small edit: this blog will soon move to https://blog.cyrneko.eu, if you want to read this post there you can do so and future edits will go there. For reference, the post here was edited at 09:36 CEST on the 16th, a day after publication. The one linked above may be newer.

Those are the contents of a post I recently made, but really that and even the replies I made are not the full story

Truth is, to get right to the point, the fact that Matrix was accompanied by a for-profit entity, funded by venture capital was the biggest mistake that Matrix as a project has ever made.

Element is not a friend

In roughly the beginning, there was two organizations that came out of the project: The Matrix Foundation and New Vector Ltd / Riot / Element. The idea was for New Vector Ltd to carry out the necessary work and bring in the necessary funding for the Matrix Foundation to thrive. Or well, so I've been told.

They had multiple funding rounds lead by the likes of status.im, Automattic, the AI and Web3 company protocol labs and others; You get the gist, lots of VC and similar funding also a questionable amount of “Web3” and bullshit generation AI. Element was then tasked with using that to build the software that would power Matrix.

And for a long time, they did that. They relied on the software themselves but kept it in the hands of the Non-Profit Matrix Foundation.

Until the 6th of November 2023 when they—in their words—moved to a different repository and to the AGPL license. In reality, the Foundation did not know this was coming, and a huge support net was pulled away under their feet.

Element's “re-focusing” on “establishing a level playing field” means hostile takeover of all important projects that were under the Matrix Foundation banner and to stop running and managing the Matrix.org homeserver despite it still being the default option in Element today.

The results of this are, as one may expect, devastating. I don't think I've seen the Matrix Foundation ring the alarm bells any more than today that they need funding to keep the foundation going. Unfortunately, all the money is being swept up by Element instead.

Of course I understand there is not really an alternative as of right now; No one else wants to take up Element's job, by which I mean the job that the foundation pays them to do now instead of it being donated to them. Yes, the high expenses for the Matrix.org homeserver are largely because they are still managed by Element, just not as donated work but instead like with any other customer.

This also means that the Foundation suffers from Element's decisions and is why they pay a hefty price for what would otherwise not be this expensive.

Today this leaves the foundation in a dire situation.

So dire in fact that they are starting to adopt things that I can almost guarantee many on the governing board do not like.

The Matrix Foundation is making Matrix.org a freemium service.

Now, and I can't stress this enough, I really don't think many people at the foundation want this. But with Element sort of just pushing whatever they need in their client and nothing else, I doubt anyone would even be able to get anything implemented in Element to notify Matrix.org users akin to what Thunderbird or KDE started doing in their respective products. As such the governing board does recognize that measures like these are kind of necessary, even if ugly.

Either way it shows that Element is seemingly cashing in on selling ,Matrix to governments and B2B as a SaaS solution without it going back to the foundation, without it funding critical parts of the core of matrix that need to be revised (like moderation, or a lack thereof) or the Matrix Foundation.

At the same time I can't help but think that this could have been prevented. Even Matthew himself recognizes that putting the future on Matrix on the line with VC funding and alike was not the best idea for the health of Matrix.

Matrix should, from the start, never have been this heavily tied-into and reliant on VC funds to keep the project as a whole afloat. Ultimately, for-profit companies will do what makes them profit, not what's the best option. Unless the best option happens to coincide with making the most profit.

Unfortunately, supporting the foundation through anything more than “in spirit” and a platinum membership is out of their budget, apparently. I think that morally they owe a lot more than that.

So, what now?

If you believe Matrix can still thrive despite, in my eyes, being sabotaged by New Vector Ltd, please do go donate.

If you're like me, and you've seen Matrix fail too many times and have concerns about the sustainability of some of the core design decisions, there are some other projects you may be interested in.

This list is split into two, things that I personally want to recommend and things that were recommended for me to include by others. Everything I am recommending here specifically isn't tied to VC funds or a for-profit entity, at least not to my knowledge.

Personal:

• Polyproto / The Polyphony Project – Made by 🏳️‍⚧️ people, aims to have Discord API compatibility and builds a “boring” identity federation protocol with multi-homing and user-owned identity and data in its design. (donate)
• Delta chat – Builds on traditional E-Mail standards like SMTP and IMAP, enhancing it with end-to-end-encryption, a custom server stack and a full instant-messenger experience. Additionally has webxdc. (donate)
• Revolt? – An open-source discord clone. The ? is there because whilst it's okay, it is not federated which makes me a bit hesitant to recommend it as an alternative to matrix of all things.

other's recommendations:

• XMPP/Jabber – battle-tested instant messaging standard with lots of client apps for major platforms. Despite me running an xmpp server I don't personally recommend it due to clients taking a while to catch up with features, i.e how Conversations currently lacks replies and most clients lack the ability to delete/retract messages.
• IRCv3 – An evolution of the well-known IRC standard with lots of quality-of-life and functionality improvements that are to be expected from modern chat applications. I haven't personally used it so I can't personally recommend it.

