Hi Fediverse, I’ve recently been interested in the ActivityPub protocol as well as the programming language Elixir. To better learn both technologies and simultaneously get a product out of it I’ve thought about making (yet another) federated social network.

The main ideas that would set it apart from other federated networks are the following.

  1. It should be easy to use for people who are not tech savy. It should appeal to local communities, like a sports club, as a platform to organize and .

  2. It should be a platform where communities can organise, share discussions and create events. I know many people who use Facebook primarily for this kind of functionality. I hope this could be a gateway for smaller local communities to move over to something more free and open.

  3. Users should be able to categorize their feeds into nested lists, like traditional RSS readers. The idea is that you could subscribe to your favorite PeerTube creators in one feed, and the blogs you follow in another, even with subcategories. You might also make a feed for Lemmy communities you want to follow.

I would love to hear what thoughts you have on this. What do you think? Do you think the fediverse is the right technology for this? Am I focusing on the right audience? Do you have comments on the attached mockup design?

I would appreciate if you shared this post across the fediverse, such that not just Lemmy users could be a part of the discussion.


For people who are interested in my project Photoview don’t worry, I love to work on that and I will continue to do so.

Jorge S.
link
22M

The mockup looks really nice and I think it could work if you keep it simple. Something I miss in the fediverse is the possibility of categorizing posts. For example, the user could select a category like work, food, politics, academic, others when posting. I am not sure this will work but something to think about.

@viktorstrate@lemmy.ml
creator
link
22M

Thank you for the nice feedback, feed categorizing is exactly what I was thinking about with the Feeds list to the left. I’m glad to hear that I’m not the only one who miss/want this feature.

weex
link
32M

Hey viktorstrate, My current operating principle is that a community is foundational to the success of a software project. Building community is therefore the first thing to focus on when starting something like this. To that end, I’m looking in your post for points around which a community could rally.

1/Easy-to use-is good, yes but it’s kind of a given for all software. Some prioritize more than others but nobody wants their stuff to be hard to use.

2/Organize, share discussions are similarly common however an event focus is more unique. Mobilizon is working on this as well but I haven’t seen that yet as a strong core problem around which the rest can be built. I think it’s better added as a feature or plugin to an existing and strong software project.

3/Lists exist in Mastodon but not nested lists. I suppose once you have more than a handful of lists, you’ll want more structure and so I think that might be a good feature, but again better added to Mastodon or a fork than something to build from scratch.

In terms of greenfield projects, I would think a TikTok-clone would be interesting. That’s heavily dependent on a mobile app with strong camera, creative and video editing editing features but it would be nice to have somewhere in the fediverse to point that fastest growing segment of social.

Getting back to community, how do you feel about the strength of the photoview community and how do you think you might improve on what was done there in this new effort?

@viktorstrate@lemmy.ml
creator
link
32M

Hi @weex@lemmy.ml, I think you’re right that a community is very important, and thinking about what kind of community will use it from the beginning is important, such that one doesn’t build a platform for a community/segment that doesn’t exist.

1/Easy-to use-is good, yes but it’s kind of a given for all software.

You are right, but when I say that I mean that it should appeal to an average user. For example, I see many federated platforms that have a thousands settings and customization options and try to do many things at once, that certainly works for some people but as a new user you can quickly be overwhelmed.

I think Pixelfed is doing a great job in this regards, the UI is familiar and simple.

2/Organize

You might be right in that it doesn’t feel like there’s a huge demand for this, but maybe that’s because the right medium doesn’t exist yet?

I feel like the problem with Mobilizon it only does events, no groups or posts etc. I don’t see that big of a need for that, especially not federated.

But maybe I sould focus on the core federation experience first, like nested lists, great integration with Mastodon, Pixelfed and Lemmy. Then move the focus from there, I also might understand the problems and needs better at that point.

3/Lists exist in Mastodon but not nested lists.

Yes, I really like Mastodon lists, but I feel that they are too simple that I would actually use them that much. For example to add a user to a list, you first have to follow them, then go to the list page, click edit lists and search for the user to add them.

Also there are no way to have someone show only in a list and not in your main feed, like on Facebook you can be friends with someone, but hide them from your timeline. I think this would be an especially needed feature for the fediverse, because in many cases if you don’t follow someone from another instance, their posts won’t synchronize to your instance.

In terms of greenfield projects, I would think a TikTok-clone would be interesting.

I’ve heard this proposal before, but going back to thinking about communities, I don’t think see a community for this. The people who are interested in fast pased, AI autosuggested videos, are not the people who are interested in a federated network. I simply see no benefit for federation here.

Instead I think federations shines for large discussions, where people comming from different places and communities can be a part of the same discussions. On a side note, I would also like to see more blogs and websites integrate with ActivityPub but not allow user registration, such that you don’t need to have an account on every single small website, just to leave a single comment.

Getting back to community, how do you feel about the strength of the photoview community and how do you think you might improve on what was done there in this new effort?

This is a good question. I love the Photoview community, we have a Discord server with 250+ users and a lot of constructive discussions on Github. I have also worked with EmbassyOS to integrate it into their platform.

I think it was much easier to build the community around Photoview, since there’s already a passionate community around self-hosting and Photoview directly targets people who want to host it themself. But the community around a social network shouldn’t mainly be the instance maintainers, but instead the users, and I don’t have a clear idea of how to target and grow that kind of community yet.

I feel like the Photoview community hasn’t been growing by adding new features, but rather by things like helping people with installation, writing documentation or by fostering discussions on Github.

The project you’re describing looks a lot like Hubzilla or Diaspora. And Hubzilla could certainly be helped with some design volunteers :)

@viktorstrate@lemmy.ml
creator
link
12M

Thank you for sharing. I have looked at Diaspora, but not much at Hubzilla. I’ll make sure to check that out, at least to get inspiration. The reason why I am thinking about building a site from scratch is that many existing projects try to do too much, or are doing somthing way different than what I want to achieve. I have looked at some potential candidates (Mastodon, Pleroma, Pixelfed) to build on top of, but none of them seem suitable to me.

Well maybe if you can write down what you want to achieve, that’ll be a good start for seeing how compatible these ideas are with other projects. Personally i don’t use it, but i think Hubzilla has got good principles: a single site/identity to publish all your federated content, with support for several protocols. You should really try it out :)

@X51@lemmy.ml
link
22M

My biggest problem with social media is that if I follow someone, I’m not interested in seeing everything that everyone posts in my feed. I want to be able to select the users separately and view their specific posts when I’m in the mood. I don’t want everything jumbled together. I usually don’t follow anyone and just click on their profile instead. That’s disheartening to the people who do post things that interest me. I’m looking forward to pixelfed introducing groups. Being old school in my tastes , I prefered the old usenet format better and using a feed reader. Spam (& google) ruined the old usenet format.

@viktorstrate@lemmy.ml
creator
link
12M

My biggest problem with social media is that if I follow someone, I’m not interested in seeing everything that everyone posts in my feed.

This really resonated with me, and I feel exactly the same way, especially for the fediverse since the content can be so diverse. Much like when I put on music, I often put on an Album rather than a playlist consisting of completly different genres and moods.

And I think this is a problem that hasn’t been fully solved yet on the fediverse.

I’m looking forward to pixelfed introducing groups.

I will also be following that tightly, and probably build something very close to their solution. Pixelfed will be a big inspiration for the project as their goals are much aligned with what I’m trying to achieve, but with another focus being images.

Being old school in my tastes, I prefered the old usenet format better and using a feed reader.

I have used usenet very little, I had a subscription for a short while but never got the hang of it, I might not be old enough to have caught that train. But maybe I should look more into that as a source of inspiration.

@X51@lemmy.ml
link
22M

The usenet was essentially public email without all the HTML baggage. You searched for a group that interested you on public news servers. You subscribed to it and immediately interacted with people who shared your interests. The groups had a moderator to keep things on track. Typically, moderation was not a problem and I didn’t even know who the moderator was. Deja news archived the old posts and you could follow old threads to know what had been written before. Eventually, spam overtook the groups and it was too much to avoid. There were pay services, but most ISP’s had a news server and you could just follow the public groups for free.

I do very little to entice subscribers to follow me and I manually delete old posts after I felt people had time to review them and comment. I don’t want 20 years of my life online. People grow and change over time. Despite that general tendency, I’ve grown followers on Pixelfed at a faster rate than I have on most Social media sites. One person said my posts were filling up his feed and that’s why he hadn’t followed my account. That’s another reason I feel that following someone should not dump everything into your main feed. Following someone should just make it easier to access their posts. It should force it to be seen.

My song playlists had themes if I listened to one. I’ve owned a lot of albums that were 50% great songs and 50% duds. I do listen to albums if I know the album is consistent throughout.

I do like the complete randomness of Mastodon, but I have a huge blocklist that probably ties up a lot of system resources in the background. Misskey is very interesting, but I feel that some of it is broken and some of it is not very intuitive (such as expanding a thread.)

Travis Skaalgard
link
22M

Honestly I really like this concept, but the other comments are right - there aren’t a whole lot of reasons people would switch from things like Facebook to this unless they were already onboard with hating Facebook.

@humanetech@lemmy.ml
link
1
edit-2
2M

You and @jedrax@lemmy.ml are right, but also I believe that we should start to think more in a different direction on the Fediverse. The biggest strength that flows from its decentralized nature, if you come to think of it, is that the fedi AS A WHOLE should be competing against the traditional social media giants. By overly focusing on individual apps, and seeing them as full-service platforms, we overlook this strength. There’s slim chance that a small FOSS-values driven non-commercial project will ever be truly competitive. But the Fediverse in its entirety could. The network effects of fedi compound to something that’s worth cnnsidering. The total experience of what is means to be ‘a fedizen’ is then important. That experience will never improve much if there’s just single apps without much interop between them, though. Note that this fedi strength is a weakness for the moloch platforms, who have to protect their castle with everything they have, locking people in, while we are essentially coop-based.

Jedrax
link
22M

I like to see people still working on social media sites, but you have to ask yourself, why would everyday users switch away from Facebook to do this? I’m not talking about the idealized aspects like privacy, daddy zuck = bad, etc. But on a practical level, it would require a lot of time and coverage to be a contender

Randoom
link
22M

the ui looks nice and good.

Halce
link
1
edit-2
2M

I think in terms of events, a useful idea would be to have the option to set an event as ongoing (i.e. non-time-expiring) for the purposes of launching like an ongoing charity drive, or a ‘marketplace’ for giving people in your local community used stuff you no longer need yourself (a family member had such a ‘second-hand’ marketplace for our town setup on FB and it went on constantly for a few years).

Vegafjord eo
link
12M

Yes, we should first and foremost focus on making these solutions appealing to organizations.

@nutomic@lemmy.ml
link
12M

Sounds a bit similar to Smithereen, which is inspired by VKontakte.

https://friends.grishka.me/grishka

@viktorstrate@lemmy.ml
creator
link
02M

Thank you for sharing, I try to look as many different existing platforms to learn from them. When I visit the link it’s mostly in Russian, and the official website (from link in footer) gives me a 403 forbidden error.

Do you know where I can learn more about it, preferably in english?

@nutomic@lemmy.ml
link
02M

Here is the code, but theres not much info either. Probably best if you talk with @grishka@mastodon.social directly.

@humanetech@lemmy.ml
link
22M

You can also approach @grhiska on SocialHub where he’s a frequent poster, and also wrote some Fediverse Enhancement Proposals (FEP’s) to give feedback on. SocialHub is in general a good place to discuss technical issues related to federated apps.

@humanetech@lemmy.ml
link
02M

I love the idea, and am advocating myself for some time to create better abstractions of “Community” in the fediverse than we have now. I call it the “Community has no Boundary” paradigm, where a community goes well-beyond a simple group of members. Instead it can have a rich set of Relationships to other actors (Person, Organization, Group, etc.), some standardized, some app-specific. “Community” as a concept might standardized on the fediverse, where a server exposes one or more Group actors. In the current fediverse the community is implicit and defined by the “instance boundary”. Lemmy is among the first apps that slice things differently and expose many communities with the recent federation update. I wrote more on this in Standardizing on ActivityPub Groups and Standardizing on a common Community domain as AP extension?

The user-friendliness is very important indeed. I see a lot of people comparing to Facebook, and I guess you’d compete with them from functional perspective. But I couldn’t tell as I’ve never used FB. The FB walled garden excludes people that have no account. Maybe your app can shine here, by allowing integrations to the websites that these local communities have. Like GoToSocial the app may be a back-end service, where different UI’s can consume the API.

@viktorstrate@lemmy.ml
creator
link
0
edit-2
2M

Thank you so much for your work on this and for sharing it here. Much of the interpretation of the ActivityPub specification has been quite clear to me by reading the W3 spec and looking at implementations.

But I currently only have a vague idea of how groups should work and since it will be an integral part of the platform I want to get it right. I haven’t read the threads you linked yet, but I will make sure to read them carefully.

Maybe your app can shine here, by allowing integrations to the websites that these local communities have. Like GoToSocial the app may be a back-end service, where different UI’s can consume the API.

That’s a great idea, I hadn’t thought of that! Maybe building a small Wordpress plugin aswell since most of smaller communities tend to use that for their websites.

i would love a good looking alternative to fb, it looks nice, i dont think i would use it much because i dont have ppl in the fediverse, same reason i dont use mastodon tbh but maybe it could work better if there are groups

Pixelfed had groups since recently

A community dedicated to fediverse news and discussion.

Fediverse is a portmanteau of “federation” and “universe”. It is a common, informal name for a federation of social network servers whose main purpose is microblogging, the sharing of short, public messages.

Getting started on Fediverse;

For devs;

  • 0 users online
  • 3 users / day
  • 12 users / week
  • 60 users / month
  • 216 users / 6 months
  • 18 subscribers
  • 265 Posts
  • 2.09K Comments
  • Modlog