[endorphone]

"I think all the hype around block chain is hilarious for the very simple reason that block chain can’t actually solve any business problem that a centrally managed service couldn’t solve"

This is a very common statement on here regarding blockchain solutions, and is effectively like saying "There's no program you can't write with 1991 Visual Basic" (or build in Excel -- I've had financial sorts try to hammer every problem into an Excel workbook). Theoretically true, but outrageously off the mark given the nuance on needs of many very varied projects.

Blockchain style solutions are not relevant for many projects. They have some legitimate uses, however, and saying "But you could..." isn't a retort.

[Sangermaine]

>Blockchain style solutions are not relevant for many projects. They have some legitimate uses, however, and saying "But you could..." isn't a retort.

Except you've got that backwards. Since none of those solutions actually exist yet beyond hype, it's the blockchain proponents who have to show how using blockchain technology in those situations is not only feasible but superior. Saying that you can already solve those issues with existing techniques IS the default.

[endorphone]

No, I don't have it backwards. Blockchain solutions offer decentralized, trustless alternatives to existing centralized, trust-based solutions. Further, it's bizarre reading this constant claim that nothing exists beyond hype, when a number of orgs have operating or in testing blockchain solutions for industries from manufacturing to finance. Inertia is a hell of a thing, as is "the way we do things", so it isn't surprising that it didn't suddenly replace everything, but it does have uses.

[zprime]

>No, I don't have it backwards. Blockchain solutions offer decentralized, trustless alternatives to existing centralized, trust-based solutions.

The problem is that the more we look at potential use cases, the more we found that trust relations are necessary to handle a load of basic process requirements that cannot be solved by blockchains alone, thus requiring the addition of trusted actors, thus removing the need of blockchains to begin with.

Blockchains proponents tend to see trustlessness as a benefit by itself. In their view, Blockchains are better than traditional databases because trustlessness is better than traditional client-server models.

System designers see trustlessness as a constrain. If you are in a bad situation where no single actor can be trusted by everyone else, then you can implement a decentralized database using blockchains. It's going to cost a lot in term of performance and efficiencies, but it's doable. But as soon as this constraint of truslessness is removed, why the hell would you bother with something like blockchains? And experience show us that, well, there's actually not a lot of contexts where trustlessness is an absolute requirement.

[scient]

Please do list a couple of useful solutions where blockchain is the right/better tool for the jobs and thats outside of crypto coins or secondary layer of payment abstraction (and even there any payment option could work better in a lot of the cases).

[m0llusk]

Property title insurance is a great example. With blockchains it should be possible to reveal the owners of all properties and all transactions in order to remove ambiguity from the system. It is true that constructing and making use of such data implies trust in some sort of organization or process, but a robust store of ownership data is currently beyond keeping of property title records as currently performed by entrusted government entities.

[cleong]

Why would this work better? Not to be cynical but it seems from a engineering perspective that the trade off is security for data robustness? Anyone who takes control of the network will take control of all property title insurance data. In return at least i know up time will be 100%...

[redog]

Isn't it just basically triple entry accounting?

Blockchain tech is ultimately about trust. Trust that the accounting hasn't been altered. You don't quite have it solved with just a double entry ledger right?

[rlpb]

> Blockchain tech is ultimately about trust. Trust that the accounting hasn't been altered.

A simple git repository offers this property. Really it's just any data structure where only growing it is permitted. You don't need a blockchain to get this.

[floatboth]

https://twitter.com/whitequark/status/946886702932557824

[alwaysdoit]

The implicit part of this argument, though, is that blockchains are expensive: in order to have this trustless system you have to have PoW which is inherently computationally (and therefore monetarily) expensive. It's not going to be easier to write either (which is the reason why people aren't using 1991 Visual Basic).

In other words, the argument here is why would you do it this other way that is more complicated and/or expensive when you could do it the default way which is simpler and cheaper.