40 ms·
4 or 5 long essays about a "simple solution" and yet I have almost no idea of what the solution might look like. Is it really that difficult to describe the use
by pbreit 14y ago
4 or 5 long essays about a "simple solution" and yet I have almost no idea of what the solution might look like. Is it really that difficult to describe the user experience in a paragraph or 2?
- lisper 14y agoI can describe it in one sentence: replace the existing protocol with one where payment authorization is done with secure digital signatures.
- pbreit 14y agoThat's really not helpful. What is the user experience at the checkout counter? Are there any new instruments or hardware involved? How are such things deployed to consumers and merchants?
- lisper 14y agoTroll. You already asked me that, and I already answered it.
- pbreit 14y agoThen please humor me and either point me to such an answer or repeat it here. When I walk into a store and pay with your service, what is the experience? Your answer need not mention digital signatures (unless you are referring to the actual signing on some sort of touch screen as you sometimes see these days; and in which case i know your solution is a non-starter for deployment reasons) or encryption. And enough with the "troll" stuff. I am not remotely such. 16 very successful years in the payments industry.
- lisper 14y agohttp://news.ycombinator.com/item?id=5296345 http://news.ycombinator.com/item?id=5296345
- makomk 14y agoThat'd do nothing to stop the $100,000 fraud he mentions, which involved Moneygram processing payments for fraudsters that tricked people into sending money - the secure digital signatures would check out fine, since the account holder authorized the payment.
- lisper 14y agoActually, the right kind of digital signature scheme would help prevent this kind of fraud (though, of course, you can never completely protect people from their own stupidity). A signature would be valid only for a transfer from a particular account to a particular account. The only way a moneygram-style fraud could be perpetrated is if the fraudster persuaded you to sign a transfer to an account that they control. In fact, with a proper modern PKE-based financial infrastructure, there would be no need for a business like Moneygram, which is one of the reasons that they will fight tooth-and-nail to prevent such a system from being established.
- makomk 14y agoYou'll still need someone to sign the public keys in order to prove that the person is who they say they are, otherwise it'd allow all the kinds of fraud that happened in the Moneygram case. (We've already seen a lot of this with Bitcoin for examples.)
- lisper 14y agoNo, you don't. It would certainly be helpful, but it's not necessary. You could, for example, get the public key directly from the recipient. If you want to send money to Uncle Joe, then Uncle Joe can write his public key on a piece of paper (or, more realistically, print it on a business card as a QR code) and hand it directly to you in a face-to-face interaction, no third party signer required. Indeed, no signer at all required in that case. But you do raise a valid point, which is that identity binding in general is a hard problem. But there are solutions (trusted third parties, web of trust). Like I keep saying, this is a long story.