Unified Patent Court is more expensive and an SME killer | Hacker News Reader

Unified Patent Court is more expensive and an SME killer

ffii.org

103 pointszoobab56 comments

The 20k and 11k fees here are for invalidating and proving infringement respectively. The second one should dissuade trolls who want to target small users. The 20k fee is not to defend yourself from an infringement claim, it is to invalidate a patent. That is very different than defending yourself.

The equivalent filing in the US is about the same price, and the process of invalidating a patent is long and complicated, so in any jurisdiction you should expect to pay hundreds of thousands in legal fees for that process. On top of those fees, 20k isn't that much.

Shouldn't invalidation of a patent simply be a matter of presenting evidence of prior art? I can imagine that you'd want to mandate that the evidence is presented in a certain format and possibly charge a very nominal fee (both to avoid time-wasters) but nothing more. The bias should always be towards invalidating patents, given the damage they do to competition.

There are a lot of different ways of invalidating a patent, and that is the simplest. However, that is also nominally what the examiner did during the issuance of the patent, so unless there was gross incompetence (eg the "shopping cart" patents), it's not that simple.

I've filed several patents and never seen any evidence that the examiner did anything apart from stamping paperwork.

Type the application number into Global Dossier if it's published, and take a look at the "file wrapper" (term for all the documents listed). You'll get a better idea of what the examiner did. https://globaldossier.uspto.gov/

In the US at least, they typically put time stamped search logs in the patent file wrapper, so you can get a fairly good idea of what was searched and how long that took. It seems to me, however, that few patent attorneys will look at anything other than the "office actions", and they will look only at parts of those. Other jurisdictions don't seem to put anywhere near as many documents in patent file wrappers as the US does.

Here's the first patent I filed (and wrote(!)) which was in the UK. Can I search what the examiner did on this? https://patents.google.com/patent/WO1998024208A2

The latest one I coauthored is a US patent which might easier to research: https://uspto.report/patent/app/20210034401

On Google Patents you can get a shortcut to the USPTO Global Dossier if you search for "External Links". Looks like the USPTO Global Dossier is down for WIPO documents right now but the Espacenet Global Dossier is up: https://register.epo.org/ipfwretrieve?apn=GB.9703212.W&lng=e...

You can see the search report for the WIPO document you linked to there. Often the descriptions are inaccurate, so in this case even though it just says "Published International Application" for all 3 documents, it's not necessarily that. The one dated 05.11.1998 is the search report. This one is not very detailed.

The EPO version has a lot more: https://register.epo.org/application?tab=doclist&number=EP97...

For the second one you mention, the USPTO Global Dossier lists a bunch of documents: https://globaldossier.uspto.gov/#/details/US/16530248/A/7839...

You can get there by changing the "Type" drop-down menu to "Pre-grant Publication" and typing in 20210034401. You can look at a lot of search logs for this one. You can also read the "Notice of Allowance and Fees Due (PTOL-85)" to see why the examiner issued this patent.

Echo the other answer to this. If you know where to look, you can find what they searched.

I'm not saying it was a thorough search, but it was a search.

The patent examination process is not well understood by most people. Once a patent is received, the patent examiner carefully places the application into a room full of other patent applications.

Then kittens are released into the room with PATENT GRANTED stamps affixed to their feet.

The kittens are then returned to their holding area to await the next round of patent examination.

The patent examiners collect the applications from the floor and look to see which patent applications were granted.

This is from an old HN post: https://news.ycombinator.com/item?id=21000034

Invalidating a patent to me is an admission that the patent shouldn't have been granted in the first place. I appreciate that there is a long legal battle to get there but the patent authority then charging a fee feels like them getting money for their own failings. If anything they should be paying those people who's business was in legal limbo compensation for their error.

So then, instead of accepting that patent offices can't find all relevant prior art/prior use (disclosed anywhere in the World) you require perfection in the discovery of relevant 'art' (closely related technical documents).

That shifts the problem from "occasionally a patent is granted that isn't valid" to "it will cost $billions for even a single patent just to do a complete search" (more realistically this isn't possible you'd have to exhaustively search all prior publications).

IMO if you want a patent system you have to make it work without expecting perfection from patent offices.

> "occasionally a patent is granted that isn't valid"

ROTFL. It's more like "with software, the patent is virtually always invalid."

I'm sure there are people better than I am, but I can invalidate practically any software patent. And I did it for Google, including in Germany in the Google Maps case of 2014.

The average time an examiner is given for a patent is a day and a half (old data). That is not nearly enough time to find the "relevant prior art." So what the patent offices are doing is handing out worthless government "monopolies" and making tons of money in fees while they're doing it.

>ROTFL. It's more like "with software, the patent is virtually always invalid." //

Yes, the establishment of prior art is particularly hard for software IMO, but that's mainly not an EPO issue (per the OP) as software patents in Europe are much more restricted compared to before the USPTO.

