It’s been six years now since I discovered Bitcoin, and yet I’m still amazed at the potency of the technology. Bitcoin has achieved something absolutely incredible, and it’s something that I think is lost on the majority of the altcoin ecosystem.
Bitcoin is a financial system that operates in the absence of trust. There’s no governing authority, the closest person to a figurehead has been gone for years, and a two year protocol deadlock is living proof that unwanted changes can’t be forced upon the network. You aren’t going to end up with a situation where the developers double the circulating coin supply in a single month to fund themselves, because nobody — not developers, not miners, not exchanges, not regulators — has the ability to introduce unwanted changes to the protocol.
Sia’s core goal is to bring that same level of decentralization to data storage. We want anyone to be able to store their data such that the uploader fully and exclusively has control over the data. And we want that data to be protected against failures, such that no group has the ability to steal, censor, or deny access to data — not the miners, not the developers, not the government. Your data belongs to you, and Sia aims to be a cloud where you have full certainty that nobody else can change that.
We’ve seen that power is a slippery slope, and that once a person or group has power, they will almost never relinquish it voluntarily. We’ve seen this with governments, with corporations, and with altcoin developers. The Etherum foundation for example has taken full control of the protocols future, including the rights to make sweeping changes to things such as the consensus algorithm, the coin supply, and even chain rollbacks. Most devs in the altcoin ecosystem maintain the same amount of control, which means users of this ‘decentralized’ ecosystem typically end up placing absolute trust in the dev team.
This is not what we want for Sia. We’ve been very careful from day one to exclude ourselves from the network as much as possible. Sia launched in June 2015 with zero dependencies on any of our servers, arriving straight out of the gate as a fully decentralized peer to peer blockchain. And in every release we have made sure that the network would be able to function at full capacity even if we disappeared unexpectedly. As developers, we have been adamant: no centralization is ever permitted in any of the core software.
As developers, we have been adamant: no centralization is ever permitted in any of the core software.
This makes life harder. It means that hosts cannot automatically adjust their prices to match the value of the US dollar (you need to trust a centralized service to know the siacoin price) without a third party plugin or extension. It means that we can’t rely on our servers to help users speed up their initial blockchain download. It means we can’t survey the network and create a list of good or bad hosts for our users to depend on. It means that features our competitors can add with just a few lines of code often cannot be added to Sia at all. We have to think completely differently when programming. But in return, we get true decentralization. No cheats, no hacks, and no promises from the dev team that we’ll eventually relinquish power — we already don’t have power.
One of the biggest checks that we’ve put on ourselves is that we don’t have the ability to push updates to the network. When there’s an update, users have to manually choose to download it. It means that it can take months before a new release is being run by a majority of the network. It means any change we make has to be compatible with every version of the software released going many months (and eventually years) backwards. It means that if we push out a bad update, it may never receive majority adoption.
The result of our ethos is that Sia is a network which requires you to download the full blockchain to get started. It means you need to acquire cryptocurrency before you can offer or consume storage. It means that in general that setup is a lot harder than on other networks. But the payoff is real: Sia is a decentralized application, and its decentralized in ways that most altcoin users don’t even realize is possible.
This is our commitment to our network: a fixed inflation curve, encrypted-by-default data, no centralized service involved with any first-party software (third party software can invoke centralization if desired — but the core protocol must be independent and decentralized), and no future hardfork will ever depart from the core promises or assumptions of the network today, unless it was clear that the hardfork was a requirement to keep the network operational (something like the DAO hack would not qualify as existentially threatening).
And the even better part: we’ve set up the network already so that the users have the power to enforce this promise. You are not merely trusting us to keep this promise (unlike with most altcoins). We cannot force you to upgrade, we cannot force you to switch to a network you do not want, there is no auto-update feature that we can abuse to migrate the network to a new fork. And the code is fully open source — there is nothing stopping you from maintaining a previous version of the software if you decide you don’t like the new version.
(note: we still have a few hardforks planned. The next big one involves changing the difficulty adjustment algorithm, because the current one is broken and it’s going to be bad news if we get ASICs before fixing it. All other hardforks have similar intent: fixing broken things. And all hardforks will have lots of warning and plenty of time for the community to raise issues if they don’t like a proposed change).
We are committed to decentralization. We are committed to building reliable infrastructure where the trust factor has been entirely eliminated. And we have proven our commitment by building a network where we have as little power as possible. You should not need to trust that developers will do the right thing. True blockchains give users the ability to know nobody can harm the network in the absence of trust, and this is what we want for Sia.