[Cryptography] Does Bitcoin have unaddressed design flaws?

[jrzx]

> The idea there might be latent unaddressed issues lurking in the core designs of Bitcoin really got me thinking, and I'm wondering how much of Bitcoin depends on Satoshi's early decisions which really haven't gotten much scrutiny since.

Satoshi's decisions have received enormous scrutiny, and many problems have arisen that were foreseen, and many problems that no one foresaw.

Many people who don't know much about cryptography or crypto currency are trying to spread fear, uncertainty, and doubt. The paper on ed25519 is just silly.

Unfortunately, bitcoin is limited to ten transactions per second or so, which is inherent in the basic architecture of the blockchain, and you cannot overthrow the world financial system with ten transactions per second.

A correct implementation for a lightning layer has been proposed, which solves all my criticisms of the existing so called lightning layer.

"Anonymous Multi-Hop Locks" which however is very abstract and leaves out a lot of practical details, such as implementation of their anonymous channel

What is needed to overthrow the existing financial system is a blockchain designed to handle six thousand transaction a second, and a lightning network that works as in the above proposal.

The proposed lightning network requires homomorphic one way functions, and unfortunately curve25519 is not quite homomorphic. Ristretto25519 is however homomorphic, so we need two rather radical transitions.

If implemented on ed25519 signatures, yes, there will likely be holes in the security. Though it is possible to implement the proposed lightning network on curve25519, that is not something that I would try.

But with that lightning network, and a blockdag architecture rather than a blockchain architecture, crypto currency is likely to eat the world financial system. It is already making a serious dent despite the ten transactions per second limit.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.metzdowd.com/pipermail/cryptography/attachments/20210206/0f6c2e88/attachment.htm>