[DennisP]

That's why they're working on scaling.

Available now but still working on adoption/UI: rollups, which essentially compress the data on chain. Those bring transaction capacity from several dozen to a couple thousand tx/sec.

In a year or two, data sharding, which multiplies the capacity of rollups. That takes throughput to 20K-100K tx/sec.

With that in place, quadratic scaling kicks in, where hardware improvements multiply both the capacity of shards and the number of shards, so e.g. a 3X hardware improvement means a 9X capacity improvement.

[prepend]

They’ve been working on scaling for six years. You know any tech platforms that took six years to scale to the point of usability?

If they cared about scaling rather than milking greater fools, I think it would be solved.

Imagine if someone presented a database or OS that worked so inpractically and they said they were still working on scaling after six years. That would be the dumbest startup ever.

[TekMol]

    You know any tech platforms that took
    six years to scale to the point of usability?

The internet comes to mind. It started in 1970 with ARPANET. And we still have scaling issues :)

[prepend]

Arpanet was really useful early on and got more useful as they added nodes and protocols.

[DennisP]

As I mentioned above, rollups are available today. In six years it's gone from 7 tx/sec (Bitcoin's max on chain) to a couple thousand tx/sec, improving by almost three orders of magnitude. If it continues at that rate, it'll be around a million tx/sec six years from now.

The roadmap I described above actually does get to that level if we assume a 3X hardware and bandwidth improvement over that period.