weak subjectivity works

[zack]

At DEVCON1 today in London, a guy named Jeff Coleman gave a talk explaining how we can use IPFS to make weak subjectivity secure. He calls his technique "universal hash time".

If someone has a block from 4 months ago in a proof of stake chain, then they should be able to prove that the block was available before other information that became available 3.8 months ago.
We can examine the merkel tree to know that order that information became available.

[psztorc]

Zack, what percentage of people who "give a talk" about Bitcoin / Blockchain tech, do you think actually know anything about what they are talking about?

I would say the number is near-zero, if it isn't actually zero. Even the Core devs, or myself, could have simply have "competence without comprehension", be fooled-by-randomness, etc.

That isn't to say that we should ignore everything new, and write it off as wrong, but anyone who speaks at an Ethereum conference is probably trying to appeal to gullible people / has fooled himself.

[zack]

I also give zero authority to people who talk at these conferences. I have talked at one, I have watched a lot of talks about very bad ideas. The reason I am excited about this guy's idea is because I know the properties of merkel trees. He uses them in a creative way. If you understand merkel trees, you don't have to trust his authority. You can understand his proof.

I summarized his proof:

If someone has a block from 4 months ago in a proof of stake chain, then they should be able to prove that the block was available before other information that became available 3.8 months ago.
We can examine the merkel tree to know that order that information became available.