Hacker News Reader

Comment by PufPufPuf | Hacker News Reader

I wouldn't be surprised if all that was just a made up jargon and this was a joke article. But again, it's about blockchain, so the line is thin.

View on HN

There's actually interesting technology being developed here in the areas of distributed computation and zero trust systems. The implementations of Zero Knowledge Proofs and ongoing work on ZK-SNARKS I personally find most fascinating.

There's a lot more to this ecosystem than just speculation. At it's core is a distributed world computer but all anyone knows about is money.exe because this stuff is immensely complex.

If you look into the researcher rather than paying attention to the soyjack youtube thumbnails you'll find the actual substance. Nobody is going do the work for you. Or you know, just write it all off with a snide joke because "crypto bad".

> At it's core is a distributed world computer but all anyone knows about is money.exe because this stuff is immensely complex.

Alternatively, because the only way to use the aforementioned distributed world computer is to engage with money.exe and buy more CoinTokens. Imagine all the kids out there who will be delighted to learn a pay-per-use code interpreter. "Hey mom, I need your credit card to cover the gas while I debug my smart contract."

But assuming you have the money to spend, it's a whole universe of possibilities! Just make sure to cash in before actually trying to use any of them.

The problem with this statement is in assuming that any of this is actually ready for the average user, like a minor with their parents credit card. It's really unfortunate that the space received all of the attention it did during the pandemic, as that only managed to bring in misaligned expectations fueled by grifters making impossible claims.

There are a number of planned upgrades on the roadmap[1], such as layer 2 blobs, that will eventually drive the cost per transaction closer to zero, however we're still a decade away from that being the case. In the meantime you can debug your smart contracts on a testnet for $0

[1] https://notes.ethereum.org/@domothy/roadmap

Layer 2 blobs aren't even a solution either, arguably. You have to then engineer the layer 2 bridge to have it's own anonymization and escrow handling technology that is disconnected to the Ethereum network entirely. And realistically speaking, "closer to zero" does not mean free (or even at negligible cost). L2 chains can only exist when transactions on the mainnet are made impossible due to unbalanced gas prices. It's a catch-22.

How did you not realize that one can run/evaluate code without actually broadcasting it? All you need is access to a node (plenty of public ones) and you can "simulate" any transaction(code) you would like.

Testing a dapp off the mainnet is like ensuring your website works on localhost. It will find some issues, but it's not representative of how it will look in deployment.

In any case, for actual usage it should surprise nobody why everyone conflates Ethereum with money. No, your L2 chain does not qualify as an official solution.

No, you can test transactions as they would happen on mainnet (tests with mainnet state). Or if you want you can fork mainnet and do your stuff there.

It's absolutely representive of how it looks in deployment. You can test transactions EXACTLY how they would happen on mainnet.

I don't get your second point.

> Testing a dapp off the mainnet is like ensuring your website works on localhost

I would argue the exact opposite. A website will be deployed to different versions of different browsers on different operating systems. A smart contract will exist on a single distributed computer. It sounds like the actual problem is people treating smart contract development as cavalierly as web app development

coti is also using dags for some reason

Since distributed blockchains / databases are distributed and therefore need partition tolerance by definition, that leaves consistency or availability for the shortfall in computation.

I'm guessing any "crypto-kinda-currency" is picking eventual consistency as a core mechanic. Think about the word EVENTUAL though.

If the core function of the crypto is a ledger, then it makes sense, it EVENTUALLY gets transacted, and in practical terms you take the faith in the distributed system from a flawless previous record in reconciliation, probably before the actual completion of the transaction.

Now, a distributed major blockchain has ... how many nodes? Thousands or more? That is a long time for reconciliation of the consistency, even with great dedicated internal networks. What? This is over a heterogenous global internet network? That implies EVENTUAL has some bad worst cases.

"Smart contracts" or "distributed trustless computation". Whatever, getting the value of calculation from a node and getting the value stored in the node is essentially the same thing in terms of determining an answer to a query.

It implies a horrendous performance, one you have little control over. I don't think Kubernetes is shaking in its boots.

It's interesting Aphyr never does any crypto analyses, although he makes his bones running a test suite. How do you test a scaled cryptocurrency?

Reply on news.ycombinator.com