You are viewing a single comment's thread from:

RE: What Makes a 'dApp' a 'dApp'?

in #steem2 years ago

I can't help but think of D-Live as an example of the use of the Dapp buzzword. People were so happy to have what appeared to be a smooth functioning "Decentralized"-Live streaming service that most never bothered to check under the hood to see if the platform was what it claimed to be. Dapp - Dupped maybe?
Seems pretty simple to just call your App a Dapp, and most people won't ever bother to check for themselves if an application is centralized or decentralized.

Thanks for this article, I've never thought of the prospect of a DEX front-running a transaction to earn small profits. With enough volume and over time this could be a substantial amount.
Definitely food for thought.


It's possible for parts of an app to be decentralized and parts not. With Steem, most people really just care about the decentralized reward system (or really they probably just care about the reward system which happens to be decentralized, but that's beside the point here). So D-Live was a centralized app which utilized the Steem decentralized reward system to reward the content.

I think that is totally fine, and would like to see more apps use that model (vs the dtube model of trying to fully decentralize everything and having a poorly working product as a result).

Indeed, d-live was one of the catalysts for me wanting to write out my thought's on this topic. I was always thinking "That's not a dApp!" but there really hasn't been a concrete definition of dApps or the degree to which they are truly decentralized.

The DEX example I think is a really powerful one. It shows a clear problem, and one that can only be solved by removing the centralization -- the middle-man sitting in front of the real dApp. Which I will explain how that's possible in part 2 :)

Right on! I'm looking forward to Part 2 as someone who has become increasingly interested in blockchain/crypto/decentralization I'm really interested in learning more about ways in which non-tech / non-developers can play a roll in helping bring actual decentralization to fruition.

Probably getting ahead of myself but any chance you'll be exploring IPFS in your follow up piece at all? I'm curious to know if IPFS or an alternative to it like MaidSafe could transform the landscape and what that might actually look like. But that's a big question, lol, I'm sure there's no simple answers here.


So, IPFS (which is actually very similar to torrent files on a theoretical level) has all the 'nice' properties that come with being stateless. Because of this, IPFS and torrents don't fundamentally need a cryptocurrency platform to enable them (though you can use one, you don't need to).

By stateless I mean this: once you solve the bootstrapping problem (e.g. finding the hash of the file you're looking for, or the torrent file you want), verification is actually quite easy as it doesn't require external validators. The hash of the file always matches -- as it has no state, the file contents don't change, and thus it is easy to verify the hash of the file matches the hash you are looking for.

For statefull content, think about your account balance. There's no way for you to validate your own account balance in constant time -- you need to know the current state of the whole ecosystem and it's entire history to know your balance, as your balance is actually not a single "thing" with cryptocurrencies: your balance is rather the sum of all your historical actions regarding your balance.

So, unfortunately file storage doesn't help with statefull content -- this is where we have that necessity for validation, and where we run into the trusted interfacing problem. If the trusted interface misses (or hides) a historical action, they hide the truth. And unlike with stateless content, you cannot immediately validate on your own that an item is missing.

Another case in point would be steem monsters. Yes your cards "belong" to you but the way they are randomly selected is not verifiable. The same for the combat system, from what I understand.

Nevertheless, this is true to many platforms. Ethereum, that popularized dApps, also has its own share of problems. Chiefly amongst these, is the fact that most apps use Infura as a proxy to the blockchain. Again, outsourcing trust... But of course people are aware of it. Until proper light clients come in this is the only way to have dApps that have some ressemblence of a decent UX... Plasma chains are another option...

Truth be told, DLive was suspicious for me from the start.

@heimindanger's DTube is the only solid truly decentralized attempt out there right now.

Posted using Partiko Android