The sequel with Neil deGrasse Tyson is also really good.

It’s not social media that did it. It’s monopolistic, unregulated, greedy, giant tech corporations that made the internet shitty.

I actually would really love to hear how “right to be forgotten” applies to an email you’ve sent. I mean you can’t force anyone to delete an email you’ve sent to them, so how does right to be forgotten even apply for emails?

The fediverse would work in the same way, I think.

No. In fact, ActivityPub has no general mechanism for even knowing where content has been distributed to. So when you ask your instance to delete something, it can’t actually know what other instances to ask to delete the mirrored content.

Mastodon tries its best by sending deletion requests to all known instances, in the hope that that will reach all instances that have fetched the content. But in fact, instances that are unknown to your own instance could have the content as well, though this is probably a very rare occurrence.

Bottom line: Don’t write anything on the internet that you don’t want publicly displayed. Anyone can save it and then you can’t force them to delete it. That applies to the entire internet. It also applies to the fediverse.

This also presumes mods are, by default, inherently non-biased, held to a standard, and never have vendettas of their own.

Of course mods are not always like that. But if mods are like that, just go to another community. If mods are bad, just leave. On the fediverse, you “vote” with where you participate.

Yea it’s still only a partial solution. Even those feeds could get very active over time (we can hope 😅). The way Piefed implemented feeds is interesting but seems almost overengineered? Sharing feeds could have been done via a simple query parameter I feel like.

Yes, but that doesn’t scale. If there are thousands of comments being submitted constantly, the All feed would just be a new page every time you refresh for the new comments sort. It would be chaotic.

It should instead be based on a recent rate of comments for instance. Much like normal votes but comments instead and not based on the age of the post.

Sort of, but doesn’t it just sort by the latest comment? I.e. any thread would be bumped to the top by a single comment? I might be wrong. But that makes it kind of less than ideal if true.

The solution is not to build this yourself. If you are sitting and building features yourself for search, stop. Use a dedicated search database instead.

I honestly personally preferred Reddit’s sorting algorithm. Lemmy’s algorithm is a bit too slow to update for my taste. This is kind of part of Lemmy’s design though. My problem with Reddit was never it’s sorting algorithm (honestly that was a big part of its strength!), it was just all the ways they enshittified later on.

Yea, the best solution is:

• Keep votes “semi-public” (visible to mods/admins) to aid in moderation and avoid vote manipulation.
• Make it very clear to users that votes are not private.

As long as users know that votes are not private, it should be okay.

Well, that’s the nature of link aggregators. Lemmy’s and Reddit’s style is a link aggregator, not really what you would consider an old-fashioned forums. It’s a different sort of use case with different pros and cons. A con is that you don’t get these super long lived threads cause they disappear in the stream of new threads. A pro is that… you don’t get these super long lived threads cause they disappear in the stream of new threads. :P

Private message here on Lemmy or just a comment here is fine :)

Thanks for the clarifications and thoughts!

after seeing this, I do not want to get my hopes up… 😂😂😂

That’s fair haha. I definitely understand that. Building open source software, or really any software, is frought with possibility of failure. I don’t claim to be the survivor who will get through it all, that would be incredibly naive of me to claim. I’m just trying, just like others have tried before me :)

optimal management of Activitypub groups

I definitely plan to support groups. Do you mean anything in particular with “optimal” management? I mean what would be “suboptimal” management? Do you just mean an incomplete implementation?

optimal management of the DB: Piefed has made great strides, while this remains the Achilles heel of Friendica

Are you mostly talking about performance here? Or how do you mean?

Definitely agree images are important. I honestly question the value of RSS feeds. It’s not something “normal” users use. It’s very much a techy thing. I don’t know any non-technical person who even knows what RSS is or has ever even heard of it.

a registration system that combines the features of Mastodon and Lemmy

I plan to support applications and/or invitation trees (like lobste.rs uses). But more could be added I suppose. What features are you thinking of?

Definitely agree moderation is important. I want to ensure moderation is well supported with good moderation tools.

if the system allows formatted posts, a formatting bar is necessary (Lemmy is a great model). Then we should figure out whether to use Markdown or a more precise simplified HTML (BBcode is not the best)

I think you need a WYSIWYG editor. That’s what normal people expect.

I’d like to support polls and events. As said before, I want to support all kinds of content ideally.

circle management (like those of Friendica)

What is a circle in Friendica? I’m not familiar.

I’m planning and architecting the system to support filtering and good search functionality. Everything is accessible via the backend API (how else?).

I hope it’s clear that this is a “methodological distrust”, but I’m rooting for you!

Honestly not really clear - what do you mean with “methodological distrust”? What method would be trustworthy? :)

A Fediverse project really needs a team. One-man-show projects are too risky and emotionally draining

I agree, bus factor is a problem. But I feel like projects like this are very hard to start without starting as one person. I mean it’s hard to gather people around something without having anything to show at first. I’m hoping to establish something and then attract people who might be interested to contribute.

Can I ask you to create a Lemmy community or a Friendica group? It would be nice to discuss in one place

I feel it’s too early for that (again, don’t want to shout about it yet). But eventually I would definitely like to do that.

Hi, thanks a lot for your detailed message!

