Hacker News Reader

Comment by wizzwizz4 | Hacker News Reader

Alright, I'll bite.

Consider a “decentralised” blockchain system that does the work of The Pirate Bay – providing hashes of particular files, along with their names, so that people can select and request data from the P2P network.

How, exactly, does this system help with RSS in any way?

Additionally, how does this system help with the “censorship” issues that The Pirate Bay has? Even when The Pirate Bay remains up, that doesn't mean anybody's seeding.

And a blockchain-based system has to have this limitation. If all the data is directly on the blockchain, nobody can actually download it all, so the blockchain becomes unverifiable and ceases to have any cryptographic properties at all. (In short, there's no benefit to a blockchain at all.)

View on HN

    How, exactly, does this system
    help with RSS in any way?

You cannot make a Twitter-2-RSS bridge because Twitter block access to the data. In a decentralized system, access to the data is inherent.

    Even when The Pirate Bay remains up,
    that doesn't mean anybody's seeding

That is where the financial incentives come into play. Seeding will be economically incentivised.

    If all the data is directly on the
    blockchain, nobody can actually
    download it all

Blockchains have come a long way since Satoshis version in 2009. We have approaches now where participants only need to download arbitrarely small parts of the chain. Yet the network as a whole keeps the chain intact.

> In a decentralized system, access to the data is inherent.

Well, yeah – but we already have that. It's called the Fediverse. No blockchain in sight. (In fact, it predates Bitcoin.)

> That is where the financial incentives come into play. Seeding will be economically incentivised.

The margin for any given piece of content will be low, so there's an economic incentive to honour takedown requests, no matter how fraudulent or unenforceable, if the effort to check whether they're valid is lower than the expected average reward minus the expected average legal cost.

> We have approaches now where participants only need to download arbitrarely small parts of the chain. Yet the network as a whole keeps the chain intact.

We had those back in Satoshi's time, too. They work by having n centralised trusted nodes, then asking all your peers to double-check what you got from those trusted nodes. That approach is, surprisingly enough, also vulnerable to censorship!

(Also, I don't really think your proposed scheme is good enough to have Satoshi's name associated with it. The innovation of Bitcoin was the coin melting pot thing, not the blockchain; Satoshi's original blockchain, which Bitcoin still mostly uses, was just a worse version of Git repositories.)

    The margin for any given piece of content will
    be low

Not when the financial incentive rises the rarer (less shared) a piece of content gets. This will be baked into the protocols. Arweave already has this.

That's not feasible either. Where does the money come from?

Where does the money to do Bitcoin mining come from? From the users who want to write to the Bitcoin blockchain.

Same with the blockchains that support larger amounts of data. Whoever wants to store data pays for it.

The specifics of how the storage of rare content is incentivized depends on the protocol.

On Filecoin the publisher pays an ongoing fee.

On Arweave it is paid upfront. The current estimate is that $10 will store a GB for 200 years.

So doing a tweet on an Arweave powered social network would cost you something like $0.001 if you want your tweet to survive for 200 years.

I think you fundamentally misunderstand Bitcoin. People don't pay to store data on the blockchain; they pay to get their transactions mined. There's no guarantee that all the blocks of the Bitcoin blockchain will stick around forever, only that there'll be a consensus as to what the current block is, and what the states of all the wallets are (since that's the only information you need to validate a new block).

Keeping historical blocks around is not a property of Bitcoin. No up-front blockchain payment system can keep data around for any length of time. (There is no mechanism enforcing the storage of the data; such a mechanism cannot exist within a computer system.) “$10/GB for 200 years” are the words of a snake-oil salesperson – or, more charitably, somebody who hasn't properly read the marketing copy.

You are confusing function and implementation. Technically "get their transactions mined" is having a piece of data written to the chain.

You are correct that in Bitcoin there is less incentive to store old blocks. But in chains crafted to store data for a long time there is.

I am stopping this discussion here because you are too negative (almost aggressive) for my liking.

But it will be an interesting thread to read again in 20 years when all this has played out.

I don't want to be a naysayer but how would content moderation in regards to spam and illicit things work in this scendario?

Spam cannot be dealt with. Illicit material sticking around is a feature. There is an economic incentive for material that creates a liability for the “seeders” to not be kept around.

To give it as much credit as I can: individual peers could do their own content moderation, deciding what they do and don't want to be around, and if enough people decide something shouldn't be there, it'll basically drop out of availability. But that goes against the whole “market forces” philosophy that it's supposedly based on, and also goes against some of the claims that the proponents of these systems keep making.

It's just a bad system, not even taking into account the pointless CPU cycle wastage. Blockchains sort of make sense for something where each block is based on the previous, but for arbitrary pieces of data? Chaining them like that is faddish unless it brings real concrete benefits over just doing the same thing without blockchain.

Reply on news.ycombinator.com