Thread by @hus_qy

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk I am a DLT researcher so when I try to determine the "scalability" of a particular consensus, then I am usually not that much interested in the current performance of the system under the current conditions but I ask questions like:

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk "How does the system perform when you increase the amount of nodes?" or "How does the system perform when you increase the amount of validators?"

So when I talk about scalability, I talk about "big O notation" kind of things rather than actual TPS numbers.

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk NANO works fine if the amount of validators is not too large (currently you have 4-5 validators that have > 50% of the weight).

If you are fine with this number of validators, then you don't even need to be worried about any of the things I wrote but as soon as the amount of ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... validators gets bigger the system gets significant problems. Validating a multi signature of i.e. 200 validators will take roughly 200 times longer than validating the signature of a single validator which severely limits the scalability of the system.

The fact that you ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... need to do this signature validation for every single transaction makes this problem even worse.

So if you assume that NANO strives to become somewhat decentralized and reach the amount of decentralization that other projects have today, then you will not be able to ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... process a lot of TPS.

Maybe you don't care about the amount of validators because you assume that they are honest anyway but having such a small amount of validators also has quite a few security related problems.

In NANO the validators need to be reachable on the ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... internet to be able to answer to requests by the issuers of transactions. This means that they can also be attacked (i.e. DDOSed) which in theory makes it really easy to just take down the whole NANO network with a relatively small cost).

The fact that the validators do ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... not earn any fees also makes it quite unlikely that they will invest hundreds of thousands of dollars into VAC style firewalls that could deal with these kind of attacks.

A lot of the mentioned problems do not exist for other consensus mechanisms because the validators ...

@tofeeornottofee @sattiota @Lazylefty1 @mira_hurley @Jackalope1357 @PeterSchiff @Eljaboom @elonmusk ... either don't need to have a publicly reachable IP address or there are simply more of them so the liveness of the system is not challenged if a few of them are attacked.