Skip to content

Piero Bosio Social Web Site Personale Logo Fediverso

Social Forum federato con il resto del mondo. Non contano le istanze, contano le persone

The open source single-user server software will be ready for #alpha #testing soon!


Gli ultimi otto messaggi ricevuti dalla Federazione
  • @rimu@mastodon.nzoss.nz Definitely. Offloading the static assets to nginx is a big win. Varnish adds a layer of serving from memory that takes it up a notch. Like having your own Fastly pop.

    It does require some configuration nuance to be sure you aren't serving cached assets to the wrong connections (e.g., authenticated GET requests that shouldn't be shared beyond a specific session).

    read more

  • @eyeinthesky

    Only as a metaphor.

    Federation happens between servers / nodes, not between networks.

    read more

  • Is it correct to say the and are "federated" by protocol bridges? I have similar question related to and and other bridged protocols. Given the is , what this larger federated social web called?

    read more

  • @silverpill@mitra.social Thanks - that solves a number of issues I've been encountering. From the outset, I wanted to use the system actor to point at the relevant administrative collections, but couldn't think of a good way to identify the actor to the client (without hard-coding it). That webfinger adjustment solves that.

    read more

  • @jdt

    >nodeinfo is a common protocol used for discovering information about ActivityPub-speaking servers

    NodeInfo is used primarily by sites like https://fediverse.observer, it is not intended to be used by clients. There is a similar entity in ActivityPub: server actor (instance actor). In your case it seems to be located at https://enigmatick.social/user/system ?

    I prefer Webfinger discovery method, as described in this FEP: https://codeberg.org/fediverse/fep/src/branch/main/fep/d556/fep-d556.md.

    read more

  • From its conception, #Enigmatick has leaned heavily on the /inbox and /outbox endpoints for client operations. There are some /api endpoints, but I avoid that were I can shoehorn operations into the #ActivityPub specification and #ActivityStreams vocabulary.

    While typical operational activities are fairly well accounted for, administration is a weak point. For example: I haven't identified a clear way to use the currently described mechanisms for an administrative user to pull up and manage instances or actors on a server.

    I've relied on CLI tools (e.g., ./enigmatick --help) to manage some of that. And in some cases, I know how to manipulate data in my database, so I haven't worried too much about building tooling. But I'd like to ship something that other folks can use to share in my efforts, so I've been thinking about how to model those activities in an ActivityPub-esque way to use in the Svelte UI.

    ActivityPub Messages

    To that end, I'm now using Block and Delete activities sent from the client to the server outbox to manage the blocking of instances and purging of data.

    { "@context": [ "https://www.w3.org/ns/activitystreams", { "ek": "https://enigmatick.social/ns#", "Instance": "ek:Instance" } ], "id": "https://enigmatick.social/activities/550e8400-e29b-41d4-a716-446655440000", "type": "Block", "actor": "https://enigmatick.social/user/system", "object": { "type": "Instance", "id": "https://spammy-instance.example" } }

    In practice, my client does not generate the id, but that attribute is generated by the server and the Activity is stored alongside other typically federated activities. These local Block activities are not federated out to other servers; they are intended solely for local server management.

    The Block activity is sent as a message signed at the client by a user with administrative privileges on the server. Enigmatick's user authentication is unique (i.e., I use a separate set of encryption keys for client-signing executed by a wasm module in the browser). That can be a topic for a future article.

    That the actor as the system Application user is important. That is used by the server to establish the scope of this action as system-wide, not just for a single user. The system actor is discoverable in the nodeinfo metadata.

    I'm using a typed object rather than just an id reference. This is so that I can use this same flow for blocking and purging Actor objects (i.e., the type would be Person, Service, or Application).

    The purge action is similar, using the Delete activity.

    { "@context": [ "https://www.w3.org/ns/activitystreams", { "ek": "https://enigmatick.social/ns#", "Instance": "ek:Instance" } ], "id": "https://enigmatick.social/activities/550e8400-e29b-41d4-a716-446655440000", "type": "Delete", "actor": "https://enigmatick.social/user/system", "object": { "type": "Instance", "id": "https://spammy-instance.example" } }

    The term, "delete" is a bit of a misnomer in this case as it applies to the instance specifically. The instance will remain, but the objects, activities, and actors associated with that instance will be fully deleted (i.e., not set to Tombstone).

    Collection Endpoints

    To facilitate the UI operations, I've created two new collection endpoints on my server: /instances and /actors. These endpoints provide typical ActivityPub Collection objects.

    { "@context": [ "https://www.w3.org/ns/activitystreams", { "Instance": "ek:Instance", "activitiesCount": "ek:activitiesCount", "actorsCount": "ek:actorsCount", "blocked": "ek:blocked", "ek": "https://enigmatick.social/ns#", "lastMessageAt": "ek:lastMessageAt", "objectsCount": "ek:objectsCount" } ], "type": "OrderedCollection", "id": "https://enigmatick.social/instances", "totalItems": 7702, "orderedItems": [ { "type": "Instance", "id": "https://example-instance.name", "blocked": false, "created": "2025-12-16T16:56:33Z", "lastMessageAt": "2025-12-16T16:56:33Z", "actorsCount": 0, "objectsCount": 1, "activitiesCount": 0 } ], "first": "https://enigmatick.social/instances?max=9223372036854775807", "last": "https://enigmatick.social/instances?min=0", "next": "https://enigmatick.social/instances?max=1765657395402834" }

    I've added some extensions in the @context to account for a few non-standard attributes.

    That collection is used by the UI.

    The Enigmatick instances UI showing the most recently discovered instances from the enigmatick.social server

    Collection Discovery

    nodeinfo is a common protocol used for discovering information about ActivityPub-speaking servers. I've extended my use of that to facilitate client-discovery of these new endpoints using the metadata object contained in the nodeinfo JSON.

    "metadata": { "actor": "https://enigmatick.social/user/system", "adminActors": "https://enigmatick.social/actors", "adminInstances": "https://enigmatick.social/instances", "domain": "enigmatick.social", "url": "https://enigmatick.social" } Final Thoughts

    As I'm reading through this, I see some opportunities for refinement. I should probably be using OrderedCollectionPage instead of OrderedCollection for my collection endpoints. I'm sure there are other tweaks to be made.

    read more

  • Agreed that forums are definitely needed, and the energy NodeBB has brought to the Fediverse has been very welcome indeed! The coexistence is often smooth but sometimes quite clunky (although of course that's true for ActivityPub platforms in general).

    Specifically for the deletes, I had also run into problems where they weren't getting propagated everywhere. Not sure if there's a similar thing happening here; If I recall correctly, the issue I was experiencing related to unsigned fetches.

    @julian

    read more

  • @klu9 @eyeinthesky

    Having multiple servers connect to each other is Federation.

    Having multiple independent servers (regardless of whether they connect to each other or not) is Decentralization.

    ...

    TS is an independent server — thus, it with others form Decentralized social-media.

    TS does not connect to other servers — thus, not Federated.

    read more
Post suggeriti
  • 0 Votes
    1 Posts
    10 Views
    The big feature in release v3.2.2 of Ktistec is pinned posts with support for the Mastodon Featured Posts collection. Federation works both ways—pin a post on Ktistec and it will show up as a pinned post on Mastodon and vice versa. When you refresh an actor profile, Ktistec also fetches and updates the actor's pinned posts. This is another small step in the direction of supporting all features that Mastodon-compatible client applications expect to access via the API. It's also useful in its own right. The other major feature, which I posted a short video demonstrating here, is X-Ray Mode. X-Ray Mode is a developer and power-user tool for inspecting ActivityPub JSON-LD representations of actors, objects, and other content. Pressing Ctrl+Shift+X on any page displays the data behind the page—like an x-ray. You can:Cached Version: View the local JSON-LD representation stored in the Ktistec databaseRemote Version: Fetch and view the original JSON-LD representation from the source serverNavigation: Click on any ActivityPub IRI to navigate to that objectHistory: Use Alt+Left and Alt+Right to navigate through your viewing historyThis feature is useful for debugging federation issues, understanding ActivityPub structures, and verifying how content is stored and represented.Here's the full changelog for the release:AddedSupport for pinned posts and the Mastodon "featured posts" collection.X-Ray Mode for viewing and navigating JSON-LD resource (actor, object, etc.) representations.Back links on thread pages for easier navigation. (fixes #1)License page for LibreJS compliance. (fixes #127)Highlighting of recently fetched hashtagged posts.ChangedImproved presentation of audio and video media.Refactored theming/styling implementation.The next release will focus on smaller features and bug fixes.Enjoy!#ktistec #crystallang #activitypub #fediverse
  • 1 Votes
    20 Posts
    20 Views
    @phnt C2S API has always been a solution looking for a problem, but it is similar enough to FEP-ae97 API, so I have no issue with people devoting their time to fixing C2S.However, almost nobody actually works on it. There is a lot of cheap talk, but anyone who actually tries to implement C2S quickly realizes how broken it is and gives up. Most progress so far has been made by a single developer (btw: I began to document some aspects of his implementation in FEP-9f9f: Collections).>fixing the complete mess of a specification and making a v2 spec that isn't ambiguous and open-ended as a typical corporate privacy policyThe working group is too busy renaming https://www.w3.org/ns/activitystreams#Public to as:Public@julian @django
  • 0 Votes
    26 Posts
    99 Views
    @julian that might be it, yeah. languages outside of javascript generally don't make a distinction between null and undefined, and even in javascript these are used inconsistently. for example localStorage.getItem will return null for a missing key. practically speaking, the "intentionally" distinction is a distinction without a difference in most processing contexts.
  • 0 Votes
    2 Posts
    21 Views
    @JohannesStarke @pfefferle@notiz.blog 🫣