schmonz.com is a Fediverse instance that uses the ActivityPub protocol. In other words, users at this host can communicate with people that use software like Mastodon, Pleroma, Friendica, etc. all around the world.

This server runs the snac software and there is no automatic sign-up process.

Search results for tag #selfhosting

[?]Mason Loring Bliss [he, him, his] » 🌐
@mason@partychickens.net

I spun up some cheap hardware on OVH today, and to my dismay, it didn't have web-based IPMI, just the old Java-based cruft. SuperMicro, ew.

Anyway, the neat thing I realized is that their rescue environment, which you can netboot, is vaguely recent Debian, and it has ZFS on it. This opens up a lot of possibilities, but mostly it meant I didn't need to try to upload an almost-two-gig live environment for a server install.

I wish the IPMI situation wasn't so grim, but man, cheap server, way more capacity than any VPS I'd find, and I can have easy root encryption.

    [?]littleFedi - Official Account » 🌐
    @littlefedi_official@littleone.littlefedi.social

    Another little step for !

    As more people start testing and running littleFedi, bug reports, suggestions and ideas are becoming increasingly important. And keeping track of everything through Mastodon posts and private messages isn't exactly the best solution. :)

    So, we now have a proper Issue Tracker!

    https://issues.littlefedi.org/

    Found a bug? Have an idea for a new feature? Think something could be improved, or that the documentation is missing something? This is the place.

    You can register, check existing issues and open new ones.

    littleFedi is still growing, and your feedback is helping us make it better, one small step at a time.

    Thanks to everyone who's been testing, reporting, suggesting and, most importantly, giving this little project a chance.

      [?]Stefano Marinelli - littleFedi » 🌐
      @stefano@littleone.littlefedi.social

      littleFedi: reverse proxy and split domains

      littleFedi can use one domain for account handles and another for the server. For example:

      • @alice@example.com is the address people follow.
      • https://social.example.com serves the web interface, API, media and ActivityPub objects.

      You can keep your existing website on example.com. It only needs a rule that redirects WebFinger requests to the littleFedi domain.

      Choose both domains before creating accounts. Changing them later changes the instance’s identity.

      littleFedi configuration

      In littlefedi.toml, with the reverse proxy running on the same machine:

      local_domain = "example.com"
      web_domain = "social.example.com"
      
      [http]
      listen = "127.0.0.1:8080"
      proxy_uri = "https://social.example.com"
      trusted_proxies = ["127.0.0.1/32"]
      

      If you already have an [http] section, edit it rather than adding another.

      Point the DNS records for social.example.com at the reverse proxy. The server hosting example.com must also serve HTTPS. A DNS alias alone does not replace the WebFinger redirect.

      Expose ports 80 and 443 on the proxy and keep port 8080 private. If the proxy runs on another machine, use a private listener address and trust only the proxy’s address.

      Caddy

      Add this to your Caddyfile, replacing the example domains:

      social.example.com {
          reverse_proxy 127.0.0.1:8080
      }
      
      example.com {
          redir /.well-known/webfinger https://social.example.com{uri} 302
      }
      

      Caddy handles certificates and the WebSocket connections used for streaming.

      If example.com already has a site block, add only the redir line to that block and keep the existing website configuration.

      Check the configuration before reloading Caddy:

      caddy validate --config /etc/caddy/Caddyfile
      

      nginx

      This configuration belongs in nginx’s http context, usually through an included site file. Replace the domains and certificate paths with your own.

