/biz/ - Business & Finance

[Anonymous No.22838400]

Explain sharding to me

[Anonymous No.22838412]

>>22838400
imagine something that looks exactly like pic relating coming out your ass when you thought it was only a fart

[Anonymous No.22838429]

>>22838412
How would it make my ass more scaleable? Wouldn't its function be impaired by it? Are you trying to tell me that Ethereum is a piece of shit?

[Anonymous No.22838435]

>>22838400
Shards are essentially their own blockchains that run parallel to the main chain (Beacon chain). So you can have a bunch of independent shards that live in their own ecosystems and process transactions, then to finalize the transactions they send the information to the main chain which validates the information. It's a form of parallelization.

[Anonymous No.22838456]

It’s when you think you just have to fart but you shit your pants

[Anonymous No.22838477]

>>22838435
Why does it matter whether they run parallel or not? Why this solution works better than if they would operate on the blockchain?

[Anonymous No.22838504]

>>22838400
you think its a fart but a little poop comes out

[Anonymous No.22838577]

>>22838477
You are right.
ETH 2.0 is overcomplicated
51% attack on 1of 64chains could be easy as fck.

[Anonymous No.22838628]

>>22838400
Less than a shit but more than a fart, it might leave some stains in your underwear

[Anonymous No.22838646]

>>22838400
It's like shitting your pants but with glass shards splitting your bum.

[Anonymous No.22838663]

>>22838412
Actually Vitalik approved this analogy on one of his ama's

[Anonymous No.22838680]

>>22838400
Sharding is like farding, but with sauce.

[Anonymous No.22838739]

>>22838477
Because it takes some of the processing off of the main chain and uncreases the transactions per second.

[Anonymous No.22838779]

>>22838739
How does it matter where the transactions are processed?

[Anonymous No.22838790]

>>22838400
>Explain sharding

[Anonymous No.22838808]

>>22838577
imagine thinking you know more than Vitalik "the drug addict" Buterin.

Go ahead and solve it yourself then faggot

[Anonymous No.22838812]

Its when sergey takes a shard allover vitalik

[Anonymous No.22838816]

>>22838477
If you have to run two programs on a single computer, the way it traditionally works is it runs the first one (say it takes 10 seconds to run program 1) and program 2 is waiting for program one to finish before it can start running (another 10 seconds) for a total of 20 seconds runtime.

If you split the blockchain into shards, each program can run on it's on shard so now running both would take only 10 seconds. So you've done two different transactions in 10 seconds instead of having a bottleneck on a single blockchain that has a long line of programs waiting to run (transactions)

[Anonymous No.22838855]

>>22838816
Alright, I get it. What are the drawbacks of this approach? Why wasn't it used from the beginning?

[Anonymous No.22839076]

>>22838779
Right now ETHEREUM can only process 15 transactions per second. If I have 2 ETHand you have 6 ETH and I want to give you 1 eth then you want to give me 4 eth. Those two transactions have to occur on the chain. If you process those two transactions on a paralell chain and instead just have one transaction processed on the main chain (record the new balances of you have 3 and I have 5 ETH.) Thats a lot more transactions per second which can be processed.

[Anonymous No.22839195]

>>22839076
>Thats a lot more transactions per second which can be processed.
yeah but eth still can only go 15tps. its just a temoparay scaling meme.

[Anonymous No.22839223]

just take the glass pipe to your mouth and enjoy the fucking gas fucking free fucking fees mother fuckah

[Anonymous No.22839258]

>>22838855
ETH wasn't designed for scalability, and probably never will be. Miners and OG eth holders made a lot of money on fees

[Anonymous No.22839294]

>>22838456
Kek

[Anonymous No.22839334]

eth 2.0 is a joke lmfao can't wait for you noobs to get rekt

[Anonymous No.22839422]

>>22839195
Eth 2.0 is supposed to be way faster. Ithink eth is on it's way down. Algorand is going to destroy ETH. Ripple and Algorand have teamed up to help Bank of America explore crypto. The Marchall Islands using Algorand for the first ever national crypto currency, where crypto will be their fiat. Algorand will be the death of Ethereum

[Anonymous No.22839622]

>>22838400
Marketing buzzword since 2015 and will never happen.

[Anonymous No.22839739]

>>22838855
Sharding/parallelization is a very old concept in computer science but was not understood when it comes to blockchain technology (particularly proof of stake blockchains). There were many years of Ethereum research needed to get to the point we are today in scaling blockchains.

One drawback is composability - ETH2.0 will allow for limited composability in that things can run asynchronously in a limited scope.

Example - shard 1 needs information from an application in shard 2 in order to complete a program - shard 1 can await the information shard 2 processes and sends back to shard 1 to complete its program. This will be supported but anything more complicated wont be. Currently ethereum (on a single blockchain) can handle crazy amounts of composability - the kind of shit that lets a single ethereum transaction touch a dozen different apps and wait for the results of interacting with that app before moving on to the next command in the program.

But even if some degree of composability is lost in ETH2.0 it is definitely worth it for the scaling capabilities. Dapp developers will just have to work around the new limitations.

[Anonymous No.22839861]

>>22839739
Does this mean transaction isolation will be lost?

[Anonymous No.22840065]

>>22838577
anon, I... do you even understand pos?

[Anonymous No.22840370]

>>22838477
because there is a latency in block reaching most of the network.

if you mine blocks faster, the network will produce a lot of uncle blocks (blocks that are valid, but turn out others are more valid), and then the network will build upon then, then try to ditch them and go on new chain of blocks that turn out to be less valid then others again.

in short, it will take more time for the network to verify the most valid blocks than the time to create block, and will simply lag behind in blocks, have parallel long chains of blocks that will eventually nullify each other.

companies now assume it is safe that a transaction is safe after 15-30 blocks are passed in ETH, with a 10 times faster block mining that can easily jump to +10'000 for example.

with a bit faster block mining it can be less trustworthy and more unstable, and with even more it will be simply unusuable.