sue someone for patent infringement, and they probably have a patent on just about anything (MS has a patent on the strlen algorithm for example, its really insane).

In other words, worry about getting sued by a patent troll not a big company. Also, perversely, I think you better off not even looking for patents in the first place because one of the factors that can go against you in a patent lawsuit is whether or not you were aware of the patent in the first place.

[fatdog789 → arockwell]

The problem isn't the USPTO, it's the patent applicants themselves who do not disclose the prior art in their own applications. The USPTO is supposed to conduct its own prior art search, but the volume of applications, coupled with funding which decreases every year, results in a very minimal search for direct prior art.

Patents can and often are invalidated for the patentee's failure to include prior art; in fact, it is the leading cause of patent invalidation.

The problem is that the USPTO doesn't have the number of employees it needs to examine each patent, as fully as would be necessary to stop "obvious" patents from getting through, within a time frame for allowing meaningful inventions to hit the market in a timely fashion.

[gravitycop → fatdog789]

The problem [is] the patent applicants themselves who do not disclose the prior art in their own applications. [...] Patents can and often are invalidated for the patentee's failure to include prior art; in fact, it is the leading cause of patent invalidation.

That seems to differ from this:

http://robertplattbell.blogspot.com/2009/01/prior-art-search...

Why would I need one?

The simple answer is, you don't. There is no requirement to search the Prior Art for previous inventions. And increasingly, many folks are shying away from searching. If you do have a search done, however, you are required to cite the referenced discovered to the Patent Office, should you later file a Patent on the same invention.

[arockwell → fatdog789]

I think the problem is more structural. All of USPTO's funding comes in the form of patent applicant fees, so they have an incentive to let us many through to encourage others to file ridiculous patents as well. The way the system is set up its really not in their interest to act as an effective judge.

[cdibona]

Two thoughts:

1) Never, ever, ever do patent investigations before coding. If you for some reason still build something and are sued, it's called willful infringement and awards triple.

2) Successful patent litigation or settlements on non-executed (or submarined) patents are rarely if ever successful. It does happen, but it is very rare.

3) Patent litigation happens, but rarely is it tried against small fry (read small fry as meaning unprofitable, less than say 20m in the bank, non-public and smallish, less than 100 employees,etc..) It just almost never happens. If your company is the one you link to, stop worrying about it. You are too small to bother with. Anyone threatening you will likely buy you or be themselves out of business before you can get anywhere near where you could make a court appearance.

4) People will threaten. So what. It can take many (4 ->7) years for a patent case to get to trial, your lawyers on retainer can do all the wrong things and it will still take years and years. You can settle any moment before a ruling, so why do it immediately?

Anyhow, I'm not a lawyer, but I've been in this space for a very very long time. Good luck.

[cdibona → cdibona]

Okay, so that was more than 2 thoughts.

[wmf]

Most patents are never used in anger, so you should probably just ignore them. There are other existential business risks that are probably bigger than patents.

[likpok]

If you're willing to pay, you could try approaching they owners to see if they're willing to sell/license.

[captainobvious → likpok]

If it involves more than one patent owned by different owners it is a classic tragedy of the commons problem: each patent holder has an incentive to maximize his demands because the downside (risk that the overall patent cost is too much for the purchaser) is shared by all the sellers (the patent owners).

[anamax]

> Are there subtle (or not subtle) arguments against a "to qualify for patent protection you actually need to build this" clause?

What does "build" mean? Can I get an idea, file a patent application and then start building? Or, do I have to build before I file? What if it takes longer for me to build than it takes the patent to issue? (Or, I can't legally sell when the patent issues, as is often the case for drug companies.)

If I make one instance and try to sell it, have I satisfied the "build" constraint? Or, do have to successfully sell stuff continuously? What if I can't legally sell because the thing that I'm building requires the use of IP that I can't get a license for?

[fatdog789]

Expense. What if the invention is very expensive to make but the inventor is a high school dropout living above his parent's garage? This one requirement now completely invalidates his rights to the invention he created; this means less incentive for people to invent. It also applies against companies as well, b/c now they have to invest resources in researching an invention and in developing the production line to crank it out if it works.

Timeline. You require them to build the product before it can be patented, which means a massive incentive before they even know if they'll get any protection.

REDUNDANT. Patent law already has a "reduction to practice" requirement. This requirement usually requires the inventor to build a working copy (and this is the most common way that the RTP requirement is met) or to provide detailed enough instructions that any person in the same art could build a working copy of the invention.

[frisco → fatdog789]

How would it invalidate his right if he's made it? That wasn't really what I meant, though-- the requirement would be more that you need to actively be pursuing the thing you're patenting, instead of simply reserving the idea. I don't think requiring it do be made _before_ a patent to be issued would be a good idea, but just the _idea_ isn't worth much, unless it's _really_ nonobvious.