      Install valid certificates and arrange automatic renewal before enabling the HTTPS servers.

      map $http_upgrade $connection_upgrade {
          default upgrade;
          '' close;
      }
      
      server {
          listen 80;
          listen [::]:80;
          server_name example.com social.example.com;
          return 308 https://$host$request_uri;
      }
      
      server {
          listen 443 ssl;
          listen [::]:443 ssl;
          server_name social.example.com;
      
          ssl_certificate /etc/letsencrypt/live/social.example.com/fullchain.pem;
          ssl_certificate_key /etc/letsencrypt/live/social.example.com/privkey.pem;
          client_max_body_size 41m;
      
          location / {
              proxy_pass http://127.0.0.1:8080;
              proxy_http_version 1.1;
              proxy_set_header Host $host;
              proxy_set_header X-Forwarded-Host $host;
              proxy_set_header X-Forwarded-Proto $scheme;
              proxy_set_header X-Forwarded-For $remote_addr;
              proxy_set_header Upgrade $http_upgrade;
              proxy_set_header Connection $connection_upgrade;
              proxy_read_timeout 1h;
              proxy_buffering off;
          }
      }
      
      server {
          listen 443 ssl;
          listen [::]:443 ssl;
          server_name example.com;
      
          ssl_certificate /etc/letsencrypt/live/example.com/fullchain.pem;
          ssl_certificate_key /etc/letsencrypt/live/example.com/privkey.pem;
      
          location = /.well-known/webfinger {
              return 302 https://social.example.com$request_uri;
          }
      
          location / {
              return 404;
          }
      }
      

      If example.com already hosts a website, add only the WebFinger location to its existing HTTPS server block. Keep the website’s other rules instead of the return 404 shown above. If you already have the same map directive, do not add it twice.

      Check the configuration before reloading nginx:

      nginx -t
      

      Check discovery

      Use an existing local account, replacing alice with its username:

      curl --fail -i 'https://example.com/.well-known/webfinger?resource=acct:alice@example.com'
      

      The response should redirect to social.example.com, keeping the resource query parameter.

      Then follow the redirect:

      curl --fail -L 'https://example.com/.well-known/webfinger?resource=acct:alice@example.com'
      

      The JSON should contain acct:alice@example.com as its subject, with an ActivityPub actor link at https://social.example.com/users/alice.

      Finally, check a follow in both directions with another instance.

      If account handles and the server use the same domain, only the reverse proxy is needed.

        [?]Rev Mike » 🌐
        @RevMike@social.tchncs.de

        Hello Mastodon! 👋

        Thought I'd introduce myself!

        I'm Mike, living in North Yorkshire, UK. By day I'm a Church of England vicar, but this account is mostly for indulging my geeky side!

        I'm a Linux enthusiast, running an OMV home server with Docker, Nextcloud, Home Assistant, Immich and various other bits and pieces. I've recently moved to a Jolla phone running Sailfish OS, and I'm enjoying exploring life beyond Android and Google.

        I'm interested in open source, self-hosting, privacy and digital independence. I enjoy tinkering, learning new things and occasionally breaking stuff before figuring out how to fix it!

          [?]dch :flantifa: :flan_hacker: » 🌐
          @dch@bsd.network

          I've been using -serve github.com/charmbracelet/soft- for ages, and will probably keep doing that, even though its now archived.

          I'd like to have a simple read-only view though, that others can click and browse through, or link from.

          Any recommendations for that? Something go or rust would be ideal.

          Maybe I should add this as a feature to soft-serve, and fork it?

            [?]littleFedi - Official Account » 🌐
            @littlefedi_official@littleone.littlefedi.social

            A new build of littleFedi (2026.11.16) is out!

            The Little Computer People have been busy, and we're happy to share a new release of littleFedi!

            Before reading the long changelog, there's something important we'd like you to know about upgrading.

            ⚠️ Important: Upgrading littleFedi

            This release includes database migrations that run automatically on the first startup. Some of them may take a while, especially on larger databases or slower hardware.

            Please read these instructions carefully before upgrading.

            1. Stop littleFedi and make a backup.

              Back up your database, configuration, and media files. Remember to save any encryption keys stored in the service environment, too.

              See docs/backup-restore.md for backup and restore instructions.

            2. Replace the binary.

              Download the new binary for your operating system, CPU architecture, and database/media backend. Replace the old binary and start littleFedi using your existing configuration.

              No separate migration command is needed.