I totally understand the lack of faith - I mean I’ve shown nothing to earn any faith so that’s completely fair. I also share your frustration with existing apps that have shown to not improve or be good enough. That’s part of the reason why I wanted to try my hand at it myself. I feel that the status quo is not good enough and I believe in the mantra that “if someone else is doing something that you think you can do better, you should do it”.

forumverse (or threadverse) projects like Lemmy have received the main damage precisely from the incompatibility wanted by Mastodon against them;

Not sure what you mean with “damage” here, but my plan is to support all kinds of ActivityPub content, both the microblog stuff that Mastodon is known for, the forum stuff that Lemmy does and anything else from other apps. I don’t want my app to feel limited like Mastodon or Lemmy. Mastodon is very microblog-focused, Lemmy is very forum-focused. I want something that can do both and more. In some ways, this makes it harder, in other ways it makes it simpler. For instance, Lemmy makes a difference between “posts” and “comments”; they are not the same thing in the database. But in my app, comments are just another post, much like how posts work in Mastodon.

new very interesting projects like Bonfire (the only software together with Friendica and Hubzilla to manage the “circles”) are being developed with difficulty and are made up of many modules and above all do not have a decent app.

I’ve heard of these projects, but haven’t studied them in detail. I find bonfire especially confusing. I can’t seem to grok what it is - is it a server, or a framework for a server, or an app? For instance there’s this app but the code link goes nowhere. There’s also this repo with commits that look super weird. Honestly just confused about it. Anyway.

I agree that having good mobile support (including an app with great UX) is super important.

Finally, the impression I had is that even among the most famous developers of the Fediverse there is a bit of ignorance about Activitypub, about other platforms and about how other developers have solved the same problems; also it seems that the “Masters of the Fediverse” are always in a bad mood and have less and less desire to learn new things (a praiseworthy exception is Matthias Pfefferle).

I’ve tried to learn a lot about ActivityPub and I understand it fairly well at this point I would say. I’ve participated a bit at https://socialhub.activitypub.rocks/ where ActivityPub is discussed at length. I’m not sure what “Masters of the Fediverse” refers to but I definitely am a curious soul and I think continuous learning is super important :)

Creating a federated software is therefore not a very simple thing neither technically nor psychologically, but if you feel capable of doing it, perhaps it could be advisable to test yourself a bit:

• developing some web utilities, some plugins or less ambitious projects
• actively contributing to other existing projects (Friendica? Bonfire?)
• getting familiar with both the Mastodon API (which is an industry standard) and with the development and definition of APIs in general: when someone wants to write an app for your software, they will look at your code and in two minutes they will decide if it is worth doing!

I appreciate your concern, but I am a professional software engineer so I’m not so worried about the scale of the project. Rest assured, I have worked on very large projects professionally and built plenty of things in side projects, most of them related to the web. I also administrate Feddit.dk so I have experience with hosting a Lemmy instance and all the complexity that brings.

I particularly enjoy Rust, and I did actually look into contributing to Lemmy (since that uses Rust in the backend) at first before I started my own project. Unfortunately, Lemmy’s code is… not where I would like it to be (both of Lemmy’s main devs learnt Rust while working on Lemmy, and it unfortunately shows in the code quality), and the direction of Lemmy is not the direction I want to take my project, as stated above. I want something more general than a Reddit clone, though it will be inspired by Reddit/Lemmy in some ways (I plan to use up/down votes to sort content, for instance).

I have no interest in contributing to Friendica, as the direction seems bad, as noted in the post above. Besides, it’s PHP and I really don’t want to touch that. Hubzilla is also PHP and seems much to technical for general users, so once again not viable. Bonfire seems to be Elixir which I don’t know either, but again I am super confused about what Bonfire even is. All these reasons and other reasons are why I wanted to do my own project.

I don’t agree with you that the Mastodon API is an “industry standard” - it may be widely used, but Mastodon is continually forcing its own ideas of the Fediverse on the rest of the ecosystem, which I don’t like and is something that is often bemoaned on https://socialhub.activitypub.rocks/. But rest assured that I am very familiar and comfortable with APIs (again, professional software engineer 🙂). I care about documentation a lot and from the start, my prototype backend has exposed its API via an OpenAPI specification so that clients can be easily generated. I’m actually about to use this OpenAPI spec to generate a client myself as I start work on the frontend 🙂.

Again, thanks a lot for your thoughts and attention! If you have any concrete feedback on the UI and/or UX of Lemmy, Mastodon, Friendica or other apps, I’d love to hear it, as I’m starting work on the UI for my own frontend these days. For instance, any favourite UI of a fediverse app, any preferred features or any common mistakes or pitfalls that should be avoided, if you have any thoughts along that direction.

I think “hundreds of thousands and even millions” is a bit of a stretch. Wikimedia’s annual report mentions donors at a level of “$50,000+”, and I’m guessing most of those are probably closer to 50,000 than to 100,000. Tbf I suppose that’s over just one year, so perhaps your statement isn’t entirely inaccurate.

I suppose this community is as good as any. But it’s difficult to talk in general about this as each fediverse app has different performance needs/characteristics, so I’m not sure if you can extrapolate anything in general. But perhaps?

Why shouldn’t the donation model keep working? Wikipedia works on donations, why can’t the fediverse?