[gravitycop → frisco]

I don't think requiring it do be made _before_ a patent to be issued would be a good idea

The courts that have ruled on these things apparently think it is a good idea.

http://robertplattbell.blogspot.com/2007/10/file-early-file-...

In Pfaff, Justice Stevens wrote:

"Second, the invention must be ready for patenting. That condition may be satisfied in at least two ways: by proof of reduction to practice before the critical date; or by proof that prior to the critical date the inventor had prepared drawings or other descriptions of the invention that were sufficiently specific to enable a person skilled in the art to practice the invention." (Id., at 12)

Perhaps you do not know that patents are enforceable back to their original filing dates (making their issuing dates largely irrelevant), and that it takes several years for patents to issue. What is your reasoning that led to your conclusion that patents should issue before they are qualified to issue?

[frisco → gravitycop]

> What is your reasoning that led to your conclusion that patents should issue before they are qualified to issue?

The observation that there are a _ton_ of patents issued to "inventions" that aren't even developed yet. That might theoretically be a bar to a patent being issued, but in practice all you need is the idea, the money, and a write-up.

[gravitycop → frisco]

there are a _ton_ of patents issued to "inventions" that aren't even developed yet.

That is in the current climate of not legally needing to reduce an invention to practice. If, instead, patent-seekers were required to actually produce working copies of their inventions (reduce to practice), we might see them regularly do that before the critical dates. I believe that that was how it was traditionally done, until recent times.

[gravitycop → frisco]

the requirement would be more that you need to actively be pursuing the thing you're patenting, instead of simply reserving the idea.

Perhaps the patent-holder could be required to maintain a website that shows how to reduce the invention to practice. However, all US patents are already published online, and, by definition, they all show how to reduce to practice. A caveat is that patents can sometimes remain secret for a number of years before being disclosed by the USPTO. Besides that, patents are enforced at least back to their initial filing dates (not their issuing dates), and therefore a company could be working on a product for several years without having any way of knowing that a patent on that product had been applied-for before they started.

-

The USPTO does have a method in place for encouraging patent-holders to market their inventions. It is the patent maintenance-fee schedule:

http://robertplattbell.blogspot.com/2007/10/understanding-ma...

U.S. Maintenance fees are progressive, so each successive fee is much more than the previous one. The idea is to encourage inventors to allow their Patents to expire unless they are actually being used. The Government does not want inventors to “sit on” Patents with the hope that maybe somebody will someday infringe, and then they can sue. With the Maintenance Fees, it is “use it or lose it”.

Maintenance Fees are raised periodically along with all other Patent Fees, so it is hard to say exactly what they will be. At the time of this writing (July, 2005) the fees are as follows:

    YEAR      Large Entity     Small Entity
   3 Years:        $900             $450
   7 Years:      $2,300           $1,150
  11 Years:      $3,800           $1,900

If these maintenance fees are not presently doing the job, perhaps they should be raised. How about multiplying the figures above by 1,000? That way, the final maintenance fee for a large entity would be $3.8 million.

[anamax → gravitycop]

> A caveat is that patents can sometimes remain secret for a number of years before being disclosed by the USPTO.

That's true only if "a number of years" is intended to mean "one year". With the exception of national security stuff, US patents are now published one year after application. That change happened several years ago.

> Besides that, patents are enforced at least back to their initial filing dates (not their issuing dates),

Since US Patents' term (if issued) also starts with their filing date, why is that objectionable? (Yes, if a US application takes 5 years to go from filing to issue, filing to issue time counts as part of the 20 year term.)

[gravitycop → anamax]

With the exception of national security stuff, US patents are now published one year after application.

http://robertplattbell.blogspot.com/2009/01/prior-art-search...

If a Patent is "pending" it is kept secret for 18 months, then it might be published. Some are not published (it is an opt-out system) others filed under the old rules (pre 2001) are not published, period. So it is possible that a patent may be "pending" that you don't know about. If this product has been on the market for several years, the likelihood of it being "pending" diminishes accordingly. But some Patents take 5, 6 or more years to issue. So you may never have a concrete answer whether one of your competitors has filed for a patent or not.

Additionally, a patent can be reserved, almost two years before filing, by the following sequence:

  1. Perform an Invention Disclosure before a credible witness.
  2. Wait almost one year.
  3. File a Provisional Application for Patent.
  4. Wait almost one year.
  5. File an Application for Patent.

So, normally, there could be 3.5 years of secrecy. If USPTO publication is opted-out, or if the filing was done under the pre-2001 rules, the secrecy term could be longer, possibly to infinity.

[anamax → gravitycop]

"On filing of a plant or utility application on or after November 29, 2000, an applicant may request that the application not be published, but only if the invention has not been and will not be the subject of an application filed in a foreign country that requires publication 18 months after filing (or earlier claimed priority date) or under the Patent Cooperation Treaty. Publication occurs after the expiration of an 18-month period following the earliest effective filing date or priority date claimed by an application."

http://www.uspto.gov/web/offices/pac/doc/general/index.html#...