Re: [Bitcoin-development] Two Proposed BIPs - Bluetooth Communication and bitcoin: URI Scheme Improvements

[Mike Hearn <mike@plan99.net>]

[-- Attachment #1: Type: text/plain, Size: 1729 bytes --]

BLE meets a different use case than regular Bluetooth. BLE is designed to
allow always-on broadcast "beacons" which are conceptually similar to NFC
tags, with very low power requirements. The tradeoff for this ultra-low
power consumption and always on nature is the same as with NFC tags: you
get very little space for data, and they are essentially one way. That's
why a common use case for it is to trigger some other mechanism like a
classical Bluetooth socket or HTTPS connection.

I think BLE has a role to play in Bitcoin payments, but probably not for
actually transferring payment data. Rather, a merchant should be able to
drop a BLE beacon in their shop, and then wallet apps can use that to learn
where to download a payment request/upload a payment message. But the
actual data transfer would still take place over Bluetooth, Wifi or the
internet.

That leads to the question of what the beacon broadcasts. A bitcoin URI is
the obvious answer: the problem is a URI contains an address. No problem
for the "throw money at a live performer" use case but a problem for the
cafe use case. If we are willing to mandate BIP70 and remove the static
address part from the URI the we get a "uri that points to a url" which is
a bit inefficient but at least lets us distinguish bitcoin beacons from
other kinds. That still leaves the fundamental question raised by the
Airbitz spec - how does your wallet download the right payment request?

Unfortunately that's a tough UI problem. I don't think comparing long hex
strings manually is a good way to go. This seems less user friendly than a
QR code.

Once we solve that problem, how BLE beacons can trigger payments will all
fall into place. The tech part isn't the hard part.

[-- Attachment #2: Type: text/html, Size: 2084 bytes --]