@f_pieper @matthuwiler IOTA will support atomic composability of smart contracts that are close to each other in the sharding space.
Distant smart contracts are not composable but smart contracts are just pieces of code and you can deploy multiple instances of them in different shards.
I think ...
Distant smart contracts are not composable but smart contracts are just pieces of code and you can deploy multiple instances of them in different shards.
I think ...
@f_pieper @matthuwiler ... that the idea to have a single smart contract that serves the whole world is pretty weird (especially since it has a maximum amount of TPS that it can process).
What happens if the throughput requirement is larger than this capacity? Then the only way to scale further ...
What happens if the throughput requirement is larger than this capacity? Then the only way to scale further ...
@f_pieper @matthuwiler ... is to "fan out" and run multiple instances of them in parallel.
So why would you not go with this model of "parallelizing tasks" from the start? It is much more natural and the equivalent of how humanity has scaled since thousands of years.
It is also much cheaper from ...
So why would you not go with this model of "parallelizing tasks" from the start? It is much more natural and the equivalent of how humanity has scaled since thousands of years.
It is also much cheaper from ...
@f_pieper @matthuwiler ... the perspective of the user because not everybody is competing for the same resources.
I do think that 2nd layer scaling solutions like rollups are bad because they completely destroy composability, so I agree that we need L1 scaling but I think that distributing access ...
I do think that 2nd layer scaling solutions like rollups are bad because they completely destroy composability, so I agree that we need L1 scaling but I think that distributing access ...
@f_pieper @matthuwiler ... to popular smart contracts over parallel instances is better than centralizing it.
In addition, Cerberus uses a classical BFT style consensus which means that it will be pretty easy to attack the validators with things like DDOS attack.
In that context such a ...
In addition, Cerberus uses a classical BFT style consensus which means that it will be pretty easy to attack the validators with things like DDOS attack.
In that context such a ...
@f_pieper @matthuwiler ... "centralization" around a single instance of each SC is even worse because you can essentially take down all dependent SC's just by attacking a single shard.
@f_pieper @matthuwiler Maybe I am biased because I dislike the BFT style consensus algorithms due to their fragility (small amount of validators + low byzantine threshold).
We will see if Radix will ultimately be "robust enough" to survive in the long run or if my concerns are justified.
We will see if Radix will ultimately be "robust enough" to survive in the long run or if my concerns are justified.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
