Hacker News new | comments | ask | show | jobs | submit login
How to Run a Blockchain on a Deserted Island with Pen and Paper (hackernoon.com)
70 points by jxub 9 months ago | hide | past | web | favorite | 19 comments

What are we trying to achieve? It’s very simple actually — all we’re trying to do is maintain a simple table of balances on a piece of paper. This table will show how many coins each of our heroes has. The trick is, because we can’t have one piece of paper that holds the only source of truth — we’re going to have to keep things equal and let each of the gang maintain their own version — this is the decentralized part. And naturally, we’re also going to hope that all 4 pieces of paper eventually show the same thing — this is the consensus part.

Nothing in the statement of the problem requires a block chain.

Maintaining distributed ledgers is not a difficult problem except for one niggling detail: double spending. In other words, if the scenario you're looking at has no risk of double spending, a block chain isn't going to help you at all.

To expand on the article's example...

Jack is down to his last coin and gets a brilliant idea. He'll pay Kate and Hugo each one coin during the day. If he succeeds, Jack can forge one coin out of thin air.

The islanders face a dilemma. Which of Jack's two payments should be accepted? Accepting the payment to Kate means that Hugo loses his. Accepting the payment to Hugo means that Kate loses hers.

Accepting both payments is a non-starter. The islanders may as well just start bartering again if monetary scarcity can't be preserved.

This is the motivating example behind block chains. The system we know as Bitcoin had been largely worked out by the time Satoshi released the white paper in 2008. Digital signatures and hash chains mean that forgery of coins or payments is practically impossible.

What was missing was a way to make double spending difficult, without invoking a central authority.

Excellent reply. One could expect HN to be less susceptible to this fad that is blockchain.

To be fair, the author never said it was the only or best solution, but articles similar to this could make use of a note mentioning that there is no need for blockchain and there are more efficient ways to do this.

HN is susceptible for two simple reasons: money and ideology. Many here have stakes in cryptocurrency and until they can find new bag holders they need enthusiasm. In this space enthusiasm and popularity are directly translated to profits. The second is that there is a small yet vocal minority of cryptolibertarians, you know ones who think “fiat is dead!” who desperately want cryptocurrency to not only thrive, but triumph.

The consensus here is trivial to control by the strongest of the group. What if when Hugo sent the 200 coins, he said “oh we’re allowed to generate 200 coins if it has been over 3 days of no coin generation” and his statement would be impossible to verify because:

1- They have no way to get in touch with the central authority, the developer (author of the blog post), who can tell the users if the new consensus rule is allowed or not.

2- They have no written account of the consensus rules, which would help to at least invalidate new consensus rules that seem to conflict with the written ones.

So what would happen if Hugo did this? Well if he’s the strongest of the bunch, he could simply impose on the group and say “this is a new consensus rule, accept it or else”. So much for decentralized consensus!

> Well if he’s the strongest of the bunch, he could simply impose on the group and say “this is a new consensus rule, accept it or else”. So much for decentralized consensus!

Isn't this always the case? If Hugo is strong enough, he can also bully his way into abandoning the blockhain, and relying on his word for what everyone's balance is.

(Of course, this kills any notion of an actual economy, and becomes a totalitarian regime powered by effectively slave labor...)

> Isn’t this always the case?

Well yea, that’s how societies work. Social hierarchies emerge as we delegate power to the strong and wise. If they abuse their power, there is strife and potentially war, and eventually the corrupt leader gets replaced with a new one.

The idea that Bitcoin or any given blockchain is “more decentralized” than what we currently have is profoundly flawed. All you have to do to control a blockchain is control its consensus rules, as the Hugo example demonstrates. All you have to do to control the consensus rules is control the developers... and developers don’t operate in some parallel universe outside of our governments’ control.

All that to say, yes indeed, Hugo runs the show here, blockchain or not... just like he runs the show in the real world.

Well the article does explicitly say:

> Is this an ideal blockchain implementation? Probably not. It can be improved in many ways.

To your point, though, couldn't the rules just be part of the block chain and documented as a "transaction"? Any change in the rules would just be another transaction and require majority approval.

