Hector Martin

Yes there's downsides to just increasing blocksize which is why it's not been done and other options are being looked at. In the meantime it *is* an electronic cash. With challenges, but it's there. Not a failure (yet). Or stupid.

Cash doesn't have a $30 transaction fee. Right *now*, all Bitcoin is is the world's most expensive checking account modality, and a vehicle for speculators.

You're declaring it a failure whilst it's in trouble and before it's failed (with fixes in pipeline). You're declaring it stupid for not having magical properties no-one knows how to achieve. Current problems are temporary. But you can still use it.

Incidentally, can you reference that O(n3) scaling thing, I've not heard that one before and it doesn't seem obvious to me.

There are N people in the network. As N grows, each individual person also has more potential people to transact with - transactions would be O(N²). Since each person has a copy of all transaction history, the network requires O(N³) storage overall.

Number of transactions is number of people squared? Nah. That implies that the newborn will transact with every person on the planet. Let N be number of transactions. Number of full nodes is constant. Therefore O(n) storage.

O(n²) does not mean n². Yes, there is a (small) constant factor because not everyone transacts with everyone. If you prefer, make the exponent somewhere between 1 and 2 if you expect sub-linear growth with user base. How is the number of full nodes constant?

Also, choose some arbitrary constant for the number of people a person will transact with in their lifetimes - not every entity in the world. n^2 is different than a large(?) constant times n.