4 ms·
Not really surprising, to be honest. I'm really tired of the overused "blockchains are really slow databases" phrase. They're not, and if you're using one as t
by tuxxy 8y ago
Not really surprising, to be honest.
I'm really tired of the overused "blockchains are really slow databases" phrase. They're not, and if you're using one as that, you're doing it wrong almost 90% of the time.
Here's a great question to ask yourself if you need a blockchain:
Do I need game theoretical enforced security?
There are many networks and protocols that can be improved with such security mechanisms. There are more that can't be improved. Just in another thread, someone asked the question if Tor could be improved with it. There's plenty to experiment with here and figure out what makes a blockchain application great.
The entire idea is that you can incentivize good behavior and disincentivize or punish bad behavior. Consensus comes into play here when deciding, "Was this operation done correctly? If so, do this. If not, do that." This has the ability of enforcing correct operation in decentralized networks. My associates and I have taken to calling these "Decentralized Incentive Networks". Yes, there are many problems that have yet to be solved, but people are working on them. I'm really interested in seeing solutions to the oracle-problem, but I'm unsure if it will ever be solvable.
Even then, a lot of these applications won't be fit for enterprise (or even small business) use. Many of these applications will benefit (and survive) from community adoption and rely on being open source.
Hopefully, when the hype is gone, we will have made some really great technological advancements to peer-to-peer and distributed applications. This is where I think blockchain apps will find their niche.
- ethanwillis 8y agoThere's quite a few people working on the oracle problem. But there's always risk at some level. And even oracle solutions get broken down into decentralized/centralized platforms.
- kllrnohj 8y agoThe singular point of blockchain is to establish a paper trail. The singular point of Tor is not to have a paper trail. Those are literally opposites. Blockchain + Tor would be the sort of complete nonsense the article is talking about. Blockchain has nothing to do with security. It's not a security technology. It's a permanence technology. It ensures that something is never modified in the absence of a trusted authority which it does with security technology and distributed proof of work.
- tuxxy 8y agoYou've completely missed the point. There are people who are using Onion routing with nodes who are economically incentivized to behave in a non-malicious fashion. If caught acting malicious, then your node can be penalized. It's not like you're actively putting all traffic on the blockchain. You're arguing a strawman here. I'm not going to speak for these projects directly because I know very little about them, but I can speak as to how you would incentivize non-malicious behavior.
- kllrnohj 8y ago> You're arguing a strawman here. You didn't really provide anything to actually talk about. You took two literally opposite techs and slapped them together and then acted like it was obvious what that meant. Blockchain has no application to Onion routing nodes, either. That, again, would be the sort of thing you don't want a permanent record to track. Tor is specifically designed for the routing to not be stable and not be known. It is fundamentallly the polar opposite of blockchain. Blockchain is a public list of immutable data, identical for everyone. TOR nodes form a private list of ephemeral data, fleeting and ever changing. Unique to everyone. > If caught acting malicious, then your node can be penalized. Blockchain literally has nothing to do with this. It has no such capabilities of any kind. Hell, it was literally built so that you can't have penalties as that's governance. > I can speak as to how you would incentivize non-malicious behavior. Blockchain doesn't do this, at all. If anything malicious actors love blockchain because there's largely no capability to recover from a successful attack.
- baby 8y agoZcash.