6 ms·
Great article. I was actually responsible for invalidating another maps-related patent from the 1990s, in Germany. I was in Patent Litigation and I also rememb
by AlbertCory 5y ago
Great article.
I was actually responsible for invalidating another maps-related patent from the 1990s, in Germany. I was in Patent Litigation and I also remember having a "ART+COM" folder in my email, but that's about all I can remember. I don't think the case was very active by 2014.
Anyhow, there's something Avi said that I have to comment on:
The test for patent-worthiness is: “would it be non-obvious to someone who is skilled in the art?” I’d been doing “stupid programmer tricks” like this since high school, as have many others in my field.
If only it were so simple! You would think that "hey, that's just a stupid programmer trick!" would be enough to render something obvious, but you'd be wrong. That's what my paper [1] is about.
No, to call "obvious" the combination of references A and B, there usually has to be some "teaching, suggestion, or motivation" in the literature that it would be a good idea to combine A and B. There are nuances to this statement, in case anyone's feeling pedantic.
[1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2399580 https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2399580
- neatze 5y agoTo me context of software patents; is like patenting a how to use hummer, and not novel innovation of hummer design. More simpler analogy can be applied to wheels, imagine patent where wheels cannot be turned left on any vehicle, because of patent. In context of wheel, I have no issue if someone invents and patents better rubber type and production technology, but patenting how rubber generally can be used on the wheel is simply wrong, it has not benefit to society/humanity.
- aidenn0 5y agoIf someone were to patent an analogue circuit, would the software implementation of the same circuit on a DSP be patentable?
- orangepurple 5y agoHold my beer for a minute while I patent the idea of reproducing analogue circuits with a software implementation of the same circuit on a DSP.
- Gigachad 5y agoI personally think patents should only apply to significant investments. Anything that could be mostly explained in an elevator trip should not be patentable. Rounded corners should not be patentable. The latest ARM core design should be.
- touisteur 5y agoBut what if what you invent is deeply simple, after you've worked out everything and spent years simplifying? What if it is 10 lines of code but it solves a big industry-wise prod or exploitation problem? The concept might be simple but might necessitate years of effort to find and validate? After reading constant praise about simplicity here on HN it always sound like cognitive dissonance to then read when patents are mentioned that something 'simple' should not be protected.
- ootsootsoots 5y agoPatents, trademark, and copyright pretty clearly serve to empower the aristocrats. A good idea will win on its merits or fail. Legal ownership of ideas just tether us to paying Disney and Google’s of the world in perpetuity as they buy everything up. We basically live in a state sanctioned system of quota and expropriation of effort. It’s not violent like the USSR, so we look beyond it. And we spend a lot of energy policing it all.
- seb1204 5y agoI can offer this link to a podcast about this exact subject. Unfortunately it's in German DE https://twitter.com/timpritlove/status/1450041042930122754?t=3cDg1INGsvlLH3-wjPsa4g&s=19 https://twitter.com/timpritlove/status/1450041042930122754?t... But Tim is surely up to clarify any questions.
- jandrewrogers 5y agoI think there are two different cases that often look similar in practice. The first is when the invention is obvious to everyone but no effort is spent on development because it would have no value until adjacent technologies reach a certain level of maturity. The second is when the adjacent technologies have been mature for some time but there is a critical problem that needs a key insight to bring it all together. To an outsider these cases often look identical, the difference is that the former case was obvious in foresight and the latter case has a tendency to look obvious only in hindsight, but only a reasonably objective subject matter expert can make this distinction. The latter case is a litmus test of non-obviousness -- all the pieces were there for some time but no one had put them together. It is distinct from when everyone puts the pieces together the second the pieces are available. Patent clerks can't tell the difference. I will say that the state-of-the-art for geospatial computing in the 1990s was pretty damn primitive, many useful things had not been invented yet. It isn't great today but it has come a long way.
- dboreham 5y agoBased on my experience working as a consultant to patent litigators, it's worse than that. I dove pretty deep over a few beers with these guys on this subject. What they told me is that in the US patent system, you can pretty much forget about using obviousness as a defense. Yes it exists as a concept but if you intend to use it, you'll probably lose so they never try to use it. Hence the focus on prior art.
- jandrewrogers 5y agoObviousness is not obvious. It requires an incredibly deep level of subject matter expertise to even make that determination.
- touisteur 5y agoAnd. Some solutions that seem very obvious in retrospect can be the result of years of shaking down kinks, letting expert intuition crystalize, until something almost 'dumb' fixes into an invention. Obviousness is the wrong caliper here, I think.
- avibarzeev 5y agoYeah, I was trying to keep it simple, but you convinced me to edit that line out for correctness. Thanks! Their patent was invalidated in 2017. Here's why: https://www.ipwatchdog.com/2017/10/25/cafc-affirms-invalidity-geographic-map-visualization-patent-asserted-against-google-earth/id=89554/ https://www.ipwatchdog.com/2017/10/25/cafc-affirms-invalidit...
