Webcash: Peer-to-peer electronic cash with no blockchain | Hacker News Reader

Webcash: Peer-to-peer electronic cash with no blockchain

webcash.org

24 pointskanzure15 comments

> The central server helps webcash wallets detect double-spending and ensure the integrity of the monetary supply according to the supply schedule.

How would it work if the central server eventually goes down? Would the payments be in a pending state until it goes back up?

The system mostly relies on the central server staying online. Technically, though, users make payments by copy/pasting webcash to each other, which can still happen even if the server is offline. When the server is offline they won't be able to check for double spending and they won't be able to tell the server to refresh any secrets.

So it requires gpu mining but doesn’t use blockchain and requires explicit servers. That’s confusing, is there an official rationale for it?

With a blockchain, a client downloads all the blocks and runs a local query to see if there has been any double spending of a coin (roughly). Here, a webcash wallet queries the webcash server for double spending protection instead. It asks the server, "has this webcash been spent already?". This is substantially faster, and it requires a different trust model from the regime of the trustless decentralized blockchain.

With regards to mining, actually right now CPU mining still works. When the FPGA and ASIC miners come, it would be (comparably) trivial to switch out the mining algorithm that the server recognizes.

If server trust is required, why not just put a database into it and store signed records like every bank does?

Yes, the server is a wrapper around a database. That's the point of this project. :)

That’s what this is.

Related: You can have a functioning and _secure_ decentralised blockchainless cryptocurrency through local quantum computation _without_ quantum communication: https://iohk.io/en/research/library/papers/one-shot-signatur...

(excerpt)

> We show that one-shot signatures have numerous applications for hybrid quantum/classical cryptographic tasks, where all communication is required to be classical, but local quantum operations are allowed. Applications include one-time signature tokens, quantum money with classical communication, decentralized blockchain-less cryptocurrency, signature schemes with unclonable secret keys, non-interactive certifiable min-entropy, and more. We thus position one-shot signatures as a powerful new building block for novel quantum cryptographic protocols.

> We define the notion of one-shot signatures, which are signatures where any secret key can be used to sign only a single message, and then self-destructs. While such signatures are of course impossible classically, we construct one-shot signatures using quantum no-cloning.

Well, okay. I don't know enough about quantum cryptography to really evaluate this idea. However, conventionally, the way to do this is a trusted execution environment that attests to the hash of the software running and it promises to have really really deleted the private key or the secret after it is transferred. Advantage of this scheme is that it does not require me to study more cryptography.

'Decentralized' ... 'central server' ???

Also, if the creator is reading here, can you give us a reason for why you decided to keep Blockchains' arguably most controversial aspect - PoW - after removing the blockchain? There are lots of non-PoW cryptocurrencies, surely you could have found another distribution model?

Not the creator, but PoW has plenty of positive qualities. The fact that the distribution schedule is much more accelerated than the others, and doesn't rely on fees should make concerns about "energy waste" less pressing.

Two thoughts:

Security: The centralized server is an open invitation to hackers.

Cash out: Man does not live by web cash alone. The often overlooked issue with crypto in general which adversely affects it's utility is how to convert back to fiat and at what cost. If this can't be done in a simple, seamless, low cost manner, I'm out.

> Security: The centralized server is an open invitation to hackers.

There are many centralized web services that operate around the world with high security. There's a large array of conventional techniques for securing servers. Admittedly most people and companies are bad at this.

There are many centralized web services that operate around the world with high security.

I know. I have run one for over 20 years. My solution --- build a unique, custom, single purpose web server software that only supports a very strict set of functionality after every single request has been thoroughly vetted.

I see the attacks mounted against my server on a daily basis --- mostly focused around canned, off the shelf server software and the endless flaws and shortcomings in either the software itself or how it is configured.

A server dealing with money will likely attract increased attention and sophistication.

Admittedly most people and companies are bad at this.

My point exactly.

"Peer-to-peer"

"Centralized server"

Pick one.

Reply on news.ycombinator.com