Your timing is broadly right IME, probably a bit generous for less complex areas (though it varies and EPO spend a bit longer).

Note that in part it's an industry issue - if software companies wanted to help them they can point out prior art before parents are granted (it's part of the process in EPO and USPTO jurisdictions).

>I'm sure there are people better than I am, but I can invalidate practically any software patent. //

EPO _granted_ patents? I'm highly doubtful you can do that in the normal search period (0.5-1.5 days).

If you want to argue that far more resources should be devoted to patent examination; that's possible but in EPO countries the balance seems about right -- countries have a hard job arguing for thousands more civil servants to save only a few (!?) court cases. In USPTO you need an overhaul of the court system, massive increase in examining staff I don't think we'll fix it.

All just my opinion of course.

> "EPO _granted_ patents? I'm highly doubtful you can do that in the normal search period (0.5-1.5 days)."

"In the normal search period" is the key phrase here. If you're granting a government monopoly which is extremely costly to challenge (as the article demonstrates), then you owe us more than a cursory look at the prior art.

In any case, this is probably an argument for just making software nearly impossible to patent, even more than the EPO already does. As (I think) Joel Spolsky said, no more than 5 or 6 per year sounds about right.

Yeah, with product lifetimes and development in general being on a much shorter cycle parents should probably have reduced in term to reflect that.

The remark about paying compensation was slightly in jest. I do think it's unreasonable for there to be a fee for issuing a patent as well as invalidated one though. Expecting perfection is a long way from expecting their incentives to be aligned with the desired outcomes.

Correct. The PTO (speaking of the US here) takes in way more money in fees than it spends, and doesn't even get to keep that money to improve its searching.

Isn't invalidating the patent one of the defenses you have against a patent troll? If they don't back off you may cost them all future business for that patent. This makes that potentially much more expensive to do.

It is, but it's a VERY aggressive line of defense. The OP was saying that this is an "SME killer" and that "No small company will be able to defend itself" against a patent suit. This is far from the case. A small company was never going to try to invalidate a patent to defend itself either way.

> On top of those fees, 20k isn't that much.

If we are comparing apples to apples, the actual court filing fees and such in the US are pretty small ($350 to sue in Federal court + $400 to file documentation), and if you cannot afford these fees the court has the power to waive them. For a small business with a clear case,this means your $20K is possibly enough to pay the lawyer and see the case though. Just recently, on this site, we had an article about SparkFun rebuffing a patent troll in US Federal Court for less than (or close to, as I recall) the EU filing fee.

> On top of those fees, 20k isn't that much.

Yes, the issue is when the fees are charged. A filing fee prevents discovery from ever happening. It is basically saying, you are too small, too poor to dispute a patent. That is not justice.

"That is very different than defending yourself." The counterclaim for validity court fees are set at 20.000EUR under the UPC.

My company does not have those 20.000EUR, especially nowadays. Similar action in front of Belgian courts cost 165EUR, it is a 120X increase according to my calculator.

> FFII says the proposed Unified Patent Court is an SME killer with its super-expensive court fees of 20.000EUR. No small company will be able to defend itself if it is accused of violating a patent, the proposed UPC court fees will deny access to justice for small companies, the cost being 100X more expensive than the current situation in the different countries.

I can't help but wonder if this is really just an honest mistake by the EU.

>just an honest mistake

I wouldn't think so, these are legislations that have every word and number fought for by the legislators, this is likely a concession to some legislator that stands to gain from this.

> I can't help but wonder if this is really just an honest mistake by the EU.

Unlike software patents, it's less clear to me that this is worse than the alternative.

The EU patent office is horribly unfair to individual inventors. There are not even fee differences for a kid in their bedroom vs Apple when filing a patent.

Take data also, EPO will charge you hundreds of thousands of dollars for patent data. USPTO gives it to everybody for free.

"Take data also, EPO will charge you hundreds of thousands of dollars for patent data."

This appears to be wrong, the full dataset is available as a free download: https://data.epo.org/linked-data/download/

There is also a freely available API, with documentation at: https://data.epo.org/linked-data/documentation/api-overview....

I am not talking about the patent images. That's not the data that most people are interested in. I am talking about ESpacenet, OPS, etc. which are the ones that people care about.

OPS data can be downloaded with a free account (up to 4GB per week): https://www.epo.org/searching-for-patents/data/web-services/...

ESpacenet appears to be available as at the very least a completely free online search: https://worldwide.espacenet.com/patent/

Am I missing something?

4GB is nothing for a dataset of that size and Espacenet is certainly available online, but it is heavily rate limited and they do not just allow you to download the full thing.

The data was only a side point. My main frustration with the EPO is the fee structure.

What is an SME?

Subject matter expert? Small and medium enterprise? Article could do better.

https://en.wikipedia.org/wiki/Small_and_medium-sized_enterpr...

Small to medium Enterprise (business)

I've come to believe there's a rule similar to Betteridge's Law of Headlines along the lines of: When an article says "small-business" it can be substituted with "big-business" to make the article more accurate.

