Thanks for calling me on this... it shows I didn't do a good enough job of explaining my key point.
The point I was trying to make was that the architecture of card processing has a fundamental weakness: the consumer authorises the merchant to generate and send a message to the consumer's card issuer (via the card network) to take funds from their account However, ensuring that the amount you authorised matches the amount they request is decidedly non-trivial. In the absence of controls, the amount they ask for could be anything. This means that every time you use your card you are open to the risk that the merchant asks for far more than you were expecting.
To mitigate this risk, the edifice of card security grew up over time.... in the paper days, the fact that both the merchant and retailer had an identical copy of the voucher allowed the consumer to dispute a fraudulent transaction. More recently, technologies such as chip+pin try to do something similar. But they're all really just sticking plasters on the funamental problem: card processing relies on the merchant "pulling" the funds from the consumer's card issuer.
Now sure... it works (usually) - but the cost and complexity of making it work safely is large.
The point I was trying to make was that Bitcoin (like cash, as you say) works differently. The key communication link is from the consumer to the merchant (i.e. a "push") and the consumer has full control over how much they send.
Now this isn't perfect... you have to solve the problem of how the consumer knows *where* to send the money and all your other objections are valid (e.g. consumer has to be online, etc, etc)
But as an opening statement, I think this characterisation could be a very effective way of destabilising/undermining any representative from the card industry who one suspects of planning to spread FUD.
Agreed... I wasn't really addressing the pain of key management, etc... this was really just a consequence of the push versus pull observation
Hopefully my comments above helped explain what I was trying to convey - but the point I was making was that when you use a card you aren't actually "paying" the merchant and you aren't transferring funds to them. You are *authorising* the merchant to pull. And a huge amount of work is needed to ensure that what they actually pull from your account is what you authorised them to pull!
I was too broad... I meant "no need to trust the merchant to pull only the funds you authorised them to pull"
I think it probably *is* fair :-) The "authorise merchant to pull and put in place controls to make sure they only pull what they said they would" is the only way you could build the system in the 60s. They have moved in terms of architecture since then but this fundamental model is still at the heart of how cards work... and my argument is that it's not an ideal way to do things.
Sorry - that wasn't my intention.
Bitcoin is almost laughably immature in so many respects... but by using "push" rather than "pull" as its core funds movement philosophy, I would argue it's on the right side of history