You are viewing a single comment's thread from:

RE: The best protocol level solution to the Tron takeover impasse: DO NOTHING

in #steem4 years ago

In the longterm, we should reassess witness voting such that one vest equals one vote and it's harder for a few entities to change the protocol.

I agree. Multiplying voting power seems absurd.

What do you think about the idea of increasing the number of "top witnesses" to something like 200 instead of the current 20?

I believe this would make it more difficult to implement changes in the future, but that could be considered a "positive" as it would have a stabilizing effect and would make things much harder for any future attempt at a hostile take-over.

Sort:  

What do you think about the idea of increasing the number of "top witnesses" to something like 200 instead of the current 20?

A larger number of witnesses seems more decentralized. The Steem/EOS/BitShares adopt a middle ground of decentralization. Centralized enough to maintain high capacity and efficiency (i.e. only requiring hundreds rather than thousands of nodes). Decentralized enough to prevent censorship and ensure immutability. I'm not sure what number of witnesses ensures the best balance.

It would be good to get feedback from someone who knows a bit more about running a witness node to know whether a higher number of witnesses would have any negative technical consequences or require a larger block interval.

That sounds reasonable. Even 40 would be superior to 20. My first thought was that The Sun would probably just DDOS the top 20 and do whatever it wants while they're flailing.

If a government wanted to force a take-down of the steem blockchain for whatever reason, killing 20 servers would seem like child's play.

If a government wanted to force a take-down of the bitcoin blockchain for whatever reason, they'd really have a difficult time and they'd probably have to invest a ton of cash into developing their own highly-specialized server farms.