Re: [Bitcoin-development] Paper Currency

[Gregory Maxwell <gmaxwell@gmail.com>]

On Sat, May 17, 2014 at 9:07 AM, Chris Pacia <ctpacia@gmail.com> wrote:
> I can't really just hand someone the note and walk away
> because they have to scan it to see if it is actually valid.

Not just scan it, but they actually must successfully sweep it—
otherwise they can be trivially double spent. This is especially bad
since any prior bearer can perform such an attack. E.g. record the
private key of everyone that passes through your hands and then
doublespend race any redemption that happens >24 hours after you spend
them. The wrong person would likely be blamed and even if you were
blamed you could plausibly deny it ("Must have been the guy that gave
it to me!").

Othercoin seems to have much better properties in the space of offline
transactions: https://bitcointalk.org/index.php?topic=319146.0

Separately, Cassius also ran into some regulatory issues selling
physical bitcoin artifacts. Especially printing things that seem to be
redeemable for a named USD value sounds especially problematic.

Some random comments— The base58 encoding is fairly human unfriendly.
It's fine for something being copy and pasted, but I've found typing
or reading it works poorly due to mixed case.  I expect the A/B side
to be difficult to educate users about. "This side is private" is more
easily understood, you could just pick one of your sides and call it
private.  I find it kind of odd that this design seems to have no
facility for checking its txouts without recovering the private key,
though considering no one should rely on such a measurement without
sweeping perhaps thats for the best.

(As far as the numbering goes, I think you should be calling these
draft-felix-paper-currency  etc. As a matter of hygienic practice I
will not assign a matching bip number for something that went public
with a number outside of the assignment.)