4 ms·

Though I understand your sentiment, please note that the protocol governance problem holds for most major currencies (i.e. Bitcoin, Ethereum, etc). So it is not

by tinco 9y ago

Though I understand your sentiment, please note that the protocol governance problem holds for most major currencies (i.e. Bitcoin, Ethereum, etc). So it is not a unique flaw to Blockstack. Please also note that though if you think about it from a centralized perspective (i.e. you using your regular browser downloading Blockstack from their own website) then yes, it is centralized. But if you think about it from a community perspective, since all code and protocol is open source, it is trivial to establish new protocol and code regardless of authority. So the essence of it is decentralized. And this is not just some theoretical thing, all of the major coin protocols have gone through community driven ('decentralized') forks that went directly against whoever was leading the projects at that time. For example Bitcoin<->Litecoin (and notable BCC recently), Ethereum<->Ethereum 'Classic' and Ripple<->Stellar.

tribby 9y ago
> Ripple<->Stellar stellar is not a ripple fork.[0] this is a common misconception. 0. https://twitter.com/JedMcCaleb/status/933358161783693312 https://twitter.com/JedMcCaleb/status/933358161783693312
- tinco 9y ago
  Thanks for the clarification!
- andruby 9y ago
  It started as one though, IIRC, and later they wrote a new consensus protocol
TomMarius 9y ago
While you're technically true, it's practically impossible to manipulate Bitcoin in this way.
madeofpalk 9y ago
That's a weird definition of 'decentralised' then. That's just like saying that apt-get is decentralised because its open source and you get point it to different servers.
- ensignavenger 9y ago
  If somebody attacks and takes down the central server that your apt-get is hosted on, your won't be able to use it (without changing your servers). If somebody attacks the central server of the token devs, the existing token network continues to operate, even if you can't get an "official" update from the original location. So it isn't really the same thing.
- tinco 9y ago
  I am making a point about the governance of the protocol, not the protocol itself. The apt protocol is centralized, but distributed to attain performance at scale. The governance of the apt protocol could be seen as centralized, because Debian tells you which packages are ok, but it is also easy for individuals to switch to other governors, such as Ubuntu. This might be stretching the definition of 'decentralized', but my point is just that it's not a bad thing or even undesirable.
- hiram112 9y ago
  This has been a central problem for decentralized or any 'P2P' protocol since day one. When the more popular music sharing sites like Napster and Morpheus were taken down due to their reliance upon 'hub servers', more distributed networks like Gnutella attempted to get rid of their need for points of failure. But even they needed a concept of a bootstrap index to obtain a list of hubs from which to get on the network. And sure enough, it turned into a game of whack-a-mole between the industry and these servers. Torrenting has the same problem. If I remember right, the only protocols that work in the long run are those that can broadcast and discover each other, but they're limited to local LANs. Once you need to go outside of your own network, you need some way to discover where nodes are located, and without an initial point of reference(s) (which becomes the point of failure), you're screwed. Same problem with trusting users - either you use a centralized authority or pre-trusted root certs, or you have to know in advance the other users (e.g. physically exchanging PGP certs).
  - pault 9y ago
    Isn't whack-a-mole good enough though? How many examples do we have through history of a large centralized organization being bled to death (or at least harassed to the point of giving up) by trying to wipe out a shifting coalition of resistance groups with no central authority? Rome vs Jewish rebels/Celtic tribes, Imperial China vs Mongolian warlords, first France, then the USA vs Vietnamese rebels (even China took a shot but I don't know if that counts since it was post-unification), the USSR vs Afghani warlords, the USA vs Afghani warlords, any future empire that tries to occupy Afghanistan... I would bet dollars to doughnuts that every centralized authority in the history of human civilization has lost a game of whack-a-mole at least once.

by tinco 9y ago

tribby 9y ago

> Ripple<->Stellar stellar is not a ripple fork.[0] this is a common misconception. 0. https://twitter.com/JedMcCaleb/status/933358161783693312 https://twitter.com/JedMcCaleb/status/933358161783693312

tinco 9y ago
Thanks for the clarification!
andruby 9y ago
It started as one though, IIRC, and later they wrote a new consensus protocol

TomMarius 9y ago

While you're technically true, it's practically impossible to manipulate Bitcoin in this way.

madeofpalk 9y ago

That's a weird definition of 'decentralised' then. That's just like saying that apt-get is decentralised because its open source and you get point it to different servers.

ensignavenger 9y ago
If somebody attacks and takes down the central server that your apt-get is hosted on, your won't be able to use it (without changing your servers). If somebody attacks the central server of the token devs, the existing token network continues to operate, even if you can't get an "official" update from the original location. So it isn't really the same thing.
tinco 9y ago
I am making a point about the governance of the protocol, not the protocol itself. The apt protocol is centralized, but distributed to attain performance at scale. The governance of the apt protocol could be seen as centralized, because Debian tells you which packages are ok, but it is also easy for individuals to switch to other governors, such as Ubuntu. This might be stretching the definition of 'decentralized', but my point is just that it's not a bad thing or even undesirable.
hiram112 9y ago
This has been a central problem for decentralized or any 'P2P' protocol since day one. When the more popular music sharing sites like Napster and Morpheus were taken down due to their reliance upon 'hub servers', more distributed networks like Gnutella attempted to get rid of their need for points of failure. But even they needed a concept of a bootstrap index to obtain a list of hubs from which to get on the network. And sure enough, it turned into a game of whack-a-mole between the industry and these servers. Torrenting has the same problem. If I remember right, the only protocols that work in the long run are those that can broadcast and discover each other, but they're limited to local LANs. Once you need to go outside of your own network, you need some way to discover where nodes are located, and without an initial point of reference(s) (which becomes the point of failure), you're screwed. Same problem with trusting users - either you use a centralized authority or pre-trusted root certs, or you have to know in advance the other users (e.g. physically exchanging PGP certs).
- pault 9y ago
  Isn't whack-a-mole good enough though? How many examples do we have through history of a large centralized organization being bled to death (or at least harassed to the point of giving up) by trying to wipe out a shifting coalition of resistance groups with no central authority? Rome vs Jewish rebels/Celtic tribes, Imperial China vs Mongolian warlords, first France, then the USA vs Vietnamese rebels (even China took a shot but I don't know if that counts since it was post-unification), the USSR vs Afghani warlords, the USA vs Afghani warlords, any future empire that tries to occupy Afghanistan... I would bet dollars to doughnuts that every centralized authority in the history of human civilization has lost a game of whack-a-mole at least once.