            3. Allow the database migrations to finish.

              littleFedi will automatically perform the necessary migrations on its first startup. One of these rebuilds SQLite indexes and may take considerable time on large databases, particularly when running from an SD card.

              Make sure your service manager allows enough time for startup.

              Do not stop or restart littleFedi while migrations are running.

            4. Check that everything is ready.

              Wait for startup to complete, then check the logs and the /readyz endpoint before using your instance.

              Subsequent startups will skip migrations that have already completed.

            Need to roll back?

            Stop littleFedi and restore your pre-upgrade backup before starting the previous binary.

            Never run an older binary against a migrated database.

            What's new?

            As with previous releases, many of the changes come directly from the people who make this project worthwhile: our friends on this instance, those who have started self-hosting littleFedi, and everyone who has taken the time to test, report bugs, suggest improvements, or review our code.

            We've continued working on the things that matter to us: keeping littleFedi lightweight, efficient, reliable, and enjoyable to use, without unnecessary complexity.

            And yes, we're still very much committed to making the Fediverse accessible even to the smallest computers!

            Here's what's changed in this release:

            Changelog

            • Account blocks now hide posts in profiles, search, threads, quotes and

              streaming, and stop new follows, mentions, replies, favourites, reactions,

              poll votes, boosts and direct messages in both directions. Blocked profiles

              keep only the account's identity and block/report actions. Existing direct

              conversations stay readable, and email and push notifications check for

              blocks before sending. Other accounts on the same server remain reachable,

              even when they share an inbox. Public posts remain available to anonymous

              readers and other instances.

            • Static blog pages now leave out unlisted replies. Only replies with public

              visibility are included. Rebuild existing blogs after upgrading to remove

              unlisted replies from pages generated by earlier versions.

            • Content warnings now display apostrophes and other HTML entities correctly

              in posts and notifications, instead of showing codes such as '.

            • Source builds now require Go 1.27.4. Updated the Go dependencies, bundled

              Alpine.js and htmx, and the security tools used in CI. Rebuild existing

              binaries to pick up the Go security fixes.

            • With housekeeping.max_delete_per_run = 0, status pruning now keeps

              going in batches of 500 until no eligible posts remain. Previously, remote

              pruning stopped after 500 posts and local pruning after 500 per account.

              Positive limits still apply across local and remote posts. Failed deletions

              are left for the next run rather than retried in a loop.

            • The inbox no longer fetches a key for an unknown actor's self-delete.

              Requests with a Signature header are acknowledged with 202 and dropped

              when the actor and deleted object are the same URI and the account is absent

              locally. Domain blocks still apply; deletes of known accounts and posts

              still need authentication.

            • Incoming HTTP signatures now have their timestamps and body digests checked

              before fetching a public key or verifying the signature. Stale requests and

              damaged bodies can be rejected without a network request.

            • Public instance counts on /about and the v1/v2 instance APIs are cached

              for 30 seconds, instead of counting the database on every request. The

              admin dashboard still reads fresh counts.

            • SQLite reply and quote indexes now leave out NULL entries. This reduces

              their size and avoids a slow query plan when loading RSS feeds. Migration

              007 rebuilds all four indexes at startup and can be slow on large databases,

              especially on SD cards. Allow extra time for the first startup after

              upgrading; later starts skip the migration.

            • The interface is now available in Catalan, thanks to Jolupa, who

              translated it. It can be picked in the settings like the other languages,

              and Catalan is also in the list of languages a post can be tagged with.

              Strings added after the translation was made were filled in afterwards and

              are the ones most worth a second look.

            • Boosts can be scheduled, to spread them out instead of filling followers'

              timelines all at once. Hold the boost button for half a second, or open the

              post, and pick a time under "Boost later". A quick press still boosts right

              away. Scheduled boosts show up on the Scheduled page next to scheduled

              posts and can be moved or cancelled there. If the post is deleted or

              becomes impossible to boost before then, the boost is dropped. Mastodon

              apps have no notion of a scheduled boost, so the API leaves them out.

