[dozzie]

>> Blockchain does not solve consensus problem. It's a timestamping service.

> Certainly Bitcoin seems to be a distributed consensus system.

It is not. Bitcoin only tells which of the pair of documents was earlier (a document describes a single transaction). It doesn't allow to negotiate a value. If you swapped producing the proof of work with sending the transaction to a trusted third party for adding a timestamp, Bitcoin would work just as well. Go read "trusted timestamping" topic, Wikipedia has a good description.

>> Why the hell would anybody think "proof of work" is a "solution to denial-of-service attacks"?

> Isn't it?

What service does blockchain protect against availability disruption? How?

[acjohnson55]

Still not sure what you're saying. What do you mean by "negotiate"? Because there are certainly rules for resolving conflicts. As long as we're using Wikipedia, the article for consensus in computer science explicitly mentions Bitcoin.

Proof of work can be used to protect against denial of service. If you provide a web service, you can choose to require proof of work before serving a client to make the client commit some proportional amount of effort.

[dozzie]

> Still not sure what you're saying. What do you mean by "negotiate"?

OK, then what do you mean by "consensus"?

> As long as we're using Wikipedia, the article for consensus in computer science explicitly mentions Bitcoin.

Yes, and wrongly, too. Especially that the specific problem of establishing consensus that the Wikipedia page talks about has, since its very first paper, the impossibility proof at 1/3 of the participants, compared to 1/2 for Bitcoin (which means that Bitcoin is something different than consensus with Byzantine faults).

[acjohnson55]

> The consensus problem requires agreement among a number of processes (or agents) for a single data value. Some of the processes (agents) may fail or be unreliable in other ways, so consensus protocols must be fault tolerant or resilient. The processes must somehow put forth their candidate values, communicate with one another, and agree on a single consensus value.

This seems to me to be exactly what Bitcoin does, where the "value" is the sum of all facts asserted on the blockchain.

[dozzie]

Wikipedia page is about a very concrete problem described in a specific paper. Have you read the paper?

Then, Bitcoin does not allow to choose "the sum of all facts asserted on the blockchain", not any more than a typical DHT or a gossip protocol in a peer-to-peer network. In fact, there is nothing resembling the end state as in "we now terminate the protocol with this value as the outcome", which is required by definition of any type of "consensus" term.

Also, have you tried, as I said two comments ago, swapping sealing transactions using proof-of-work with a trusted third party that provides unique timestamps for the transactions? Have you read what timestamping is?

[acjohnson55]

What paper are you even talking about? You might even be right, but you are a poor communicator.

[dozzie]

Well, I sort of assumed that since you voice your opinion about consensus as defined in computer science, you at least should know what it is. I talk about Leslie Lamport's (plus Shostak and Pease) original publication about Byzantine generals problem, from 1982.