EFF’s Stupid Patent of the Month shows how patent reviews stop dumb patents

A patent litigation factory was stopped from suing hundreds of small printers.

Read the whole story

 

[loquacio]

I'm hugely overworked, leading to large backlogs, with no end in sight. Do I:

1) say "no" to stupid applications with obvious prior art and no advancement of the state of the art and have to spend a lot of time documenting, arguing, and then seeing the re-application?

OR

2) say "yes" and everyone is happy? The backlog shrinks, the applicant is pleased, etc...

I wonder how much this occurs in real life at the patent office...

Yeah, if 1 was obvious, it would happen, but too often, it’s not only not obvious, but totally unknown, so the patent goes through.

There’s really nothing wrong with stupid patents. I’ve seen plenty of those over the years. But stupid patent almost always means valid patents that patent something stupid that no one will ever want to do, so no one cares.

Stupidly patenting something that shouldn’t be allowed to be patented is something else.

Well, the problem with #1 is who is the one judging if it is obvious or not. Some patent office clerk at the turn of the millennium may not see some network printing tech as obvious, while someone who works with networks would throw the application away wondering why someone wanted to patent something so stupidly obvious, like the sharing of resources in the network, which is what networks are for. On the other hand someone can come up with some patent that I would read and think "this is some breakthrough shit" only to have a person well versed in the subject come up and tell me I'm an idiot for allowing that to be patented when there's prior art going back to when token ring networks were still a thing.

Office clerks do the same everywhere, they file things and retrieve them. Office clerks don’t make any of these decisions, as you should well know. Patent examiners make them. They are experts in the field they examine. But there are so many patents out there, and so much work being done, with hundreds of papers in many different journals filed every year in every area, that it’s impossible to keep up.

You can’t properly blame the examiners. Every one of us makes mistakes. Anyone who claims otherwise is either naive or narcissistic.

To understand the concepts of patents, you need to understand that it’s not what the patent claim is about, it’s how the claim does it. You can have 10 valid patents on ways of sharing resources over a network. There’s no exclusivity in many things.

You can take a valid patent and add to it in a way that makes possible something that the original patent can’t do, and patent the combination as a new patent.

The patent system is versatile. But errors creep in. Modern technology is extraordinary complex. Some patents take dozens of pages to explain.

Exactly. People tend not to realize that the only thing in a patent that really matters is the broadest claim, in which every single word limits the scope of the patent. When it comes to articles like this, simply reproducing the patent's title will often give the impression that a patent is far broader or more abstract than it actually is, which leads to misconceptions about the scope of the problem. Patent trolls frequently have terrible patents, if you actually read the words of the claim. Most patent troll lawsuits have little if any merit. The patent trolls aren't expecting to go to court and win - rather they're counting on the defendant not wanting to deal with the hassle of litigation and opting instead to pay a settlement or license fee.

People seem to think that at least some patent examiners simply rubber stamp every patent application that comes across their desk. In reality, almost all patent applications are initially rejected by the patent office. It's practically a matter of course for patent examiners to reject an application the first time they see it. Most patents are only allowed after a multi-year back-and-forth between the patent office and patent lawyer, during which the patent lawyer will add language to the claims to significantly narrow the scope of the patent until they arrive at something the examiner is comfortable with. A patent with the title "Networking System" might end up with a broadest claim that fills an entire page, and which no one could infringe even if they tried.

Now, that's not saying that every patent that gets issued is perfect. At the end of the day, patent examiners are people who have to trawl through massive databases looking for prior art that may or may not exist. Their searches are only as good as the keywords they come up with, and sometimes prior art gets missed. While they have to meet certain technical qualifications to review patents in their field, they may or may not put in the effort to keep up with the state of the art. Fortunately we have the IPR system to catch patents with prior art problems, even in the absence of rock-solid guidance on the section 101 issues that often come up when we talk about software patents.

Automatic initial rejection then approval when sufficient bullshit is added is not a functioning system.

Patent Office doesn't get a free pass because they're "just doing their job" and "it's hard". Their job is a cancer on society. Framing it in legalistic and procedural bullshit doesn't change that.