[Cryptography] ghash.io hits 50% of the Bitcoin compute power

[Jon Callas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

> I'm running Bitcoin Core 0.9.1 on my laptop, which I spin up every
> once in a while to update my copy of the blockchain.  I have no idea
> what it would do if it started getting lots of messages saying whoa,
> there was a fork 100,000 blocks ago and here's the new version of the
> blockchain.

It would do the same thing that every other piece of software that has to do complex things and error detection on hard-to-simulate, won't-happen-in-the-real-world conditions does: fail massively in some spectacular way. Usually the spectacular way includes *not* throwing an error. 

I find it hard to think that the only people who screw up chain-checking are people who deal with X.509. No one *ever* gets that right in their early software, and then when it's finally right, it suffers bitrot years later.

	Jon


-----BEGIN PGP SIGNATURE-----
Version: PGP Universal 3.2.0 (Build 1672)
Charset: us-ascii

wj8DBQFTnLn1sTedWZOD3gYRApM7AJ9g9VFAabjfxqnZe1zPb8cXIWhniQCfc9Z0
jklxNkLcFbnH+EdRQNSqrGM=
=qlyH
-----END PGP SIGNATURE-----