Anti-recommendations:

• SimpleX Chat – Many suggested this and I will explicitly recommend against it due to the founder's positions on various topics. This includes being anti-vaxx, believing COVID-19 was a hoax, trans- and homophobia, climate denial; In the SimpleX Groupchat he's also been seen basically bootlicking trump a couple times, but I've lost receipts to that.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

Wanna make some sort of new gadget? Chances are you'll have to use USB-C for charging. Wanna make a nearby-share like feature? You better use open WiFi standards to facilitate connections.

These are just two examples you might heard of, namely Apple and others being forced to use USB-C or forcing AirDrop to use Wi-Fi Aware instead of a proprietary standard. Or how WhatsApp was forced to allow interoperable messaging.

Some may argue the reception to this has been mixed, but really in my classes, my family and my friend group everyone uniformly loves changes like these because it breaks down barriers. It allows people to have chat messages between ecosystems, use the same charger, 'airdrop' to their android friends, or have tightly integrated smart-home setups.

These kinds of things are constantly being seeked out by people, remember Beeper? They were hugely popular because one thing they did was break down iMessage's barriers.

What I'm trying to say here is, time and time again we've seen that people get excited over standards being established and barriers broken down, and that it generally ends up just being integrated into every day life until people can't imagine a life without it; We're starting to see this happen with USB-C, people don't even have Micro-USB or Lightning devices around anymore, and when we say “do you have a charger?” it's just assumed that it's going to be USB-C

And quite frankly, you should care about that.

It may seem a bit mundane in nature, it's just some standards, just every day life, but I firmly believe you should care about open standards just a bit more because without them, we would slowly be locked into corporate ecosystems and under their control.

It really doesn't take much, just look for standards-based alternatives to the things you already use, or try looking for something like that before you make the decision to use a particular thing; I.e if you're looking for an instant-messenger, look for one based on open standards and see if you like it. And if you don't, be sure to tell 'em! Many of these projects are not only based upon open standards but often Open-Source, meaning they work in the open, often for the public, and they depend on feedback and (financial) support from people like you.

And even if you don't end up using something based on open standards, that's okay! I'm not asking you to like every single open standard there is, but I believe them to be incredibly important to our freedom and our personal independence, or more specifically our independence from big-tech corporations that dictate how we use technology.

I mean, that is also why my devices run Linux, for independence and openness of the ecosystem; For the same reason I've started using XMPP in addition to Matrix (sort of) for my instant messaging needs.

(this post is from my backlog of posts, I tried to clean it up a little but quite frankly lacked the energy; hopefully my point comes across)

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

Not out of malice, but because the current Matrix ecosystem is unsustainable.

In the Foundation's attempt to become more independent, they've given control over Synapse – The most feature-complete, most widely deployed, and only long-standing Matrix Home-server Implementation – over to Element and ended up making the entire Matrix ecosystem dependent on a for-profit, VC-funded company that does not care for core issues in the process. The way that other projects like Dendrite have faded away (and also been given to Element!), or how Conduit's development cycle is incredibly slow does not inspire confidence in other projects either.

To that comes the fact that many parts of the Matrix protocol are due not just fixing, but entire re-structuring to keep with the times. The server-to-server parts of Matrix have major issues especially in regards to state resolution and resistance to malicious federation. Not only is it almost impossible to defend a room (or letalone a whole set of them) against spam and abuse, there is also an attack vector in breaking federation in a room altogether that is being abused in places like #main:postmarketos.org

As it stands right now, I do not see the mainline Matrix.org Foundation fixing any of these issues. Not just because of them being closely adjacent to Element still, but also due to a lack of funding; Even if the Matrix Foundation was truly more independent by now, they would still lack funding required to do their work, and to do it well.

As such, I express my wish for a hard fork of the Matrix protocol. It should be lead by those who truly care about the core of the protocol, trust and safety and user-experience first; Not by who is funding a certain feature in Element; Not by the needs of B2B customers, all of which are mostly unaffected by these core issues that the core community has suffered from over the past few years.

And if it doesn't work at all? Shit happens.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

This situation was now resolved. Please do refer to this document here: https://doc.micr0.dev/s/yjVU8vCdc#

Micr0, AKA @micr0@fuzzies.wtf, @micr0@wetdry.world, and micr0.dev has accused me of being a pedophile.

You may know Micr0 for making AltBot, for co-running the FediMC Minecraft Server [archive], or for just being another person on the Fediverse.

Over the past roughly 20 days, Micr0 has accused me of, and convinced the administration of wetdry.world that I am a pedophile, behind their closed doors. He has reached out to me only after I noticed and talked in public about passive-aggressive comments that Micr0 left in a groupchat we have shared membership in, and how I noticed I was suspended from the Fuzzies.wtf Mastodon instance.

This all starts a few days ago, in a groupchat that Micr0 and I share membership in, I noticed that I could not see any of Micr0's messages and asked about it, then had someone else forward my message so Micr0 would see it.

The question was whether Micr0 had blocked me, and why.