I hate Hugo

This is great. Planet Money mentioned how it was/could be used in prison: https://www.npr.org/sections/money/2017/02/10/514577243/epis...

Inefficient, mostly not new either because in the past you'd have one person with that responsibility: the one who has experience with ledger. Now everyone's a notary. How it'd happen with centralized/specialization is another person would cut wood, yet another would hunt fish, and everyone would do their part (you can see this in effect in the movie The Beach [1] though the group is larger than 4 like in this example; also in Cast Away its just that there's only 1 person there who has to do everything by himself which makes it a redundant example). Everyone specializes in something because that is more efficient when HR are scarce. Because the group is small its easy to keep track who's doing what; instead of pet peeves like ledger.

So in short either you're so bored that you don't have anything better to do, or you're busy with surviving. If you're bored and you happen to have pen and paper I can recommend something different: PnP RPG. Such relies on something you'll always have with you even when you're alone (see Cast Away [2] as well): your imagination.

[1] https://www.imdb.com/title/tt0163978/

[2] https://www.imdb.com/title/tt0162222/

For anyone interested, there is literally no evidence that any society anywhere, ever, switched to using money because using barter was so onerous. [0] On an island like this one the inhabitants would inevitably create a system based on mutual aid and ongoing negotiation of responsibilities to one another. While this is a fun article, to me it exposes how cryptocurrencies are primarily useful as a value store and a means of speculation rather than a solution to actual human-scale problems.

[0] From "Debt: The First 5000 Years," by David Graeber https://www.mhpbooks.com/books/debt/

Love it. Added the IslandCoin White Paper [1] to the Bits & Press Bookshelf (along with the Bitcoin and Ethereum White Paper).

[1] https://github.com/bitsblocks/islandcoin-whitepaper

SpaldingCoin is the hottest ICO in the archipelago

If there are four of you on a tiny island and you're not running an informally communist society, something's wrong.

Possibly what's wrong is that one of you is a libertarian.

I'm no purely being flippant here - four people are going to be so inter-reliant and form such a small, tigh-knit group, that keeping track of balances and debts is likely to cause more problems than it could possibly solve.

This is a situation for trust, not currency.

I agree. You're more like a family at that point than four LLCs.

Then again, I wonder how many people reading that blog post and nodding along are already sketching out a brilliant idea to bring this back home, and have their spouse and kids get on the handwritten blockchain wagon.

How do you decide that unlimited trust is better than properly accounted currency?

This is great. We just signed a contract to take ownership of an entire remote tropical island to create an eco hacker space.

I’m inspired for us to implement something based on this as a class.

The island is here > http://Majagual.org

Would love to find other cool ideas for projects and inspiration.

There is nothing eco about taking over an island to "create [a] hacker space"

I appreciate your perspective.

Our first priority is to protect the space and be great stewards of the land. We also believe it is possible for humans to live integrated with nature and for the space to have a positive impact.

Current projects include:

Plastics clean-up and processing. Can the plastics that wash up on shore be used for anything useful?

Working to protect the native plants and animals -- there is currently a pretty sad practice where locals will try to set part of the island on fire to flush out iguanas and then shoot them. They like the pregnant ones the best. This is detrimental to the space as well as the iguanas -- we will work to establish some sort of sustainable relationship so this does not occur.

Development of electric fishing boats that can re-charged via solar.

We have a relationship with the MIT FabLab in Panama City and hope to bring students to the island for immersive experiences as well.

We will be hosting very light leave no trace events as well as building very light eco-infrastructure such as rainwater collection.

Additionally, the space will serve as a platform for inspiration, healing and leadership development. The intention is that by continuing to work with folks in this type of environment they will be exposed to new ways of living and will return with a deeper connection to themselves and nature.

We believe we can work closely with the natural environment and contribute to the increased health of the land and surrounding communities.

We haven't taken outside investment so there is no mandate to generate a financial return. Our plan is to move slowly and with intention. We will also welcome insights from folks like you to help keep us accountable.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact