[cryptography] [Cryptography] BitCoin bug reported

Lucky Green shamrock at cypherpunks.to
Sun Feb 16 21:40:55 EST 2014


On 2014-02-16 14:41, ianG wrote:
[...]

> But lack of transactional integrity is a far greater sin.

Those of us that were around in the days of yore may remember how Mondex
failed to implement the simple concept of a two-phase commit. Why would
they incur the overhead hit on transaction times?

No reason to unless your database is actively working to defraud you.
And why would your DB do such a thing? (Especially if you aren't the one
operating the DB?)

I know IanG has been around for long enough to remember the chip card
payment systems that would first credit the payee and only afterwards
debit the payor. The latter part being preventable if you cut the power
to the card before the write. Which was real easy, since a write takes
far more mA than a read.

Often, I think of some of the mid 1990's payment system innovators as
incompetent. Which they were. Yet they were rocket scientists eligible
for Nobel Price compared to some of the Bitcoin outfits that I have met
over the last two years.

Never will I forget sitting down (as a favor to a friend) with a Bitcoin
online wallet outfit crew that is holding an absurd percentage of
Bitcoin bits, who took offense at the very notion that there might be
something to be learned from several millenniums of financial services
best practices.

I say, let them and all of Bitcoin burn to dust.

--Lucky


More information about the cryptography mailing list