Except in rare cases, policies that help smaller businesses will also help larger ones proportionally. To hide the big-business benefit, promoters of the bill will spotlight how it helps the "little guy". Because SME/SMB have become politically untouchable. Being accused of hurting small business owners will get a congressperson out of a job. It's a think-of-the-children argument. So when I see something being touted as good for small business business, it sets off warnings that a spin doctor is at work.

This is a problematic proposal, but, unfortunately perhaps, I do not think one can plausibly say that it raises the specter of "EU economic suicide". I would characterize it more along the lines of being a step in the refeudalization of the developed world.

Not sure what to make of this.

Patents should be much more expensive. They are government granted monopolies.

Most patents are also not really original too, so I'd expect the world would be better off with 1/10th the number of patents.

They probably should ongoing gradually escalating fees (for example maybe getting a patent for the first 5 years can be $100k, next 5 years $1M, next 5 years $10M, last 5 years $100M).

But the fees for resolving disputes should come down (mostly lawyers fees, not court fees). Probably the best way is to eliminate the government granted Monopoly on attorneys (and related restrictions on law).

All of these suggestions are a right step. The fundamental issue is that the current system is designed to issue when in doubt; without considering the cost to society from inhibiting progress by pointlessly throwing everything through the legal grinder with extra sand (in the form of a patent minefield).

Ideally we'd start looking for an entirely different way to encourage innovation, and just cease issuing patents entirely. For example, patent-like systems with a more restrictive scope (not exlusive access to pretty much "any idea", but exclusive access to e.g. "medical approval"), so we don't accidentally hit every future sector with this junk. And perhaps we'd look for some system that's more self-regulating; i.e. where abuse exists rather in the form of ripping of the taxpayer, and less in the form of court cases. The justice system is so terribly inefficient and everything it touches, yet is just keeps on growing until everything is some kind of zero-sum game.

I think this bias against small companies and individuals is easily explained, at least in the USA where I live. This probably applies to the EU also:

Large corporations buy influence via lobbying, revolving door policies, etc. so expect governments to set in place policies to protect large incumbents from smaller rivals. So many modern problems would be reduced if we could get rid of the deep systemic corruption.

Annoying that politicians feel entitled to set enormous fees for government services as a form of social engineering or profiteering, rather than simply providing the services at cost.

Court fee to invalidate a patent in Belgium is 165EUR, Netherlands 667EUR, France around 400EUR, Italy around 500EUR, Czech Republic 80EUR, Sweden 160EUR.

We are very far from those 20.000EUR.

Luckily the UPC is a complete pipe dream and may never see the light of day. It’s been ‘around the corner’ for nearly a decade.

They are now very close to get it running, they want to organise a signing ceremony of a patch for the Brexit issue, without any debate in the Parliament(s):

http://patentblog.kluweriplaw.com/2021/10/28/member-states-w...

If there's one thing I've learned from the world of think tanks it's that you have to be extremely critical when someone starts arguing about the woes of SMEs.

It's not clear from this press release that there's a fee to defend oneself. A 11k fee to file would surely make it less likely there would be trolls?

If someone attacks you with a bogus patent, then it would cost 20.000€ to request invalidation of the bogus patent.

From a quote near the bottom of the article:

> Typically, in response the sued company could be advised to demand cancellation of the Unitary Patent by means of a counterclaim but in that case, it will be required to pay a court fee of EUR 20,000.

11k is nothing to a patent troll like a FAANG (MAANG now?) but quite a lot to the kind of SME that - afaik - doesn’t engage in this behavior to the same extent.

> MAANG now?

I keep hearing MANGA

If we substitute Netflix for Microsoft (I’m never sure why Netflix was chosen instead), we can have GAMMA which is my personal favorite.

Why not have both? "The GAMMAN companies" has a nice ring to it.

I think it was originally for "Silicon Valley" companies, and Microsoft was up in Seattle.

So's Amazon.

It was "high-growth large-cap tech stocks" at a specific point (where MSFT was flat) - it was a Wall Street thing, not a tech thing.

> (MAANG now?) FAANG still has G for Google, not A for Alphabet. In a similar vein, we can (and perhaps should?) simply ignore Facebook's rebranding attempt - names are a social constrict, you don't get to unilaterally decide how others will call you.

>names are a social constrict,

typo as truth.

We can perhaps now just shorten it to "the man". I think at one point it was generally used to refer to exciting tech companies where people got paid really well and had an almost unreasonable standard of employment (meals provided for them etc).

For the most part those companies have grown up and stuff like free coffee and fruit has become the standard across most offices.

If only this were true! People wouldn't be kicked off open source projects.

It is being referred to as MANGA in other places

Trolls have a reasonably large amount of money to burn if necessary, and they normally try to get their victims to settle before things reach any court.

So no, this won't stop trolls, it will only deter a lot of legitimate cases from SMEs.

Reply on news.ycombinator.com