The response by Micr0 was as follows:

Idk why that matters or how that is relevant I am in this gc for my friends, i previously left because of alexia and choosing not to support her ways but ow I came back so that I can be with my friends. I blocked her so that I can be with my friends sorry I don't need help from a well [if you know] then [you know]

receipt is here

Following this, I made some vague posts on my Fediverse account, starting here. because this thread is followers-only, here is a screenshot of said thread as a receipt: https://files.catbox.moe/vi90f9.png

(sorry, I can't provide a full transcript of that right now.)

That was then followed up with the following posts: Thread number one: – https://shrimp.starlightnet.work/notes/a6v7iva6ddnpdpzv – https://shrimp.starlightnet.work/notes/a6v7ji237hkkxp5x – https://shrimp.starlightnet.work/notes/a6v8vlrsbtr32dj0 – https://shrimp.starlightnet.work/notes/a6v8wxhdzk6a5ncn

Thread number two: – https://shrimp.starlightnet.work/notes/a6v997347806sg5w – https://shrimp.starlightnet.work/notes/a6v9bgsarvz5apnb

Thread number three: – https://shrimp.starlightnet.work/notes/a6vac7eegywc95h9 – https://shrimp.starlightnet.work/notes/a6vaeztqhys1pdp1

To keep it short, I had noticed that I was entirely banned from Fuzzies, and my account was hidden on Wet-Dry World. Following this I contacted Essem on XMPP to inquire about my hidden profile. I did not contact Micr0.

Essem read the message, and swiftly handed the situation over to Micr0 which then contacted me instead. The full transcript of that conversation (with some personal details redacted) is here: https://files.catbox.moe/vntyxq.txt

This covers all conversation I've had with Micr0 since, but I'll only cover part of it for this next section.

Effectively, Micr0 talked to me about the situation and brought forth two arguments for why I've been banned, outlining that these are “pedophilic behavior”

1. Micr0 claims that I had text interactions sexual in nature with a minor, implying them to be under the age of consent.

This is quickly disproven, as said person is over the age of consent in their respective jurisdiction. We have an age gap of 1 year and 4 months. To respect their wishes, I will not be publishing their name.

1. Micr0 claims that in a prior conversation about a predator that is free-roaming on Fedi, that I sympathized with the pedophile after Micr0 had asked me for advice on how to deal with the situation.

This requires some more context, but I cannot make the chat history public as that would invade the privacy of Micr0 quite a lot. Instead, I offer a summary:

Micr0 had contacted me about feeling powerless over someone that is known to have engaged in erotic role-play with a minor at the age of ~30. After some back and forth to clear up some confusion I had in the conversation, I was given the information that the way Micr0 wanted to deal with it is by 'cancelling' said person. I reiterated twice on how I think that is a bad idea because of not having sufficient evidence, and 'cancelling' or public exposures usually not being taken seriously, or it delegitimizing evidence that could be given to authorities.

Micr0 saw me disagreeing with the idea of cancelling as me sympathizing with the predator.

Fast forward to the last two days.

I have reached out to Essem as previously described, but still have not gotten a reply. Instead, all the information is still handled through Micr0. Micr0 had told me that the post I made should be edited to reflect that the decision was “not personal”. Hoping that I could get unlimited, I complied, and even went a step further and made a standalone post stating this.

48 hours pass without any kind of interaction. Micr0 had said that either he or Essem would reach out to me “tomorrow”, yet that had not happened, so I sent another message asking for an update.

And an update I got. My appeal was rejected. I was explained that my account would continue to be limited, and it may potentially be banned from Wet-Dry World, my old home, altogether.

At the same time as that message, Niko had contacted me, telling me that Micr0 had said:

im just facing a difficult decision right now since a pedo got banned from fuzzies and wdw but they want to use altbot and stuff .... – Receipt

There is is. No doubt, Micr0 is convinced that I am a pedophile.

I was later told by Nelson that he was contacted by Micr0 as well during that time.

During these 48 hours, I was constantly on-edge, but especially after it was confirmed that my appeal was denied, I had an incredibly high pulse, had trouble controlling my breathing, and generally felt like I was on the edge of a full on breakdown and/or panic attack.

Once I had calmed down a little, I went outside for a walk to get some fresh air, in which moment Micr0 contacted Niko again this time with the following:

Niko? Niko: Yeah Micr0: I have a random question this is very stupid but I need to know [in reply to “I've been tasked of collecting some evidence to support the recent moderation actions of wetdry world”] did you tell someone about this? – Receipt

In other words, Micr0 was tasked with asking all of my closest friends behind my back with the explicit purpose of finding receipts of me showing “pedophilic behaviors”.

So just as my pulse went down far enough that I could breathe, I immediately found myself furious again.

This leaves us at now. Micr0 is probably still looking for evidence in my contacts and between my friends to justify the decision of banning me from Fuzzies and Wet-Dry World, and for evidence to use against me to call me a pedophile.

Small update before we continue: Whilst I was preparing this document, another message entered my inbox. Namely:

Micr0: hi again, just wanted to clarify that we won't be contacting other instances about this situation. the limits will remain only on wetdry.world and fuzzies.wtf – nowhere else. i thought knowing this might ease some concerns.

