Show HN: A new blockchain that can do email on-chain
As a long-standing member of this community (2007!), I'm really pleased to announce the launch of the Saito project here.
Since you probably haven't heard of us, Saito is a terabyte-level blockchain designed to support email, social networks, payment channels, distributed forums, and other big-data applications directly on-chain. We achieve scale for this by solving the fundamental economic problems with the proof-of-work and proof-of-stake consensus mechanisms. If you look into the network design, we think we'll blow your mind!
Anyway, we've been reading the discussions here for the past several months about whether blockchains are useful and... well... hopefully Saito is going to end that debate! So if you're interested in blockchain technology or just open-minded, I really hope you'll check us out and let me know if you have any feedback or suggestions or questions. Right now we have an online demo, downloadable code on Github, and some resources explaining how Saito works and how to get started building apps with it.
Also, I should probably also mention that we're hiring, with the caveat that we're in Asia so anyone interested in joining the team should either be in this part of the world already or happy to work remotely.
14 comments
[ 3.2 ms ] story [ 42.0 ms ] threadYes -- Saito is the first "transient blockchain" and everything will technically fall off the network eventually unless someone acts to preserve it.
It is possible to maintain data on-chain, although in most cases the users sending transactions and the users receiving transactions will not want to pay for that and will prefer bear the burden of archiving them. There is no reason for the entire network to backup my email inbox, for instance. So "permanent storage" will only really be needed for tokens that represent monetary value.
We do have two ways to maintain data on-chain in perpetuity though:
1. applications can manually rebroadcast transactions whenever they near the end of the chain. This is dirt-cheap and helps preserve the security of the network for complicated reasons that you can read up about in our explanation of how Proof-of-Transactions / Proof-of-Relay works.
2. nodes creating new blocks are forced to rebroadcast transactions that meet consensus/code-level criteria qualifying for rebroadcasting (i.e. their "rebroadcast" flag is set and they have enough money to pay for it. We intend this as a fallback rather than the main mechanism -- it will be more expensive.
Generally though, shifting to a transient chain is important as it is one of the steps needed to "fix" the economics of existing consensus mechanisms. It allows us to price the cost of storing a transaction forever by dividing infinity up into manageable chunks during each of which storage costs are accurate and market-driven and fees are handed over to the nodes that are actually doing the work in that point at time.
The thing about government is that no one, not even government can be trusted. Government however, has power. For obvious reasons, no one can govern the government, so all we're left with is improving visibility, heavily.
Absolute power corrupts, and without oversight, they basically have absolute power right now.
The reason the demo defaults to plain http is that the server that feeds out the applications is part of the full-node software and it does not support HTTPS yet. We are intent on fixing this, but it didn't seem like a critical thing to worry about for this dev release, which is focused on getting people a working version that can be run on localhost and give people the tools to build genuinely distributed applications. We hope to get this fixed in the next two months before launching our public testnet and DNS system. Right our biggest dev focus is overhauling the network code for multiclient connections.
With that said, in the long-run it won't matter if users http or https to connect to the blockchain. Saito is not vulnerable to MITM attacks and users can use Diffie-Hellman key exchange mechanisms to swap keys directly over the blockchain. We can think of it as an improved version of TCP/IP that is actually secure but that will cost a few fractions of a cent for every message we need to send to unknown and anonymous peers.
There is actually -- see comment blelow -- but if you're concerned about SSL just go through our reverse proxy that adds it. Some links may break:
https://demo.saito.tech
We should have default SSH in the full-node client itself by the time we launch our public testnet in a few months.