            • Scheduled posts can now be edited, not just moved to another time. The

              Edit button on the Scheduled page opens the post in the composer with

              everything it had, including the time, and saving puts it back on the

              schedule. Before, the only way was to cancel it and write it again. A post

              that is already being published can no longer be edited.

            • Posts can be saved as drafts. The composer has a "Save draft" button next

              to Post, and the new Drafts page lists them, newest first. Opening one

              brings back the text, content warning, audience, language, attachments,

              poll and expiry, and the post it replies to or quotes. Posting or

              scheduling a draft removes it. Everything is a plain form, so it works

              without JavaScript and in text browsers. Drafts live on the server and are

              not shared with Mastodon apps, which have no API for them.

            • The image viewer can now move between the pictures of a post. Use the

              arrows on screen, the left and right keys, or swipe on a phone. Before, each

              picture had to be closed and the next one opened by hand. Without

              JavaScript nothing changes: each picture still opens on its own.

            • In French the image viewer had no name for screen readers. The apostrophe

              in "Visionneuse d'images" broke the expression that sets it. Translated

              labels are now read from data attributes, so quotes in a translation can no

              longer break the script.

            • The Roost theme works better on phones. Post actions fit on one row, the

              back link is a flat row instead of a pill against the screen edge, the

              profile banner and avatar no longer fill the first screen, and the bottom

              bar marks the current page without a filled block.

            • Unread mentions and direct messages now stand out from the other unread

              notifications. In Roost the mention colour was the same blue as the unread

              mark, and in every theme their rows took the same background as any unread

              row. Herd and Ink also get mention colours further from their accent.

            • Filters set to warn now work on the catch-up page. The filtered posts were

              shown in full, without the filter notice. Filters set to hide already

              worked there.

            • Direct message notifications have a magenta edge, an envelope icon and their

              own label, so they stand apart from public mentions.

            • Boosts from a followed account can now be hidden from the home timeline,

              with the "Hide boosts" button on their profile or the matching switch in

              Mastodon clients. Their own posts still show up. Boosts already in the

              timeline go away too, and come back if boosts are shown again. The
              reblogs parameter of the follow API was accepted but ignored until now.

            • Push delivery now follows up to three 307/308 redirects within the same

              HTTPS origin. The Mastodon app relay redirects endpoints without a trailing

              slash; those pushes previously failed after all retries.

            • Pushes for favourites, boosts and federated mentions now carry
              notification_id. Those jobs were built before the notification was stored,

              so the ID was missing, and the official iOS app showed only the relay's

              placeholder trumpet instead of the notification.

            • The bookmark button on boosted posts now stays in sync after refreshing

              the home timeline. Before, it could appear inactive while the post was still

              bookmarked, and clicking it again removed the bookmark.

            • Replies no longer show the previous post as unavailable after it has been

              downloaded. Refreshing or reopening the reply now keeps the correct reply

              target in timelines, boosted posts and Mastodon clients. The open thread

              page also updates its reply label as earlier posts arrive.

            • Redrew the quote icon. The old one, two squared-off hooks side by side,

              could be read as "SS". It now looks like a pair of quotation marks.

            • Logging in with toot from the command line now works. toot opens
              /oauth/authorize/ with a trailing slash, which returned 404. URLs with a

              trailing slash are now routed like the ones without, as on Mastodon.

            • With require_approval on, new accounts were created already approved: they

              could sign in at once and never showed up in the approval queue. They now

              wait for an administrator. Accounts registered while the bug was present

              stay approved; check recent signups if the setting is on.

            • Filter keywords created from a Mastodon client with whole_word turned off

              were saved with it on. Leaving whole_word out still means on, as in

              Mastodon.

            • Mastodon apps no longer show littleFedi's own notification types (follow

              accepted, unfollows, blocks, emoji reactions, blog rebuilds and others) as

              broken entries, and their unread badge no longer counts them. A client can

              still ask for them with types[], which the notification endpoints now

              honour. exclude_types[]=pleroma:emoji_reaction now works too.

            • The user:notification stream carries only notifications. Before, it also

              sent home timeline posts, which some clients (Tuba) showed as empty

              notifications.

            • The instance documents report image_matrix_limit as a pixel count, taken

              from max_image_pixels and max_image_dimension. It was 4096, which clients

              read as 4096 pixels in total and used to shrink uploads to a thumbnail. The

              video limits Mastodon's iOS app requires are there too.

            • New client API endpoints: GET /api/v1/accounts?id[]=,
              /api/v1/accounts/lookup, /api/v1/accounts/familiar_followers and
              /api/v1/notifications/policy. Mastodon for iOS and Tuba call them.

            • Push notifications now reach the Mastodon apps (Mastodon for Android and

              iOS, Tusky, Ice Cubes, Ivory, Mona). Apps that don't ask for
              standard: true get the older aesgcm encoding, which is the only one

              several of them can decrypt, and every push now carries the access token

              and a numeric notification ID, without which the apps drop it. Form encoded

              subscriptions are accepted, and the subscription ID is a number as the apps

              expect. Subscriptions made before this release keep the old behaviour:

              turn notifications off and on again in the app, or sign in again.

            • Polls no longer break Mastodon for iOS and Android. own_votes and emojis

              are always lists, and expires_at uses Mastodon's date format.

            • Deleting a post from an app returns the deleted post with its source text,

              so "delete and redraft" keeps the text. Deleting it twice returns 404.

            • Timeline paging: min_id returns the page right after the cursor instead of

              jumping to the newest posts, which made apps skip everything in between when

              filling a gap. A cursor pointing at a deleted post keeps working, and an

              unknown max_id returns an empty page instead of the top of the timeline.
              min_id and since_id work for notifications too.

            • Link headers are now absolute URLs and keep the other query parameters.

              Followers, following, blocks, mutes, favourites, bookmarks, lists and list

              members send them too, so apps can load more than the first page.

            • Media sizes are reported under meta.original as well, and remote media

              keep their blurhash and size. Apps used them to lay out images before they

              load.

            • Accounts without an avatar or header point to a placeholder image instead of

              an address that returned 404. Accounts also report uri and
              hide_collections, and relationships requested_by and languages.

            • The default post visibility and language set in the profile are reported to

              apps and used when an app leaves them out.

            • /api/v2/instance reports the VAPID key under configuration.vapid and
              translation.enabled. Mastodon for iOS reads both.

            • Loading images, avatars and emoji no longer counts against the rate limit.

              Scrolling a busy timeline could hit it, and every API call after that failed.

              An API request over the limit now gets a JSON error with Retry-After, and

              unknown API routes return JSON 404 and 405 errors instead of HTML.

            • littleFedi now logs a warning when requests come from a local reverse proxy

              that is not listed in http.trusted_proxies. In that case every visitor

              shares the proxy's rate limit.

            • CORS preflight allows the headers the browser asks for (Idempotency-Key

              among them) and exposes Link and Retry-After to web clients.

            • Idempotency-Key is honoured when posting: a retried request returns the

              post created the first time instead of a duplicate.

            • Tokens from client_credentials (the app's own token, before a user signs

              in) are treated as anonymous on public endpoints instead of being refused.

            • Filters: keywords_attributes works when creating and editing a filter, in

              JSON and form encoding, and the /api/v2/filters/:id/keywords and
              /api/v2/filters/keywords/:id endpoints are new.

            • New endpoints: GET /api/v1/accounts/search,
              /api/v1/statuses/:id/history, /api/v1/timelines/direct and
              /api/v2/notifications/unread_count.

            • The streaming API accepts the token in Sec-WebSocket-Protocol, as the

              Mastodon web client sends it, and tokens limited to read:statuses and
              read:notifications.

            • The "Clear all" button on the notifications page is now called "Delete

              all", in every language, since that is what it does.