To this, I simply replied “I had no concerns of that nature.” I personally believe that this reads as Micr0 not reaching out to other instances, as you usually would with a case like this, because of the evidence not holding up. – Receipt

I find the delays, the accusations, and the emotional roller-coaster this has caused me to be unacceptable and disgusting.

This has cost me my sleep. This has cost me my peace. This has cost me almost passing out. This has cost me so much time writing up this document TWICE as I lost it once due to power outage. It has cost me being on the verge of a panic attack for hours on end just desperately trying to calm myself down.

I am NOT letting my name get dragged through the mud.

Micr0 and Essem,

I am deeply disappointed in how you handled this situation. I am disappointed by how you went off of assumptions of my nature, the age of those I interact with, and false interpretations of my words to paint the picture that I am a child predator. I am disappointed that you did not even care to do the bare-minimum age and jurisdiction check to see if any of these claims even hold up. I am disgusted at how, behind closed doors, I explicitly got called 'pedo' without remorse.

But most of all, I am angry. I am angry that the person I spent so much time dragging out of the mud would turn around on me and use dirty tricks to justify calling me a pedophile, and I am angry at how Micr0 seemingly has failed to realize what kind of consequences an accusation of this magnitude could have for my life and my future. It is devastating.

And I know you two are going to read this. Good Night.

This document was signed off by:

• Nelson
• Neil [archive]
• Niko
• and 1 other person that signed off privately. Contact me for details on that.

All of the above approve of this document, and can confirm that what is written here represents the truth.

Update as of 19:46 CEST, April 24th of 2025:

I have had people reach out to me saying that this is not an isolated case. I am not giving more information at this time to preserve the privacy of those which contacted me, but it seems this kind of internal misinformation may allegedly not be a one-time mishap.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

This post is part of a series of posts about viable alternatives to discord. You can check the #discordAlternatives tag above for all blogposts in this series.

Revolt is a pretty young, centralized platform initially started as a hobby project by Insert sometime in (I believe) 2020 to see if they could build a clone of Discord. Pretty simple goals.

Over time it grew, insert had a couple more friends join for moderation and development, and now there is a small team working on it in their free time—including an Android app and soon an iOS app And whilst I certainly see the connection between seeking an alternative to Discord and Revolt...I don't see how anyone would recommend Revolt today as an alternative for communities, especially large ones. Quite frankly, the platform is not built out in some key ways yet for it to really be viable for most communities, especially those that like to meet up regularly

So, let's go over what I think is still missing before I can recommend it personally, shall we?

Voice and Video chat

This has been a long one for Revolt, with initial work on, well, re-working that entire part of Revolt happening in the backend in...2023? 2024?

Either way, reworking this has been on the roadmap for a long time, as there is technically speaking a Voice Chat feature but it is very rudimentary, kind of buggy, and entirely nonexistent on the mobile apps

That said, the plan is to have voice & video out by the time the new Web/Desktop client releases, which is gonna take a while still as Insert needs to find the time to actually be able to do that

Pinned messages

Currently, Revolt has no concept of pinned messages. Well, maybe in the API, I remember hearing something about that, but definitely not in any of the clients. As such I am gonna say that pinned messages simply aren't a thing on Revolt as of right now, which is kind of a bummer as a small yet still very frequently used feature.

I guess this one could be lived without...for however long it's gonna take still

End-To-End-Encryption

The main reason I am including this here is because quite frankly it's been quite some time since this has been on the roadmap and, sadly, there is a lack of transparency in especially this part unless you happen to catch one of the developers in the official Revolt community

Effectively, this was confirmed to become a killer feature a while back (like, literal years) and since then not much has happened. Really, huge features like this probably shouldn't be confirmed until you've got even a vague idea of what's gonna happen. Or if you do have a vague idea of what's gonna happen, be transparent about it in the context of an Open-Source project.

API-Only features

For some reason, some features still haven't made it past the “we have it in the API!” stage despite being used a lot on the platform.

Masquerade

Masquerade has technically been a thing for a while now, but only really usable if you installed a specific plugin with the current/old web/desktop client.

It allows you to define a profile per-message, e.g having a different profile or avatar for just one message

This is especially useful to plural systems, but also for bridge bots and the likes, and the fact that it was added and supposed to come at some point and just never did is kinda disappointing. At least it can be used via the API I guess...?

Gradient role colours

Another feature that only made it to the API and never the Clients/UI is gradient role colours. Not an essential thing, but it is weird how all clients can display it and none can configure it. You'll HAVE to use the API to get gradient role colours, or some bot.

Spam on the platform

Another issue is that right now, there are no robust anti-spam or anti-flood measures built-in to the platform. There is AutoMod, maintained by a Revolt contributor, but it isn't built-in functionality and as such only discoverable at worst when something has already gone wrong.

But, worse than that, their servers regularly get DDoSed, leaving the platform unresponsive. And when they aren't getting DDoSed, the most common part to break is their CDN / Media service.

Point is, most unpleasant to anyone that tries to run and moderate a community. At least the tools that exist actually exist and work unlike certain other projects, but it's not good either.

Small Stuff

Here is just a collection of small nitpicks that I wanted to include but don't warrant their own entire sections.

