I found

> the patent will expire and the technology behind it will be free to be used by any e-commerce site

particularly specious. There is no "technology" behind it. It was a concept or busines process that the vaguaries of patent law deemed worthy of rewarding.

Not to mention that any technology ("source code") Amazon used to implement the patent will remain under copyright essentially indefinitely.

Are you just quibbling about the semantics used in the article?

Or are you saying nobody else can implement "1 click" even after the patent expires... because of copyright?

Edit: Was an honest question, I've upvoted all the answers as I found them interesting, not understanding the barrage of downvotes shrug

They're saying that, unlike usual patents, the content of the one-click patent filing does not serve as a useful basis for anyone else's implementation of a one-click shopping cart. This is the whole point of patents: to publicize a trade-secret process in exchange for protection. If your patent can't be used by your competitors to (make it easier to) start doing the thing you're doing, then you didn't really file a patent.

The one-click patent has an implementation, but the implementation is so trivial that there are only two things you can do with it: either copy it wholesale, or write your own based on the same idea (which will have nothing in common with the original because they're both too small to "share DNA" like that.) In this case, you can't copy wholesale, because copyright. So you're stuck writing from scratch. At that point, the patent has given you no advantage in implementation over the simple knowledge that a process for doing X exists at all.

Exactly. If the teachings in the patent aren't necessary or even useful to future developers, then what was the point of granting a 20-year monopoly in the first place? No research expenses needed to be recouped, and there was never any danger that some arcane knowledge would be locked away forever as a trade secret.

Many patents don't reward the first person who solves a given problem, so much as they reward the first person to encounter a given problem. That person adopts a trivial, obvious solution and the USPTO cheerfully rubber-stamps it at everyone else's expense. This was clearly one of those patents. It cost Amazon essentially nothing to research, nothing to develop, and nothing to implement, but the rewards were immense.

> If your patent can't be used by your competitors to (make it easier to) start doing the thing you're doing, then you didn't really file a patent.

This is a good way of describing the point of patents, thanks :-)

This is the most insightful description of patents I've ever seen. Thank you.

> At that point, the patent has given you no advantage in implementation over the simple knowledge that a process for doing X exists at all.

This depends on whether you consider X to be "converting ecommerce sales" or "ordering items with a single click". Just because the implementation is easy to reverse engineer from a one line description of the novelty doesn't diminish the resources spent by Amazon in coming up with this insight and testing how it improved their business.

I'm not strongly in favour of software patents in general, but the argument that this patent is "too easy to copy" misses the point a bit.

I think the comment is both quibbling about semantics (the article used the term "technology" overly broadly because the value of the one-click idea isn't in the implementation but really just in the concept) but I think also voices a legitimate complaint.

The same failure of understanding that caused the author of that blog post to misidentify Amazon's one click patent as being related to a "technology" rather than just an idea is also behind the patent office's troubling tendency to issue patents to other ideas. Advocates for patent reform make a convincing case that issuing patent protection to ideas like this actually harms innovation more than it fosters it.

The fact that the actual implementation (technology) of the one click shopping idea, the source code, is protected by copyright is true but isn't particularly relevant. And to answer your second question, copyright prevents anyone from using Amazon's particular one-click source code, but that's largely irrelevant, since it's an idea that can be implemented a million ways, which is a major reason that many people feel it shouldn't be eligible for patent protection.

I don't think it's a quibble. I think it's a point about what even makes sense to be patentable. The expiration of the patent frees up exactly zero technology.

To clarify, it's not that the special technology behind it doesn't get freed, it's that there is no special technology behind it.

Business process can be considered a technology, i.e. technical procedure. Also, since they got the patent, wouldn't they have to pass tests for previous works, novelty, and obviousness? And, why shouldn't something be patentable, just because it's easy to replicate?

Being easy to replicate often suggests that it is, in fact, _obvious_. Which is supposed to invalidate the patent.

Lots of things become easy to replicate after they've been done once, despite being non-obvious before. The "no obvious inventions" part of patent law is supposed to prevent people from patenting things that are already broadly used (or small variations on things that are already broadly used). The cliched example being that you can't patent the wheel.

Edit: I see that Animats has covered this much better elsewhere in the thread.

> Lots of things become easy to replicate after they've been done once, despite being non-obvious before

Maybe, but this is not why we have patents. Patents justification is to help innovations to spread, by making them public and, in exchange, granting a monopoly for some time as would have been the case should the innovation been kept secret.

Patents are not, in theory, a bonus for whoever implement something first.

In practice though, this seems to be largely the case, at least in software, that patents are like those mine field maps during the gold rushes, partitioning the hills and delimiting every tiny area belonging to one miner, probably to police what would otherwise be a violent, perpetual quarrel.

> Maybe, but this is not why we have patents.

Yeah, and? I wasn't saying so.

this is nonsense

Explain?

I used the Payment Request API in a project and it was great. Chrome currently has it implemented for Android, you call a JavaScript function, a prompt appears asking the user if they'd like to pay, they enter the security code and hit pay.

All this feature actually does it give you the users credit card, phone number, and billing/shipping addresses. You still have to give that information to someone like Stripe.

They're focusing on mobile since that is where the checkout experience is the most painful. https://ride.lyft.com has it implemented if you want to check it out.

> All this feature actually does it give you the users credit card, phone number, and billing/shipping addresses. You still have to give that information to someone like Stripe.

I used to sit on the working group for Payment Request - the primary intention of Payment Request is to get rid of checkout forms and parallel implementations of the same logic. Right now most for most implementations that means credit card entry, but the standard is designed to support any type of payment.

So if I'm running an eCommerce site and want to support PayPal, Apple Pay, and Android Pay I shouldn't need to implement three discrete SDKs and APIs - I can instead implement Payment Request (although you'll still need to process the payment information output by each of those payment mechanisms separately. Ultimately for many merchants the implementation of Payment Request will be abstracted away by their payment processor, such as Stripe).

However, there is a parallel standard to Payment Request, called Payment Handler, which is designed to allow third party payment providers to write service-worker-esque plugins to browsers to take payment directly. They're intended to work together, but it's possible for a browser to implement just Payment Request to replace autofill if they so choose.

There's also a lot of other work going on at the W3C around payment, but Payment Request and Payment Handler have the most support from browser vendors and are as such the most actively developed.

Do you know if this is being used in iOS Safari for Touch ID in the browser too? Or is that something non-standard?

Apple Pay on the web exists and it's fantastic. As Tarek points out it is custom and doesn't use this standard.

Apple has not implemented the starndard yet, the WebKit Status page lists it as "Under Consideration".

I hope they end up implementing it for the sites hat won't do Apple Pay natively.

That's the non-standard Apple Pay JavaScript API [1].

[1] https://developer.apple.com/documentation/applepayjs

Well, they are trying to undercut Chinese with web payments, one part why mobile payments took off so well in China is because Tencent and Alibaba took out browser makers out of the equation.

If you open a website that requests your Visa or MasterCard CC data in any Tencent product, it will be blocked unless "their special hack proof API" is used. And yes, they blocked apple pay too

> Well, they are trying to undercut Chinese with web payments

Alibaba are very active participants on the Payment Request Working Group. I'm not sure they can undercut themselves. You will find several very large Chinese companies taking part: https://www.w3.org/2004/01/pp-impl/83744/status

On my memory, they were very very anxious about anything with capacity to subplant zhifubao. Paypal famously had a one click payment plugin for China, and it got banned by Alibaba nearly instantaneously.

Paypal later banned Alibaba from using Paypal outside of China

Eons past, the HTTP status code 402 was reserved for future use; "Payment Required".

Please to god W3C, take 402 and make it part of that standard! I so wish for us all to be able to respond 402 and have a simple standard way to charge.

Payment Request is a JS / client-side API (it's all about getting payment information from the browser to the website), so HTTP codes don't really come into it.

However, there's a proposed server/HTTP specification that does include it. It lacks the same broad support that Payment Request has, so is still in quite an early stage of development - https://www.w3.org/TR/webpayments-http-api/

This is what GNU Taler [0] does. You can send additional headers[1] to decide how the payment will be processed, such as X-Taler-Contract-Url and X-Taler-Offer-Url

[1]: https://docs.taler.net/merchant/frontend/python/html/tutoria...

[0]: https://taler.net/en/index.html

  > madness

Why shouldn't something be patentable, just because it's easy to replicate?

Because if an "invention" is obvious it isn't patentable ( http://www.bitlaw.com/source/35usc/103.html ), and many many people see this as being far too obvious to be patentable.

"Obvious" for patentability purposes means obvious in advance, not obvious in hindsight. Before Amazon, nobody had one-click purchasing. They wanted people to sign up, to fill out forms with irrelevant info, and "engage with the customer". After Amazon, many people wanted one-click. That means it wasn't "obvious to one skilled in the art" in advance. For patentability purposes, the best way to demonstrate non-obviousness is to show that lots of people have worked on the problem and didn't get it right. The one-click patent meets that test.

Amazon's real innovation was combining one-click ordering with easy order cancellation for a few minutes after the order. This makes enabling one-click ordering low-risk. Many on-line sales operations still don't get this, as I mentioned the last time this came up on HN. Most shopping cart programs have a built-in attitude of "We have your money now, muhaaa haaa". Yes, you can "contact customer service to cancel." ("Please listen carefully as our menu options have changed... Due to unusually high call volume...")

Now the people who don't get this will implement one-click ordering without easy cancellation and anger their customers.

  > obvious in advance, not obvious in hindsight

Thanks! That's what I was looking for. Many things we take for granted, which are so simple and obvious, are not obvious at all when they don't exist... I remember exactly those experiences everywhere - you have to register to add an item to a shopping card... And yes, you MUST give us your mother's maiden name...

Additionally, the meaning of "obvious" is non-obvious. It means obvious to somebody with knowledge of the relevant field.

Which, in my estimation, makes this "invention" even MORE obvious.