[merrickread]

Couldn't agree more. We're using some blockchain technology to solve specific problems at my company. It's a part of our stack, but we don't mention blockchain unless someone specifically asks.

Blockchain is sensationalized beyond belief. I can't wait for the rose tinted glasses to fall off and we realize it's just a data management technology for specific use cases.

I think the biggest danger to the technology is that the community focuses on "blockchain will change the world" not "how can we make this better".

[pfftcst]

I've always wondered about this- all the hype about 'Blockchain'. Can you explain this a little bit more? How do you have a blockchain without bitcoin/ether or at least mining? Isn't that just a glorified database? What does a (I assume centralized) blockchain do for your company that a database or any type of electronic ledger cannot do?

[merrickread]

Yes, right now it is completely a glorified database. It's always been a database. It's a different database.

We're using it for compliant sharing of and collaboration on private information records between non trusting organizations and businesses. You could do this with something like a highly distributed Cassandra cluster. But some of the inherent rules to a blockchain allow us to provide higher security, higher data persistence and availability between nodes, auditable records, and accountable participants / contributors.

And it's not centralized, it's completely distributed - just not publicly available to anyone who can download a mobile app.

[ufo]

If there is no mining isn't this just like git and other systems that use merkle-trees under the hood?

[merrickread]

This does a pretty good job of explaining the differences if you're curious: https://medium.com/@philippsandner/comparison-of-ethereum-hy...

[TazeTSchnitzel]

how to get people to care about your computer science work

1. s/Merkle tree/blockchain/

2. there is no step 2

[imtringued]

It's a database that is designed to be very expensive to update but easy to verify. If someone still manages to update it with an invalid state other nodes will validate the new block and not accept it.

[acangiano]

The trouble with blockchain is that the technical discussion is often overshadowed by investment narrative. If you'd like to learn more about the actually useful applications of the technology beyond "the future of money", research Hyperledger and smart contracts. We recently launched a free introductory course on the subject of blockchain if you are interested in the business use case: https://cognitiveclass.ai/courses/blockchain-course/

[neilwilson]

I'm rather more impressed by the DAG systems than the chain. A tree always seems more sensible that a linked list.

[merrickread]

If you're into consensus algorithms take a look at Algorand:

https://people.csail.mit.edu/nickolai/papers/gilad-algorand-...

It has some novel approaches to node / data agreement.

Also take a look at the many projects Hyperledger projects. Hyperledger is incubated by The Linux Foundation. Their attitude towards open source development is that of purely technical curiosity and experimentation. Brian Behlendorf is directing the project so it's in good hands. He speaks about this attitude here:

https://www.youtube.com/watch?v=pr4Hb0jb0lo