• Still no way to see where an emote came from; Android app only lets you do this if you're already in a guild
• You cannot put guilds into folders, making it really annoying to be in many of them—especially if you're just in there for emotes
• No way to index outside of Revolt—Content here will be stuck inside guilds. Discord has the same issue and third-party indexers are NOT helping.
• Lots of features are in theory ready, but held back by missing client rewrites and updates
• The only channel type, not counting the upcoming Voice/Video channels, is text. Whilst I'm not for making un-indexable pseudo-forums, having threads-only channels is still useful for support communities and alike; inevitably hosting a forum takes a server and money. We live in an age of (cg)NATs.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

This post is part of a series of posts about viable alternatives to discord. You can check the #discordAlternatives tag above for all blogposts in this series.

Matrix is, nowadays, one of the most recommended “Discord Alternatives”, especially if your audience resides on the Fediverse where one half will absolutely hate it, and the other much louder half will be those that believe in Matrix for one reason or another. And I get it! It is federated, replicates and decentralizes data¹ has tons of servers and a corporate entity behind it that provides it with funding...is what some would say

The reality is, matrix lacks funding², lacks moderation tools, lacks spec-compliant clients, has hundreds of open MSCs with important changes that can impact the ecosystem as a whole, make it more secure, more robust, more reliable, safer for everyone

In real-life scenarios, projects have had to give up Matrix entirely—despite self-hosting and despite having a moderation bot—because the abuse that is happening on the platform is simply too large. Projects like Librewolf are regularly getting spammed with literally illegal material—the same reason other projects simply left—and have to fight against this for themselves.

They cannot restrict who can send media, they cannot setup any raid protection, federation is simply too slow for anyone to react in time, including bots, which aren't even fast enough either. There is also no way to filter events before things happen because a malicious server could easily just patch out this filter.

In other words, moderation is probably one of Matrix's largest weakpoints. It makes it absolutely unsuitable for any even moderately sized community, especially those that are potentially already vulnerable (i.e queer communities). Yet somehow it still gets worse, on a more technical level.

Due to how Matrix is designed, servers that participate in a room (i.e users from different servers all talking in the same room) have to do state resolution to figure out what the current state is. This is because there is no “owner” to a room in the sense of “a server where all the data resides”, all servers have that role and as such they require a mechanism that allows them to synchronize this data. The algorithm however is brittle and far from perfect. When a fatal error happens, it often happens that two participating rooms get different state, and as such...

State Resets

A state reset is when one or more servers decides to reset their state back to a “known-good” point. This leads to any state changes done since then, like membership, permissions and more being entirely undone up to the point that effectively, they never existed.

This is of course a major issue, as it means that any moderation done up to a given point is effectively undone. It makes people unable to communicate properly with those who aren't also on their server, and that is just the start.

Especially when many changes to state happen frequently across many servers, state resets become more common. Like, you know, when trying to moderate.

To this comes a whole host of features that Matrix is simply lacking.

• Up until recently, there was no native VoIP, only Jitsi, and Element Call is still not enabled by default unless you're using Element X on a mobile device. But Element X is considered unfinished so...
• Good, platform-wide (e.g widely supported in apps/clients) support for custom Emoji and Stickers. Element still lags behind here and the relevant MSC(s) still isn't/aren't merged.
• Scheduled Messages.
• Content Warnings and/or Spoilers for media. There is an MSC, to my knowledge only Cinny supports it.
• Multiple attachments (i.e images) per message. Not a single client supports this.
• Portable Identities—i.e not being stuck on one server and having to move everything manually. There are MSCs about this, none are implemented anywhere.
• Metadata Privacy. Currently lots of metadata is still available to a given server that really should not be, like replies and other relationship events (e.g Emoji Reactions, which are never encrypted.)
• Profile Descriptions. There simply are none, you cannot add any extra information to your profile apart from a timezone in some recent versions of Element and Gomuks(-web). Some clients may also support status, but the largest homeserver (matrix.org) has that entire feature disabled.
• Proper message forwarding. Some clients may support this but they effectively only copy/paste a message into a different chat. There is not even any indication that it's been forwarded.
• Sharing live location.

On top of all the above missing features, some of the already existing features are not even consistently implemented between clients.

Some STILL don't show media captions, Element X (the in-beta mobile rewrite) cannot display custom emoji at all, and NO ONE can decide how to do text formatting properly (seriously, maybe <br> shouldn't be usable y'know?), some implement MatrixRTC and others stick to legacy VoIP, some have widgets and some don't... And that is just the tip of the iceberg. Seriously I've barely scratched the surface of weird inconsistencies between clients–I dare you to find one client that implements the spec 100% accurately, and mind you, it's NOT Element.

My final point is that, essentially, Matrix is—as it stands—simply not suited for anything serious especially when it comes to hosting communities. Notice how I didn't even touch on the unable to decrypt meme that often goes around...that would've opened a whole new can of worms

If matrix wants to become viable, it must find the funding and invest into some core areas of the protocol that are quite frankly, broken. This includes not just moderation and T&S³, but also finding solutions on reliable key exchange even when the wider network is unreliable, finding solutions to “split-brained” rooms and state resets, getting (security) audits from the get-go and not just after an unaudited library was used and recommended for years. The list goes on, there is a lot of issues to unpack and tackle.

