[jjeaff]

I think it's worse than the patent office rubber stamping everything. They actually don't rubber stamp things that your average, small inventor might try to patent. They seem to always initially reject applications, requiring multiple submissions with small modifications over and over until they relent and grant.

This serves to filter out everyone except those with deep pockets and highly paid patent attorneys.

When I filed for a patent a few years ago, we got back prior art from the patent office that was very far fetched. We paid $3500 to the attorney for the initial application and then each time it would come back, we would pay a few hundred more.

Our patent attorney estimated that we would end up paying around $10k, but he was fairly certain that we would get through. He said most of his applications followed this same pattern.

We decided to cut our losses at around $5k and move on. A few years later, some large company patented very similar technology. We submitted prior art but they just had to pay for a few more iterations to get around it.

[matthewdgreen]

Patent examiners are given a relatively limited time/resource budget to examine applications. For that reason they will often reject early as a strategy [ETA: often, in my experience, with very weak "prior art" that basically matches some keywords], then allow more dedicated inventors to revise. This is mainly a filtering strategy, as best I can see. When your attorney quoted you $10K and suggested the patent would eventually be accepted, they were speaking as someone who understood the business and how the process would play out. It's surprising to me that you disregarded their advice and sunk $5K into the process only to walk away. Better to spend nothing than to waste $5K and get nothing from it.

[jjoonathan]

Your surprise comes from a perspective of understanding and trusting the process, presumably from experience. Is it really such a surprise that someone with a different experience -- that of extending trust and having it betrayed -- would find it hard to extend trust a second time?

And yeah, this system seems to be the worst of all worlds: needlessly punishing for small players and needlessly lucrative for large ones.

The patent system, as it is implemented, clearly behaves as an instrument of power, not an instrument for fixing incentives.

[matthewdgreen]

I'm not here to defend the patent system. But in general, patent lawyers are your guide to that system. If you don't trust your lawyer's instincts, by all means get a second opinion! But don't just ignore their advice.

[auggierose]

It's ridiculous that you have to spend $10K to patent your idea. Let's just get rid of all those patent office people and lawyers, and save everyone a lot of time and money.

[pclmulqdq]

The $10k is a feature, not a bug. Society shouldn't allow patents to be filed without commercial intent (because otherwise you are depriving the world of an idea for 20 years without meaningfully using it), and a $10k pricetag is a good signal of commercial intent. In the EU, patents have escalating fees to make sure that patent rights are only guaranteed as long as a commercial enterprise continues to be successful.

And no, the intent to file and then sell the patent is not a good "commercial intent."

[auggierose]

Which means that for rich people / patent trolls what society allows doesn't matter. Patents are definitely a bug, at least in software. Time to get rid of them.

[robertlagrant]

Patent examiners are given a relatively limited time/resource budget to examine applications

But they still take the money. If you take the money you should do a good job.

[jjeaff]

The $10k was his high end estimate. We thought we would try rolling the dice and see if we were getting anywhere. After the first $5k, it didn't look like we were going to luck out and end up on the low end.

[tgsovlerkhgsel]

The more examples I hear, the more I think patents should just be thrown out altogether in many cases.

At the very least, software patents in the sense of "if your implementation uses exclusively general purpose computers and software, it cannot be violating a patent".

[pclmulqdq]

Alice vs CLS bank threw out most patents saying "do X on a computer." Patents basically have to have a physical effect or give something a property it wouldn't otherwise have, so compression (saving bits) and encryption (giving security properties) are patentable, but most things like web apps and online shopping carts (an infamous patent troll sued many people over this pre-Alice) are not.

[tgsovlerkhgsel]

> so compression (saving bits) and encryption (giving security properties) are patentable

Which is a huge problem, because now you again need an army of lawyers just to determine whether something is patentable, and can redesign/argue around stuff to make it look patentable even where it should not be.

It would be much nicer if you could say "Here is a mass market consumer grade laptop I purchased at Walmart this morning. Here is the software running on it. Your patent claim is invalid. Your honor, I'd like my legal fees reimbursed and the plaintiff's attorney sentenced to sixty lashes for wasting everybody's time".