Live data from Hacker News

Achtung: Decentralize, decentralize, decentralize

drewdevault.com

101–110 of 174 posts

Re: Achtung: Decentralize, decentralize, decentralize

#101

I would like to decentralize, but projects like Masterdon are not the way forward. If you want to decentralize the servers then you MUST also decouple user accounts from the server they signed up with. Today, if you switch to Masterdon (or perhaps someone you know is running a diaspora node), everything appears to go well until a year from now when the volunteer who was providing the server loses interest or moves on…

Actually, what you're describing is a problem of federated services. True decentralization has no servers, and thus none of those problems. Take twister for example [1]: * Identity management via blockchain (one of the use cases where IMHO this is a good, or at least defensible choice). * Metadata storage in a DHT * Content storage in torrent swarms None of this depends on any single peer/server. [1] http://twister.n…

> Take twister for example

Y'all got any more of them "examples"

I'm now aware of two social media alternatives that aren't dead accounts walking. Twister is in beta, but that looks interesting enough to play with and keep an eye on.

Re: Achtung: Decentralize, decentralize, decentralize

#102
post #30
post #25

Earlier quoted context omitted.

> If every user paid $5 a month, Facebook's shareholders would do fabulously well. > But of course, that's a ridiculous proposition. Only an infinitesimal number of Facebook's 2.2 billion users are willing to pay for the service. Well, at least some tens of millions of users seemed to be happy about the idea to pay USD 1 a year for WhatsApp. They were running this operation with something like 50 engineers so it shou…

WhatsApp without any server-side storage nor content discovery? These are primary requirements towards a social network and they'll cost more than another dollar to fulfil.

You can rent a dedicated server for less than $10/month that has enough resources for an estimated 100+ users. Heavy users that upload a large amount of data (videos, pictures) may be willing to pay more than casual users.

Re: Achtung: Decentralize, decentralize, decentralize

#103
post #88

Earlier quoted context omitted.

> Obviously only my personal intuition but I'd wager there's not much of a business in a pay-to-play social network. Then it's not much of a business, period. In my ideal world, we'd only have businesses where the costs are open and direct (e.g. no "user is the product" or selling of personal data, or even "ad-based"). If they can't make it in this way, then, sorry, they don't have a business model.

We would never have newspapers in that case. Very few have ever been able to support themselves on the price the user pays.

>We would never have newspapers in that case. Very few have ever been able to support themselves on the price the user pays.

Well, if there are no buyers, then why have a product?

Re: Achtung: Decentralize, decentralize, decentralize

#104

Earlier quoted context omitted.

No, but some of these may be legally required in order to operate as a competitor to Facebook.

I don't think anyone is saying the alternative should operate as a business competitor to facebook. I think the idea would be for the service to be fully decentralized, where no one has the power to remove anything

Services where no one has the power to remove anything quickly become flooded with illegal content leading to LEAs taking them down.

Re: Achtung: Decentralize, decentralize, decentralize

#106

Earlier quoted context omitted.

Actually, what you're describing is a problem of federated services. True decentralization has no servers, and thus none of those problems. Take twister for example [1]: * Identity management via blockchain (one of the use cases where IMHO this is a good, or at least defensible choice). * Metadata storage in a DHT * Content storage in torrent swarms None of this depends on any single peer/server. [1] http://twister.n…

> Take twister for example Y'all got any more of them "examples" I'm now aware of two social media alternatives that aren't dead accounts walking. Twister is in beta, but that looks interesting enough to play with and keep an eye on.

Scuttlebutt is a pure p2p social network. Active and lovely https://www.scuttlebutt.nz

Re: Achtung: Decentralize, decentralize, decentralize

#108

Or just build a dapp on eos and solve all the issues at once : decentralize, data ownership, users can make money by choosing what they want to share with advertisers, everyone is essentially an owner.

Let users generate tokens based on how much data they choose to share. Advertisers purchase tokens and proceeds are split between token holders based on their share of total tokens. Everyone wins - users are in full control, advertisers reach an audience, everyone gets paid.

Re: Achtung: Decentralize, decentralize, decentralize

#109
post #29

The post repeats the myth “There is no promise that a company can make to its users that outweighs the fiduciary duty that obligates them to maximize profits by any means. The only defense of this is legislation and consumer choice.” As a counterpoint, see, for example, [1], that quotes the Supreme Court Hobby Lobby decision[2]: “Modern corporate law does not require for-profit corporations to pursue profit at the ex…

Its not about law, its about investors. Big investors have outsize control of the boards of most companies, and they pretty much only care about profits. The public doesn't blame them either because we don't talk about how decisions made by those on the board shape the entire company. Investors have little to fear from unethical decisions, so they make them.

The article actually contains a good point regarding investors and external influence:

> Companies like Facebook, publicly traded, have a legal obligation to maximize profits for their shareholders. Private companies with investors are similarly obligated. Nowhere in the equation does it say that they’re obligated to do anything for you - the only role you serve is to be a vehicle for exploitation.

But I see an exception to this rule: Self-financed companies where the founders are still the shareholders and where the founders are in the company for the long run, and where the company is financed by the users.

If your company is financed by the users, then in order to make the business sustainable you need to optimize for the users. If the users demand privacy, then you optimize for privacy.

I'm aware that slow, sustainable growth is not popular in silicon valley, but that's probably a question of culture (in many european countries a lot of startups still follow this path).

Re: Achtung: Decentralize, decentralize, decentralize

#110

Earlier quoted context omitted.

No, but some of these may be legally required in order to operate as a competitor to Facebook.

I don't think anyone is saying the alternative should operate as a business competitor to facebook. I think the idea would be for the service to be fully decentralized, where no one has the power to remove anything

The advantage of Mastodon is that people do have the power to moderate and take things down.

But only on instances they are administrators off.

So while I can take down illegal content on my instance I can't go to the japanese instance and ask them to stop posting all those anime girls. They follow their own rules and if I don't like them I can just not allow their content on my server.

If you want to follow both my server and the japanese server you'll have to find an instance that doesn't ban either of us or run your own.

(Though I trust that there are a good number of lightly administrated instances that it'll be easy to find one before you're forced to run your own)

Post reply on HN