Oh also, I KNOW THAT THERE ARE MSCs TO FIX SOME OF THESE THINGS. Shut up. I've heard too much about it already, infact I tried writing some in the past and most of them ended up dead before I even submitted them

1: Well, to some extent. One notable exception is how Media is not decentralized. 2: Especially now that the Matrix Foundation wants to fund everything themselves and not have their bills paid by Element anymore. 3: Trust and Safety

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

Background by Maxim Tajer (unsplash), logo belongs to Discord Inc.

Recently, news about Discord looking into an initial public offering made the rounds online.

Usually, an IPO does not mean anything good for end users, especially for software products or anything-as-a-service situations. For us, it only means that external investors will gain more control over the platform than they do now and will demand a return on their investment. If they did not already have a grip on Discord through its venture capital roots, they will definitely have a grip once Discord goes public.

I expect existing features to become paid, new features to become increasingly egregious, and the platform as a whole to become more hostile toward its everyday users. They will likely leverage the network effect even more aggressively to lock users in, and I believe already established a tight gripthey may even introduce true advertisements—unlike what they have done already, which some are still defending as “not ads”.

All in all, one would think that people would get tired of Discord and finally want to move. That is easier said than done though.

So, why hasn't anyone done so yet?

This question is not a new one, and the search for a replacement has been a long one.

If you're anywhere even remotely adjacent to digital privacy or tech, especially around decentralized technology, you've probably heard about Matrix or—God forbid—XMPP as potential replacements. Some may have advocated for Signal, while others might have even mentioned Revolt.

The issue with all of these platforms—aside from their unique challenges—is that none of them can truly compete. Not because they lack the resources, but because of the network effect.

The network effect is, in itself, a neutral force. It simply means that the larger a network is, the more well-integrated and functional it becomes.

Companies understand this, which is why they will exploit it. They will make it difficult to leave the platform or create an illusion of a utopia where users have no reason to leave.

You can bet your ass that if Discord weren't already trying to guilt-trip users into staying because they have a long-lasting Nitro subscription with a funny badge, or due to the various reward and gamification systems they've implemented over the years, the new shareholders would find a way to exploit network effect.

In other words, people have tried and failed to do it because Discord has already established a tight grip on its users.

but even if we imagine that network effect didn't exist...

What got people here in the first place?

If you were part of the gaming scene in 2015-2016, you’ve probably used Skype or even Teamspeak to hang out with friends until a new option started to become more prominent. More and more forums replaced their IRC and Teamspeak links with Discord ones. With the promise of not needing to pay to host a server, the best option became clear for many.

No longer would one hear the Skype ringtone, as it was swiftly replaced with the Discord ringtone as people started to move.

Over time, Discord got robust moderation and community-oriented features, customization, custom emotes, and so it quickly became not just the home of many friend groups but also of larger online communities. Nowadays musicians and content creators alike find their home there.

In other words, for a long while, Discord simply provided a good product. It works, it works well, it has many features and allows you to stay in touch with friends.

Sunshine does not last forever

Eventually, Discord needed a way to make money. After all, they had to pay back the venture capitalists who got them here in the first place, right?

So, they got to work on implementing Discord Nitro. Higher file size limits, the ability to use emotes anywhere, and streaming in high quality—the whole shtick that you probably already know.

These are all paid of course, monthly or yearly, and now all that Discord needs is a way to make people use it

So, they started putting references to it everywhere. Click on an emoji? It tells you about Nitro. Open the emoji picker? It tells you about Nitro. Open your profile? It tells you about Nitro. Open someone else's profile? You get blasted with an animation, and then it tells you about Nitro. The point is, they became very aggressive in reminding you every couple of minutes that Nitro was a thing. Not to mention the full-screen pop-ups that would appear every once in a while after an update, along with all the A/B tests they conducted to see what they could do to encourage more users to purchase Nitro.

Even in just trying to make VC investors happy, they already began sharpening their game in upselling. And so, more and more users became upset with the platform and started to look for alternatives

What little options we do have

Let me just say this from the get-go: our options kind of suck.

They are buggy, too early in development, just as VC-ridden as Discord, underfunded, or simply a hobby project by a few students. Or any combination of the above.

Optimally, we would have something that could replace discord in it's most fleshed-out and core features, whilst also making it easy for developers to port their software, offer some of Discord's most egregiously advertised features for free, provide stronger privacy guarantees, and all that whilst being decentralised to a reasonable degree.

One possible candidate for many is probably going to be [matrix], but as much as I'd love for that to be a good option, it simply is not—especially not for larger communities (e.g., 500-1000+ members). This is primarily due to the fact that significant specification work must be done by the Matrix Foundation before it becomes viable. Element also needs to step up and undertake serious refactoring of their client and UX across the board—some of which is happening—before it becomes viable.

There are few tools available to protect against this spam; even matrix.org's own moderation tool, Mjolnir, offers little to no protection. The same goes for downstream forks like Draupnir, which significantly improve the situation but are not a complete solution to spam either.

