Failing projects pray blockchain works as 'magic middleware'

“This is the year of pointless blockchain projects” and anything you build with blockchain will need to be ripped out and replaced within 18 months, according to Gartner fellow Ray Valdes. Speaking to The Register in Sydney today, Valdes said blockchain is among the most secure technologies he's ever seen, having survived …

[Anonymous Coward]

Spot on !

I've been tasked with deep understanding of blockchain for my employer and have just submitted a report saying exactly this.

It's amazing how many people out there are "ready now" to sell you a blockchain solution who - being kind - haven't the faintest idea how blockchain works.

Don't me wrong - the underlying hype *is* true. But needs a whole infrastructure to grow up around it first.

My prediction (and where you could invest a few spare quid) ? "Blockchain brokers" who will take on the computing (and energy) intensive part of actually *running* a blockchain, and act as "BSPs" for the wider world. They will also keep publicly available copies of *previous* blockchains. Because one thing my research has shown is blockchains will grow. And grow. And grow.

Oh, the acid test for seeing if you are being sold snake oil, is to let the putative vendor sell you the "next big thing" and then ask them for a job in "the next big thing".

AC because I have already pissed off some Hipster start ups ...

[A C Oward]

Re: Spot on !

Thanks for sharing your knowledge.

Can you please further share how and where you went about trying to get the deep understanding?

I wanted to learn to similarl levels and so far haven't found any useful training.

Thanks in advance.

[Anonymous Coward]

Re: how and where

start by spinning up some VMs and joining the test ("Morden") network. Then see how slow it is.

Then go back and read up on genesis blocks, and create your own 2-node Ethereum network.

Then run the Wallet, and watch as you see how CPU intensive everything is. Learning one is that even with an iPhone 27, your average punter is *never* going to be "part of the blockchain".

(Immediately balanced by all the starry-eyed hipsters telling you that "everyone will be on the blockchain").

This leads to conclusion one: the future will see an emergence of firms running blockchain nodes o behalf of blockchain players. It's where I'd invest a few quid (only these firms don't really know they exist yet).

Having created a basic "private" blockchain, then marvel as - after some time - you run out of currency.

Learning two: you need to understand the maths in order to efficiently generate currency. Conclusion two: "consultant mathematician" could be a new job of the future.

and so on.

It's why I am bloody good at R&D, and why most UK firms aren't.

[Jim Mitchell]

Re: how and where

"Having created a basic "private" blockchain, then marvel as - after some time - you run out of currency."

This is a feature of bitcoin, not the blockchain itself. Corporate use of blockchain is not going to be for currency, but making transactions verifiable and non-refutable. As noted, "private" blockchains do not make sense.

[Anonymous Coward]

Re: running out of currency.

it's a feature of the implementation. However, nodes charge "currency" for processing. This is one reason why Joe Soap ain't ever gonna run a blockchain node on his iWiz

[dajames]

I don't believe it!

Someone from Gartner has apparently said something sensible. I think I need to go and lie down.

[asdf]

Re: I don't believe it!

Yeah kind of rich of Gartner considering they are usually the ones that plant these fads in CIO's ears based what company bribed them the most.

[streaky]

Yep

I've been saying this for some time now, it's a solution looking for a problem and instead of letting it be applied naturally where it makes sense (it genuinely could be a solution to some actual real-world problems) it's going to get shoehorned into a lot of things where it doesn't belong or simply isn't needed.

This would all be fine because Darwin - the issue is a lot of it being floated by government departments and it's going to be money wasted at taxpayer expense.

[Anonymous Coward]

There's no point for blockchain inside a network perimeter

No company needs blockchain for its own transactions, a database works just fine. The only real value of blockchain is for distributed transactions traveling over insecure networks. There are potential issues though if the transactions are high volume, as the blockchain size will quickly become unwieldy unless you have some method for expiring/archiving old transactions.

Use inside a corporate network where it doesn't belong will result a lot of duplicated storage. Deduplication will help in theory, but a lot of large enterprises have many file servers, so you'll end up with one copy of the blockchain on every storage device - a point in favor of very large centralized storage for all those file servers I suppose (that will make EMC/Dell happy)

[Dadmin]

Re: There's no point for blockchain inside a network perimeter

That's a very great point. Have some Uppers!

It's funny to think of CIOs running around wanting this "latest new thing that sounds so cool; Block! and then there's Chain! It's giving me a CIO boner already!" And even funnier to imagine them scratching their heads and staring off into space for an inordinate amount of time once they have to come to grips with the distributed and open nature of this security method they plan to run on individual machines behind big, mean firewalls.

"Wait! He did it again!"

What?

"Firewalls! It's like Fire, and then some Walls! Where can I buy some for my enterprise?"

Can you go away now?

[Anonymous Coward]

Missing the point

blockchain is a state machine, not just a database. It can actually compute in ways databases can't.

[Sorry that handle is already taken.]

Re: Missing the point

It can be. As far as I understand it, bitcoin's isn't, but ethereum's is, for example.

But is that even useful?

[Anonymous Coward]

Re: But is that even useful?

Presumably, that was what Alan Turing heard when he published "On Computable Numbers, with an Application to the Entscheidungsproblem"

This is the distinction between being "in IT" and (like me) being a "Computer Scientist"

[Sorry that handle is already taken.]

Re: But is that even useful?

He may have but that doesn't necessarily prove anything (remember what Carl Sagan said about Bozo the Clown?) A blockchain is a specific solution to a system that needs to be trustless (i.e. cryptocurrency) and is tremendously inefficient as a result. Sure, it's not widespread technology and is relatively new (although it has been in use in Bitcoin for seven years now), but it sure hasn't got off to an auspicious start...

Ethereum for example is currently a complete disaster zone as a consequence of its blockchain's programmability.

[Anonymous Coward]

Re:Ethereum

platform is pretty good.

Any problems are due to badly written contracts. Don't blame the language for the programmers snafu ... there's plenty of shit C/C++/C#/VB/Python/PHP/Fortran/COBOL programs out there too.