z Kalevalla

Modest Proposal for Decentralized Social Platforming

09-11-2025

The idea of decentralized or distributed social networks comes as a response to the many problems that centralized (mainstream) social networks have developed over the years like:

  1. Privacy: the company that owns the servers owns all the data you produce on their platform and has the right to use it as they see fit. Of course, they have some form of “Privacy Agreement” where they say they don’t do any nefarious thing with it, a) but can you believe them? And b) are they really bound to honor that agreement?

  2. Censorship: They own the platform so they have right/responsibility to “moderate” to make sure no “disinformation” or “harmfull” information spreads on their platform. The problem is that there is a very fine –almost transparent, depending on who you ask– line between “moderating” for the safety and comfort of the community and “censoring” for the benefit of their own political agenda, whether that skews to one side or the other.

  3. Algorithms: “If it’s free, you’re not the customer, you’re the product”. And what is their product? Your attention, which they sell to the companies that advertise in their platform. And in order to have more product to sell they have to secure as much of your attention as they possibly can. And in order to do that they will resort to many devious, manipulative methods to get you practically addicted to being constantly connected. This is in short what’s known as “algorithms”. Over the years they have found many different ways to incentivize your constant prescence, and many can be extremely negative; one of the worst, and also one of the most effective, is the idea that very few things keep your eyeballs stuck to the screen like controversy, polarization, and general negativity.

In short, mainstream social platforms have become toxic, almost cancerous, and need to be extirpated from society.

But, are the current idea of decentralized social platforms the right answer?

They do try to adress these problems in some way or another.

The concept of decentralization itself is an answer to mostly the problem of censorship: since the very structure of the platform means your data is not centrally hosted in just one particular server or data center, owned by any one entity, then it can’t be censored, because it could be removed from one server, but still exist in another completely independent server.

As for privacy and algorithms, I don’t think they have a good enough answer yet. Their main answer seems to be that there is no monetization whatsoever, thus there is no reason for them to sell your data or to create any algorithm to manipulate you. But, even assuming we can trust them because they say they are all about open-source, non-censorship, etc., that generates it’s own problems, namely: they become unsustainable and there is no incentive for growth. They depend on maintainers who must be both techically savvy and ganerous with both their time and money. There aren’t too many like that. Enthusiasm is great, but it’s not a real strategy for the long run.

The Answer:

Free as in freedom doesn’t necesarilly mean free as in free beer

Some one has to cover the costs of hosting the servers, right?

I say it is, in fact, a more free environment when everybody knows exactly what they can expect, demand, and what their role is in the ecosystem.

So what does this look like in reality?

The Protocol

In one sentences: A lightweight, decentralized social network powered by **small, independently hosted PHP nodes**, with a unified client experience, no ads or algorithms, and fair, transparent economics.

The idea behind the protocol is quite simple, and it could potentially apply to many things, not just social platforms. Like any such application it has two sides:

Everything is open source so it’s auditable; the financial incentive is covered by the user, so there is no room for evil algorithms or data brokering.

The registration/installation fee is very cheap and is calculated to cover the expense of renting a regular shared hosting where the server software is hosted. An average of low-tier hosting service may cost around US$6/month ($72/year) and the capacity of said host is good for 100 active users; a $1/year subscription fee would generate $28 gain average per server at full capacity. It’s no get-rich-quick type gains, but it’s enough passive income to be more incentive than just sheer enthusiasm for the platform. Or you could make it $2; it’s still cheap, but how much would the average user be willing to pay for something they already get “theoretically” free from Meta, X, etc. But for the assurrance of no data-mining and algorithm manipulation and with enough network effect, $2/year may very well be worth the price.

In summary:

1. The Problem

2. The Solution (in one sentence)

A lightweight, decentralized social network powered by small, independently hosted PHP nodes, with a unified client experience, no ads or algorithms, and fair, transparent economics.

3. How It Works

  1. Global Entry Point (agora.monster)

    • Acts as a registry and lightweight load-balancer.

    • Maintains a heartbeat list of all healthy nodes.

  2. Nodes as Infrastructure

    • Anyone can install “Agora Node” on basic shared hosting (e.g., IONOS $6/mo PHP/MySQL plan).

    • Each node caps at ~100 active clients to preserve decentralization and performance.

    • Node operators add a PayPal address for automated revenue share.

  3. Client App / Web Interface

    • Chronological feed, no “recommended” algorithm.

    • Simple UI reminiscent of classic Facebook—focused on conversation, not metrics.

    • Auto-discovers and connects to optimal nodes (geo- or load-based).

  4. Data Distribution

    • User profiles and posts are redundantly stored across 2–3 nodes (eventual consistency).

    • Clients read/write from any peer-node without “instance” lock-in.

PS: I would love to say that I'm the right person to develop this monster of a project, or even to collaborate with a team and do it. I am in the process of teaching myself php but that's a bit far from maerializing yet. So I'll just leave this here and if someone sees it and wants take a stab at trying, be my guest, it would be awesome if you invite me to the project, or at least talk about it.