There is another up-and-coming project, namely Polyphony / Polyproto that meets most of the criteria above.

It has a simple model to decentralisation in which authentication across servers is simplified, does not try to do complex data replication across servers, does not have a complex DAG or state resolution, and has goals to have Discord API compatibility (to some degree). It has history in the Spacebar.chat project which has previously attempted reverse-engineering the Discord API

This makes it a really good candidate for fulfilling both the decentralisation and Discord-Compatibility check-boxes. Sadly it does not really compete with Discord yet as most of the work is currently in backend, nothing for the end-user yet. They also did not get the NLNet grant they were hoping for, which would've accelerated their work on Polyphony as a whole massively.

What I wanna say with that is if you have a project you wanna contribute to, it's probably Polyphony. If you know rust, go help them out!

A Dire Situation

All this is to say, we desperately need a good replacement. We need something that the average user can pick up where these qualities we want to avoid another Discord situation are met without them having to think much about it. Where the user is better respected, their privacy protected, where they are given choice and control.

And we just don't have it yet. That is the reality of the situation.

I have hope this might inspire some to take action, help these projects which are trying to build something better or maybe even build something of their own. Whichever way this goes, my hope is that this might inspire change in the landscape of...whatever category Discord belongs in nowadays.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

This also means I've learned a trick or two about optimizing the resource usage of my laptop to make it...well, usable.

Don't get me wrong, it is in no way good, my phone is faster; But it IS usable.

I got an Asus X551CAP for 20€ off of (eBay) Kleinanzeigen, because some woman fucked up the windows install beyond recognition and thought it was broken. Luckily for me, it actually just needed a re-install

(Part of this blogpost is also just so I remember in the future should I ever have to setup a craptop again)

#1: Modern Windows just...won't cut it

Very short section but, as if I didn't already prefer to run Linux on my devices anyways, I pretty quickly learned that Windows 10 and 11 very obviously would not work on this thing. It has 4G of RAM, and when your desktop shell is written in react native (which I remember the W11 one being), you'll have a bad time on a device with such specs. Windows 10 did not really fare that much better, at least no recent version of Windows 10. I did not and do not intend to test LTSC or other non-user-facing builds.

#2: Use Compressed Swap on RAM, AKA ZRAM

Once I had Linux installed, I quickly learned that even slapping an SSD in, swapping to disk simply took too much time. So instead, I set up zram.

Zram allows you to, instead of swapping to a disk, swap to a compressed region of memory. So, instead of swapping to a swapfile or partition, the swap data gets compressed on-the-fly and moved into RAM. With algorithms like zstd, this compression is incredibly efficient.

At the time of writing, with (Ungoogled) Chromium, Strawberry (music player) and a foot terminal open, I store 1.1G of data in 341.4M of RAM. This means the compressed data takes up only roughly a third of the original size. Because my (sad, pathetic DDR3) memory is a lot faster than my SSD, swapping to Zram is still faster and ends up giving me a better experience than swapping to disk.

...And don't get me started on when I still had an HDD in this thing, it was unusable to swap to disk 🫠

#3: Minimize the amount of services that are running

Distributions like Fedora or Ubuntu come with lots of services out-of-the-box, which is great to enable all kinds of functionality, but can also hurt you if your goal is e.g quick boot times or trying desperately to save on memory so you can have 2 Firefox tabs open

Initially, I had installed Arch Linux on this laptop, which was actually the first time I ran mainline arch and not some downstream distribution. I then later (and now, still) installed Void Linux and have been using that since.

One nice thing about Void is that, generally speaking, when installing a package any services that it ships are not automatically enabled. So, instead of installing e.g sddm and all the dependency services automatically get enabled, I can decide myself what to enable and what not. This is how I ended with a total of just 19 system services, and 7 user services:

Alt Text: A listing of all system and user services on my system. System and users services in order, with system and user separated by a semicolon and starting with system services: NetworkManager, agetty-tty 1-6, avahi-daemon, cupsd, dbus, docker, nanoklogd, polkitd, sddm, socklog-unix, turnstiled, udevd, zramen; dbus, gomuks, pipewire, pipewire-pulse, syncthing, turnstile-ready, wireplumber

All things considered this makes my system pretty lightweight, although I could still probably strip back some of them by not having e.g gomuks autostart, or by not using plymouth and instead just living without a boot splash – although it has barely made a dent in boot times, if at all.

#4: No Firefox for Low-Memory systems

Despite it's reputation, in my experience Firefox tends to perform worse on low-memory devices, like this laptop which only has 4G of Non-Upgradable DDR3 RAM.

Especially with it's “Memory Saver” option enabled, modern chromium blows Firefox out of the water in terms of performance, and especially keeping that performance as you open more tabs.

This is disappointing to me personally, as I generally prefer running Firefox or Librewolf, but this is a compromise I've now come to terms with, and is also why I'm writing this very post in Ungoogled Chromium and not Firefox.

#5: Set up an Out-Of-Memory mechanism!

