[testesttest]

Have you tried storing the list lately? I don't think you have a 10TB plus flash drive.

https://etherscan.io/chartsync/chainarchive

Crypto is more and more centralised because the hardware to run your own node is now thousands of dollars. That is without even getting into miners...It is kinda failing.

[yokem55]

This chart is kind of dated and misleading at the moment.

1) You don't need an archive node to fully verify the current state of the chain. A full geth node in full sync mode (about 600gb) will process every transaction of every block and get you the current state. You only need an archive node if you want the history of each account and contract separately indexed.

2) Folks that do want an archive node can now use much more space efficient clients then geth such as erigon (1.4tb) or besu (1.2 tb) if they want to run a block explorer or query the state of an account in the past.

[testesttest]

1) you need an archive node to get historical events which is the only way to do lots of things. E.g. scrape all NFTs

2) Erigon has horrible bugs and is missing data. I filled a ticket against one such issue and they "fixed" it by making the exception go away and returning null. Their lack of attention to detail made me abandon the project.

If you know someone in the project I can produce MANY examples and would be willing to do so because running an archive geth node sucks.

[yokem55]

I don't know any of them directly. But they do have an email on their github readme that you can send a request to for an invite to their discord. And if they don't respond on that, you can reach many of them in the Eth R&D discord.

[influxmoment]

That's Ethereum. Largely was consider a bad ideas because it's too large to independently validate

[testesttest]

Name a main stream one that is easy? I run archive nodes for SOL, ETH, Polygon, others.

[mhluongo]

Bitcoin.

[MBCook]

You’re right. 324 GB is far more manageable. Their point is totally invalid.

[influxmoment]

Agreed. 324GB and this level of growth is manageable and can be done with a raspberry pi and an ssd