That'a what I'm wondering too.. on where that data lives.
It appears that it will NOT be part of the blockchain.
Much of the information that app developers want to leverage is not consensus-critical; that is, it either does not belong in a blockchain node or it doesn’t need to be accessed directly from one. Instead, we can store all the relevant information in a traditional database and keep it in sync with the blockchain, while adding new custom layers of logic and features.
How can they boast that they have a superior blockchian when their solutions is to add layers where the bulk of the data will live. Databases which ARE NOT decentralized and worse, the features which separate a blockchain from a regular system WON'T be part of that layer...
I don't see this as a good thing.
Fragmentation DOES NOT make a system superior.
Hive processes only the data on the blockchain, nothing proprietary. To "talk" to hive you write to the blockchain.
What does this mean then:
and:
That's what I refer to. How does this equal decentralized.
I understand what HIVE will be. I want to know what the entire system will look like once all these layers are implemented?
If it's all fragmented how can it be considered secure and blockchain based if more and more of the data is going to exist on regular systems?
It just means hive specializes in certain types of data for specific applications, and providing convenient APIs that wouldn't make sense as part of the steemd consensus node.
There's really not all that many layers here, and it makes for a more agile environment. If there was one daemon that did everything, nobody would be able to run it. So now we can have more lightweight engines talking to each other. You could run steemd+hive yourself and it will be less painful than a full steemd node while offering more power. Or you could run hive and sync off a public steemd node.
Because more users can run more lightweight services, I see this as increasing security and anti-fragility. The blockchain remains the source of truth, I don't see the issue with storing copies of blockchain data on "regular systems" for easy access.