With so little memory to work with, you are inevitably going to hit some limits when running multiple things at once. Especially if you try and run e.g Discord and a few browser tabs next to one another you will suffer issues with hitting the high 80's or even 90-95% memory usage.

As such, it is VITAL that you set up some sort of OOM Mechanism. Whether that is in-kernel, or something like systemd-oomd, or another mechanism. It really does not matter, but to recover from low-memory situations something like this is worthwhile to have.

#6: Choice of Desktop matters

Generally, I've learned two things: 1. Try and avoid GNOME. The GJS runtime is quite heavy, together with the shell in general. Especially if you have a slow GPU you'll probably rarely hit 60 fps 2. Wayland desktops or compositors tend to perform better than X, even on hardware that doesn't do (full) Vulkan, or only Vulkan 1.1 – e.g you can't use any of the fancy vulkan renderers of these compositors.

This does limit you, but if what runs well on this shitty laptop is anything to go by, this is probably a good approach to take.

Funny how I am saying this whilst running a full KDE Plasma 6 desktop, which you would think is not particularly lightweight either, but it runs much better than GNOME or even some X desktops I tried. The only thing that was quicker on this so far was some standalone Wayland compositors, e.g sway or niri

#7: Try to stick to native apps where possible

Sure, native apps can be slow too, but with RAM this limited it means having multiple instances of Chromium/Electon will quickly slow your computer to a halt.

Pretty self-explanatory overall. Something that uses a native toolkit will generally be more lightweight, as a rule of thumb.

#8: Proton-Sarek and PROTON_USE_WINED3D=1 are your friends

If you too do not have Vulkan 1.3 support, or even your VK 1.1 support is broken, not using vkd3d-proton or using something like Proton-Sarek will definitely be your friend in trying to get games running

I've successfully played through Half-Life 2, Portal and Portal 2 on my shitty laptop with the Linux native builds and openGL, and then later played the windows builds through Proton-Sarek (with...DXGL? Which is a thing apparently?) as most Workshop content for Half-Life 2 does not run under the Linux builds beyond just swapping the viewmodel of some weapon.

Generally speaking there is options to get things running on crappy devices, you just gotta know where to look :P

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3

However, all of these projects in the field generally have stuck to one goal, which I feel is short-sighted – even as someone that helps with a tumblr copycat – everyone just kinda goes for “Open-Source/Federated alternative to [Social Media Site]”

That in of itself is great! I can applaud an effort at making more ethical approaches to e.g microblogging or video streaming viable, or even possible to begin with. I also understand where it comes from: It is much easier to take an existing concept and adapt a new one – like Federation – on-top of it. But in the grand scheme of things I feel this has limited us quite a lot on the Fediverse as a whole, in terms of new ideas for us to try out.

For instance, even when we just limit ourselves to microblogging, the fact that we are trying to neatly fit into certain boxes already limits us; Especially in the context of federating with Mastodon, we routinely have to make design decisions we otherwise would not have made, had Mastodon taken a different approach to their software.

One such example is what you are reading right now. Writefreely, the software this blog uses, has activityPub integration, meaning these articles are not just stuck on paper.wf, but can also be viewed from many other pieces of software that support ActivityPub. Of course, AP is a little (very...) messy, and what “supporting ActivityPub” means is very vague, but generally speaking you can hit up a Mastodon instance, a Misskey instance, an instance running one of the many misskey forks, or different software altogether like Wafrn.

But even just then, Mastodon has weird behavior around how it handles Articles (so, usually long-form posts like this one)

Instead of showing the actual post and, idk, truncating it with a “read more” button, it instead takes the URL of a post and it's title and displays that!

You can see this in action by visiting this very blog on Mastodon and Wafrn respectively: – @alexia@paper.wf">https://mastodon.social/@alexia@paper.wf (this might redirect you...sorry!) – @alexia@paper.wf">https://app.wafrn.net/blog/@alexia@paper.wf

Constraining ourselves to these neat, tidy boxes of what Social Media can be, and what we can do with our standards means that we end up in situations where our supposedly agnostic way of sharing and displaying data suddenly gets handled in very different ways. Another good example is Misskey's Markdown language, MFM

MFM allows you to do more complex formatting for your text, including operations like scaling or translating an element, overlapping things, animations and more. All of these only being supported by Misskey and co., completely unsupported on anything else; It has become standard practice to outline the fact you're using MFM in a content warning

The point I am trying to make here is that despite our supposedly agnostic nature, and our chance on building features that can be compatible with one another whilst at the same time not being limited to one website or 'instance', we rarely actually make use of that and instead – for one reason or another – try and stick to tidy boxes of what “microblogging” is, or what “video streaming” is, or what “blogging” is and so forth.

I really think Fedi could benefit from just...getting a bit more quirky with it, trying out an entirely new idea on-top of ActivityPub and the infrastructure we've built over the years, without being constrained into what one or two pieces of software decided 'The Fediverse' should be. Especially in the realm of microblogging.

↓ it'd mean a lot if you supported my work ↓

https://liberapay.com/cyrneko https://ko-fi.com/cyrneko Any little bit helps <3