[GTALUG] Blockchain, the solution to nothing

[ac]

I practiced with my first top post in 2006, before it became
fashionable :)  top posting is quite useful in communicating a singular
idea. Not trimming the original post simply states that imnsho the
quality is high enough that you can read it again. If you happen to be
someone that only scanned it the first time - or if you missed it
completely do read it, it is very cool and very valuable.
(I am re-adding it to the bottom of this reply)

On Thu, 27 Aug 2020 06:54:40 -0400
Scott Allen <mlxxxp at gmail.com> wrote:
> On Thu, 27 Aug 2020 at 05:56, ac via talk <talk at gtalug.org> wrote:
> > +1  
> That's got to be one of the lowest signal to noise ratio posts I've
<snip some crap here - to save a few bits> 

Yes, there is such a lot of crap, opinionated trash and mind numbing
idiotic arrays of bits that when one discovers gold, it simply cannot
be "trimmed" as it is actually worth something.

Unlike of course my opinion of your own waste of bits :)

my 2c

Andre

On Wed, 26 Aug 2020 14:59:12 -0300
Mauro Souza via talk <talk at gtalug.org> wrote:

> It's really funny. It's more the result of "lets use blockchain!" on
> everything that creates those issues.
> 
> And blockchain is hyper-overhyped... It can help in a few specific
> problems, but not on every single problem without exception. But
> everyone liked the concept and ran away screaming blockchain all over
> the place. Blockchain on email. Blockchain to rent a car.
> Blockchain-powered lettuce salad... that's insanity. I like the
> technology, I study it and sometimes I answer things on forums, but
> all the time when people ask me if I believe blockchain would help
> them, I say NO
> 
> The "everybody must have everyone's data" is an issue. As the solution
> grows, the ledger grows, increasing the storage and processing
> requirements. With less people connected, you need an intermediary to
> access the blockchain, and those intermediaries can be attacked and
> take down the network. And there's few solutions to that.
> 
> One solution is the blockchain used by former Raiblocks (now Nano).
> It's a DAG (directed acyclic graph), so any client will have only his
> own transactions, and not the entire ledger. That allows one to run a
> client on an ESP32, for example. there will be special nodes that
> store the entire ledger (like the Dash supernodes), but the clients
> don't need that. Who/what would use it? No idea... crypto coins, and
> not much else.
> 
> And the smart contract solution... They aren't smart, they aren't
> contracts, and they aren't solutions. One mistake and everything
> crashes down instantly.The Ethereum DAO disaster, the Parity
> multi-signature contract, and that new DeFi that melted down in a day
> because people realized they made a mistake on the code, and nobody
> would ever have enough tokens to decide anything because of an
> arithmetic mistake. And many more examples.
> 
> If the "contract" can be changed after creation, you cannot trust it
> because it can be useful now (like tokens on a web game that you pay
> to play), but later the owner changes the contract, pockets all
> tokens, and takes down the game. If it cannot be changed, any error
> on the contract is set on stone forever with deadly consequences.
> There are some countermeasures to that, like proxy contracts: a main
> contract references secondary contracts with the functions, but the
> main contract holds the data and the tokens. if a secondary contract
> is found to have an error, you deploy an amended version, call a
> function on the main contract to reference the next one, and done.
> The downside is that it is more expensive to run this contract, and
> the owner can, you know, replace the secondary contracts and steal
> everything. You can add multi-signatures, quorum, external oracles,
> but those only increase the cost and put a little protection against
> a rogue owner.
> 
> There are very few things that blockchain can be useful, and one of
> them is distributed storage. Siacoin and Storj, for example, let you
> rent the extra space on that 4tb disk you already have for some
> coins. It is not profitable enough to make you buy storage just for
> that purpose, but you already have the space, right? And you can rent
> some space on the network for backing up things when you will
> reformat your computer, and want to store your data in case something
> breaks. It's cheaper than anything else, even cheaper than amazon
> glacier.
> 
> Those supply chain management things are useful too, if you can
> integrate it correctly (and that's a BIG if). If you own a bakery,
> for example, you scan all the ingredients you have, store them all on
> your wallet (or have your purchasing software do that automagically)
> and any time any of your suppliers have any recall on something you
> got, you are warned. And everything down the line gets warned that
> they bought a recalled bread from you. Your customers don't even need
> to get back to you so you reimburse them, the next time they come to
> buy something they already have a credit for that contaminated bread
> you sold to them last week. But to integrate everyone is a nightmare,
> there are lots of privacy issues, industrial secrets issues... on a
> limited scale, it can work. Too limited and doesn't solve anything.
> Too broad and all that privacy issues get into your face.
> 
> The fact is that there are very few cases that blockchain can be used
> that a database cannot. And databases are here since a long time ago,
> everyone knows how to build them, operate them, backup them, and
> extend them.
> 
> But just wait for the AI-generated, solar-powered, graphene-based
> multi-cloud stored 6G-capable IPv8-addressing blockchain...
> 
> Mauro
> http://mauro.limeiratem.com - registered Linux User: 294521
> Scripture is both history, and a love letter from God.