[usrbinbash]

And how does a distributed Ledger solve these problems?

Monopolies like Google and Facebook exist, in no small part, because the amount of computation and data they handle is vast, and they have the data centers to deal with that.

How much data gets has to be stored on the servers of such services? Per second? I'd assume its in the range of several GiB...again per second.

Okay, so how does the blockchain compare to that? Ethereum can store data, each byte requires about 600 of its "gas" computational equivalent. A block represents 30,000,000 gas, 1 block is generated every 15sec, so we can store a grand total of about 1MiB every 300 seconds...that is, if none of the gas is used for anything other than storing data, which means, no other computations running.

So how is "blockchain technology" going to solve the problems that come from such highly centralized services exactly?

[ique]

Monopolies like Google and Facebook exist because of advertising. Remove advertising and they don't exist.

The fact that they're processing a lot of data is in large part because they need to for advertising.

Still, your understanding of blockchain tech is misleading here. Ethereum is public key registry at best and is not and should not be used to store or process data.

In a blockchain world you can still have service providers, but the user is the one with the power, not the service provider. Users are free to switch service providers as they see fit because their identity and data isn't tied to a single company.

[usrbinbash]

> Remove advertising and they don't exist.

Remove advertising and the amount of users for most social networks drops to oblivion, because barely anyone wants to spend money every time they post a picture of their cat.

> The fact that they're processing a lot of data is in large part because they need to for advertising.

They need large amounts of storage because millions of hours of video & audio, billions of food-pictures, tens of billions of lines of text, and a megagagazillion of references on who-like-clicked-what-when, take up a lot of storage.

> and is not and should not be used to store or process data.

Well then, what should be used? What decentralized storage solution can handle something like youtube, where 500 HOURS of video were uploaded PER MINUTE in feb. 2020?

And storing is half the deal. The solution also has to have high availability, consistency, low latency, and needs to be environmentally sound.

[dmitriid]

> Users are free to switch service providers as they see fit

No they are not. That is, they will need just as "free" to switch in the non-blockchain world.

Why? Obvious, isn't it?

- Blockchains can't store the amounts of data required (Youtube/Instagram/TikTok on blockchain? What a nice joke)

- Even if you somehow can, companies will store data in their own proprietary ways incompatible with each other

- And, of course, this data will be on different blockchains, some of them invented specifically for the purpose

In reality though, as we're seeing it with NFTs all the data will be centrally stored with only some meaningless tokens referencing it stored on blockchains

[ique]

Data will never be stored on blockchains, its not what they’re for. It’s amusing that you’re so vehemently against something you know nothing about.

[usrbinbash]

So if its not stored on the chain, where will it be stored, if not on central servers, where its ultimately under control of whoever owns them?

In a distributed network? How does that handle the data loads and requirements (availability, latency, security) of services on the scale of fb or youtube?

[dmitriid]

ique: Users are free to switch service providers as they see fit because their identity and data isn't tied to a single company.

me: data can't be stored on the bockchain, it will remain proprietary, so good luck "being free" and switching between service providers

ique: Data will never be stored on blockchains, its not what they’re for.

So, how exactly are users going to be "free to switch service providers" if their data will literally remain in a walled garden of the service provider?