@Cătă @Shauna GM There are also the "descendants" of Friendica, created as forks by Friendica's own creator over more than a dozen years now, which have even more advanced permissions systems.

One is Hubzilla, a fork of Friendica from 2012. The other one is called (streams) by the community, and it's a fork of a fork of three forks of a fork (of a fork?) of Hubzilla from 2021.

I've made a series of tables that compare Mastodon, Friendica, Hubzilla and (streams) in a number of categories.

Hubzilla and (streams) give you the following possible target audiences for new posts:

• everyone on the Internet (Hubzilla, (streams))
• all your connections ((streams) only; on Hubzilla, you can emulate this with a privacy group containing all your connections)
• only the members of one particular privacy group (Hubzilla)/access list ((streams)) (think "Mastodon's lists on coke and 'roids")
• only those to whom you've assigned a specific custom profile of your channel (Hubzilla; like Friendica, it actually allows you to have multiple profiles and show them to specific connections so that they see different sides of you)
• only one specific group/forum (Hubzilla, (streams))
• only an impromptu selection of connections of yours (Hubzilla, (streams))
• only you yourself (Hubzilla, (streams))

The permissions systems of Hubzilla and (streams) are compatible to one another, i.e. one understands the permissions defined by the other. I'm not sure how well they and Friendica play with each other, though.

As for Mastodon: Any post that isn't public is understood by Mastodon as a DM and treated as such. Your contacts can't boost it, for example. The downside is that this kills any chances of meaningful discussions for Mastodon users.

So Hubzilla and (streams) have this advanced permissions system, and they understand threaded conversations and treat each one of these as an enclosed object with exactly one post and any number of comments. A thread always has and enforces consistent permissions all over, including all comments.

If you're on Hubzilla, and you send a post to Alice on Hubzilla and Bob on (streams), not only do both see your post, but both also automatically receive each other's comments, and they can comment on each other's comments.

Mastodon understands your post with restricted permissions as a DM. But Mastodon DMs only ever happen between two actors. This means: If Alice and Bob are on Mastodon, then both receive your post as a DM, but neither receives any comment from the other, and they can't comment on each other's comments either.

Also, fair warning ahead:

Neither of them is "Mastodon with some extra features". They're all very different from Mastodon. They all have steeper learning curves than Mastodon. Friendica's learning curve is significantly steeper than Mastodon's. (streams)' learning curve is quite another bit steeper than Friendica's because the permissions system is not optional, and not everything is public by default. Hubzilla has an even steeper learning curve.

Also, none of the three has a full set of dedicated native mobile apps. For Friendica, there are basically only Android apps. In the Apple App Store, there's nothing. Friendica can also be used with some Mastodon apps, but they only cover maybe 20% of Friendica's features, namely those that Mastodon has, too, so you'll be very, very limited. In fact, they do not cover any permission settings.

For Hubzilla and (streams), there are no phone apps at all. They don't support Mastodon apps either, and they never will. It simply wouldn't make sense because a Mastodon app would not cover important key features.

So if you're on an iPhone or iPad, or if you want to try Hubzilla or (streams), your only option is the Web interface, either in a browser or as a Progressive Web App. At least, all three have Web interfaces that adapt to mobile devices.

Lastly, you won't find (streams) on FediDB or Fediverse Observer. It's intentionally kept away from there, and it intentionally does not submit any stats. There are currently only two public, open-registration instances anyway, one in the USA, one in Hungary with a German admin who also speaks English.

#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Friendica #Hubzilla #Streams #(streams) #Facebook #FacebookAlternative

@Anna Maier I don't know what constitutes a "good" example in your opinion, but I've got two examples of how bad AI is at describing images with extremely obscure niche content, much less explaining them.

In both cases, I had the Large Language and Vision Assistant describe one of my images, always a rendering from within a 3-D virtual world. And then I compared it with a description of the same image of my own.

That said, I didn't compare the AI description with my short description in the alt-text. I went all the way and compared it with my long description in the post, tens of thousands of characters long, which includes extensive explanations of things that the average viewer is unlikely to be familiar with. This is what I consider the benchmark.

Also, I fed the image at the resolution at which I posted it, 800x533 pixels, to the AI. But I myself didn't describe the image by looking at the image. I described it by looking around in-world. If an AI can't zoom in indefinitely and look around obstacles, and it can't, it's actually a disadvantage on the side of the AI and not an unfair advantage on my side.

So without further ado, exhibit A:

This post contains

• an image with an alt-text that I've written myself (1,064 characters, including only 382 characters of description and 681 characters of explanation where the long description can be found),
• the image description that I had LLaVA generate for me (558 characters)
• my own long and detailed description (25,271 characters)

The immediate follow-up comment dissects and reviews LLaVA's description and reveals where LLaVA was too vague, where LLaVA was outright wrong and what LLaVA didn't mention although it should have.

If you've got some more time, exhibit B:

Technically, all this is in one thread. But for your convenience, I'll link to the individual messages.

Here is the start post with

• an image with precisely 1,500 characters of alt-text, including 1,402 characters of visual description and 997 characters mentioning the long description in the post, all written by myself
• my own long and detailed image description (60,553 characters)

Here is the comment with the AI description (1,120 characters; I've asked for a detailed description).

Here is the immediate follow-up comment with my review of the AI description.

#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #AI #LLaVA #AIVsHuman #HumanVsAI

@Peter Vágner @Dieguito How conversations work is not unified all across the Fediverse. Even how connections work is not unified.

Mastodon has taken over the follower/followed principle from Twitter which is always illustrated with arrows with one point. A following B is illustrated with an arrow from A to B. A being followed by B is illustrated with an arrow from B to A. A and B following each other mutually is illustrated with one arrow from A to B and one arrow from B to A.

It appears to me that Friendica has adopted this to become more compatible with Mastodon. But its several descendants, created by Friendica's own creator, starting with Hubzilla, haven't.

Hubzilla, (streams) and Forte still have the bidirectional "connection" or "contact" as the default. It's illustrated with one arrow, but with one point on each end.

Also, all three understand a threaded conversation as an enclosed contruct entirely owned by the conversation starter. Everyone on these three who has the start post on their stream always actually has the whole thread on their stream.

In fact, all three have Conversation Containers implemented. This feature was originally created in the streams repository in 2022. Forte has had it from the get-go as it started out as a fork of (streams). It was eventually turned into FEP-171b and backported to Hubzilla last year.

All three make sure that everyone who has a post on their stream also always has all comments on that post, at least those that are made after they have received the post.

This works on two basic principles:

• All comments go directly to the original poster because the original poster owns the thread.
• Those who have the post automatically receive all comments from the original poster.

In a pure Hubzilla/(streams)/Forte system, your above example would look like this:

• User 1 and User 2 are connected.
• User 1 and User 3 are connected. (This doesn't even matter.)
• User 2 and User 3 are connected.
• User 2 and User 4 are connected.

Much simpler than explaining everything with "following" and "being followed", isn't it?

Now, the conversation works like this.

• User 2 sends a public post, thus creating a Conversation Container of which they are the owner.
  User 1, User 3 and User 4 receive the post.
• User 3 comments on User 2's post.
  The comment goes from User 3 to User 2, who is the owner of the conversation, and it is automatically forwarded to User 1 and User 4 who already have User 2's post on their streams.
• User 4 comments on User 3's comment.
  The comment goes from User 4 past User 3 straight to User 2, who is the owner of the conversation, and it is automatically forwarded to User 1 and User 3 who already have User 2's post on their streams.

The only mentioning that occurs here, if any, is User 4 mentioning User 3. This is not necessary for User 4's post to reach anyone. This is only necessary to make sure on Hubzilla (which doesn't have a tree view) that User 4 is replying to User 3's comment and not to User 2's post.

On Mastodon, for comparison, everything depends on who follows whom, who mentions whom and whose instance knows whose instance.

#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Mastodon #Friendica #Hubzilla #Streams #(streams) #Forte #Conversations #FEP_171b #ConversationContainers