Would an EOSIO-based blockchain with a single Block Producer (BP) active and producing blocks at any given time work?
Of course it would.
Think about it: why would it not work?
Would it not work for technical reasons, such as vulnerability to DDoS attacks?
Nope. A single BP node can be shielded by any number of public nodes, just like “centralized” application servers today are shielded by an arbitrary number of “session” web servers who actually deal with client connections.
But wouldn’t the single BP operator cheat the game state on purpose and just print unlimited virtual money for itself?
Nope. All transactions are still broadcast to everyone that is also running a full node. Anyone can reconstruct the system state in real-time and verify that all rules are being followed. The output from the BP’s public nodes can be verified against the output locally computed by full nodes, which can expose their own APIs for others to compare results vs. the BP.
Any effort to run a single-BP DPoS blockchain by a group of sufficiently serious and known people will very much probably be forever safe as far as intentional cheating by the original group in charge of the BP is concerned.
But wouldn’t the single BP operator censor transactions?
Probably not. Again, why would them? Not in theory, but in practice?
But maybe that chain wouldn’t work for geopolitical reasons, such as legal or even physical targeting in the single state where it is located (i.e. vulnerability to SWAT raid)?
If we are to approach this from reality, we can see that concern is mostly paranoia. In practice that would never happen if the single producer hosts the BP in any state with reasonable laws. In the event it does happen, then the EOSIO BP election system simply promotes the first most-voted backup BP to the top position. The only difference from the EOS mainnet would be that instead of 21 active BPs and 200-ish backup BPs, it would have a single active BP and all the other BP candidates would be back-up replicas.
But maybe the most interesting question is: “Why?”
Why would one start a single-node EOSIO-based DPoS blockchain? For two main reasons.
The first one is the “Honeypot” angle: it would be just massively interesting and educational to see if an EOSIO blockchain with a single BP would work exactly the same as the main EOSIO blockchain for years to come (it totally would). Logic dictates that the “Honeypot,” single-BP blockchain would be compromised far, far earlier than the EOS mainnet. Thus, the continued existence and correct functioning of a Honeypot EOSIO blockchain with one BP would provide a continuous reminder that betting against the EOS mainnet quorum of 21 BPs is pretty stupid.
The second is simply because it is something that a single, small group of people, or even a single high-skilled person with lots of free time can do. The EOS mainnet bootstrapping process, where hundreds of IT startups got together to run for a 21-spot BP election, will hardly be replicated. It’s too expensive. Blockchain projects start small, and starting an EOSIO main-chain is much easier done with a single active BP and zero backup BPs (for starters) than trying to conjure a diverse set of social actors to support a single business.
The first “Honeypot” EOSIO chain to appear will demonstrate that single-BP DPoS chains are possible and that they can be trusted, if the single BP can be reasonably trusted, both socially and technically. Its existence will work as a mental handrail that people can use to gain a better grasp of how secure DPoS and replication really are (e.g. all other things being equal, security increases exponentially with the quorum size–and humans are notoriously bad at understanding exponential functions). And thus it will support others wanting to start their own small-quorum DPoS blockchains.
So there’s the base idea for an EOSIO fork. If you want to seriously attempt it, let’s chat. I might be interested in helping you do it.