- AlbertCory 5y agoMy paper was an attempt to cast our "native common sense as practitioners of the art" into a legal framework that already exists. If a technique is "obvious to try" then you're not entitled to a patent for trying it. You can laugh at "native common sense" as being meaningless when the law is involved, but actually it's not. In the right circumstances, the fact that most practitioners of the software art see things a certain way would be "dispositive" (as the lawyers like to put it). We haven't seen anyone create those circumstances. I had a senior attorney review this paper for me, and he had a lot of interest in the concept. He called it a "toolkit" and found it intriguing that software engineers had a toolkit they applied to any problem. You know that we do. The forces arrayed against this concept are formidable. Even Computer Science departments in universities aren't unified against software patents. I actually talked to some profs at Illinois, and one of them said "you're assuming that I agree with you!" He didn't. He had a startup company and the VCs had told him he had to have patents.
- 1vuio0pswjnm7 5y ago"The test for patent-worthiness is ..." Novelty is also a requirement, in addition to nonobviousness. The OP suggests there are more important issues for programers to focus on: "I'm openly critical of their ad-driven business model, especially since the damage is clearer. That's broached in the show, but only to point out how much money Google would theoretically owe these guys if they'd agreed or been forced to license at 10 cents per use (which was very unlikely IMO). Invading people's privacy for profit is not something to yearn for. It's one of the reasons I sold all my Google stock. I respect Google's engineering work and the great "free" information they provide. There are a lot of smart and dedicated people working there. But the biz model is not ok, IMO." Sad that Google appropriated this guy's work for a "business model" that he does not approve of. Perhaps there are many more folks like him, who have done solid, valuable work in some area that has nothing to do with invading people's privacy or online ads, only to have the product of that work usurped and given away for "free" in the interests of furthering Google's prime directive: selling online ad services.
- lbriner 5y agoNo-one has mentioned the Egg of Columbus yet: https://en.wikipedia.org/wiki/Egg_of_Columbus https://en.wikipedia.org/wiki/Egg_of_Columbus A tale of something that is "very obvious" only after it has been done. I mean MP3 is about reducing information for frequencies hidden by other frequencies right (I think that is roughly correct!) and that is a really obvious way to reduce bandwidth....once someone worked it out.
- jasonkester 5y agothere usually has to be some "teaching, suggestion, or motivation" in the literature that it would be a good idea to combine A and B There's the rub. Because patents get issued for things that are so obvious that you would never bother mentioning them in "the literature". Take dragging a map with the mouse as an example. First, if you're doing online mapping, the first thing you think of is to just grab the map and drag it wit the mouse, so you implement that. (Even though some clown has a patent for the act of doing so). But assuming you've gotten away with that somehow, now imagine the first day of implementing it, when you "drop" the map while your mouse is still moving. It'll stop in a jarring way and your first thought will be "That's ugly. What if it kept going with momentum and coasted to a stop." And that would get you sued out of existence. Now, the patent system seems to suggest that the appropriate way to handle this would be to search the patent database for "dragging a map with a mouse", find the patent for "momentum", license that from whoever has it, read it and learn how to implement it, thus benefitting from some novel invention, saving yourself time and effort, and reimbursing the inventor for his effort. But in actuality, you spend the time it takes to draw a breath thinking about the problem, then code it up in the next few minutes and carry on with your day. Because there are very few things in programming where describing the problem doesn't also describe how to solve it. I've never seen an example of a software patent that covers an actual non-obvious "invention".
- michaelt 5y ago> First, if you're doing online mapping, the first thing you think of is to just grab the map and drag it wit the mouse When online mapping was first around the javascript slippy map hadn't been invented. The page would show you a given map tile, and provide left/right/up/down/zoom buttons that would load a tile one step in the direction you chose. Sometimes with a complete pageload for every step. For example: https://web.archive.org/web/20040805053745/http://getamap.ordnancesurvey.co.uk/frames.htm https://web.archive.org/web/20040805053745/http://getamap.or...
- aardvark179 5y agoWhile that may be true there were definitely various slippy maps in GIS and CAD applications. There were also other interesting navigation schemes that relied on multi button mice or tablets. I still like the model where you hold down a button and the navigation speed and direction is controlled by the distance and angle from that click point to the mouse cursor.
- pacman2 5y ago"The test for patent-worthiness is: “would it be non-obvious to someone who is skilled in the art?” " Well, nothing is harder to say than if something is "obvious". Locking backward, basically most things are "obvious". "someone who is skilled in the art?" This is another questions. What is "skilled"? Most patent courts assume an average person from the field of the invention and not an extraordinary specialist.
- pmarreck 5y agoInteresting paper! I've also noticed that many insights also seem "obvious" in hindsight (the proverbial "egg of Columbus" thing). In fact, the better the insight, the more "obvious" it seems!! Thus, "hindsight obviousness" is not an argument against an innovation