            • Profile fields: up to four label and value rows on your profile, edited in

              Settings or from any Mastodon app, and shown under the bio on the profile

              page. They are sent to other servers like Mastodon's, and the ones remote

              accounts publish are now shown on their profiles. Remote accounts already

              known get theirs the next time their profile is fetched again. Links work;
              verified_at is always empty, as littleFedi does not check rel="me"

              links.

            • Apps can mute a conversation (POST /api/v1/statuses/:id/mute and
              /unmute). Mentions, likes, boosts, quotes, edits and reactions in that

              thread stop notifying you.

            • Apps can keep a private note on an account

              (POST /api/v1/accounts/:id/note). Only you see it.

            • Followers can be removed from an app

              (POST /api/v1/accounts/:id/remove_from_followers). A remote follower's

              server gets a Reject of the follow, as from Mastodon.

            • Conversations can be removed from the list from an app

              (DELETE /api/v1/conversations/:id); a new message brings them back.
              POST /api/v1/conversations/:id/read works too.

            • Saving the profile from an app with an empty bio or display name now clears

              it. Before, empty values were ignored and the old text stayed. The default

              post visibility and language set in an app are saved too.

            • MastoBlaster could not read its push subscription: the alerts object had

              no update key. It is now there, and apps can turn on pushes for edited

              posts. Renaming a list and changing push alerts from MastoBlaster failed

              with 405, because it sends PATCH; both endpoints now take PATCH as well

              as PUT, like Mastodon.

            • Three database migrations run at startup. One adds the standard and
              access_token columns to push_subscriptions. One adds
              accounts.fields_json and the thread_mutes, account_notes and
              hidden_conversations tables. The last adds
              push_subscriptions.notify_update. All of them only add, nothing existing

              is changed, and all can be rolled back.


            A big thank you to everyone who has been following this little adventure, testing new builds, sharing feedback, and helping us improve.

            littleFedi is still a young project, and there's plenty more to do. But every release brings us a little closer to what we've wanted from the beginning: a federated social platform that's simple to run, easy to maintain, and doesn't require a powerful server just to participate in the Fediverse.

            Small computers deserve a place in the Fediverse, too. ❤️

            Happy upgrading, happy self-hosting, and stay tuned!

            The Little Computer People

              Miod Vallat boosted

              [?]Valère » 🌐
              @valere@hostux.social

              New service : tools.hostux.net

              More than 400 utilities for developers and sysadmins: JSON/YAML formatting, JWT decoding, subnet calculations, UUID generation, cryptography and more.

              Based on the maintained IT-Tools fork by ShareVB, with the Hostux theme applied to keep the same look and feel across services.

              No account required. Tools run locally in your browser.

              Screenshot of IT-Tools at tools.hostux.net An orange diagonal line separates the light and dark versions of the interface. A sidebar on the left lists tool categories, while the main area displays a grid of utilities, including unit converters, email analysis tools, and LDAP search. Bold typography, thin dividing lines, and orange accents reflect the Hostux visual identity.

              Alt...Screenshot of IT-Tools at tools.hostux.net An orange diagonal line separates the light and dark versions of the interface. A sidebar on the left lists tool categories, while the main area displays a grid of utilities, including unit converters, email analysis tools, and LDAP search. Bold typography, thin dividing lines, and orange accents reflect the Hostux visual identity.

                Jay 🚩 :runbsd: boosted

                [?]Dragon of BSDCafe [he/him] » 🌐
                @evgandr@snac.bsd.cafe

                Well, 2TB HDDs with 5400 rpm were cheap, but ... ETA: 200:31:46 😀


                Photo of NetBSD text console, showing the ETA 200 hours for RAID parity rebuild process.

                Alt...Photo of NetBSD text console, showing the ETA 200 hours for RAID parity rebuild process.

                  [?]Julian Oliver » 🌐
                  @JulianOliver@mastodon.social

                  What happened with Outlook will happen again.

                  Before you jump to another host, consider .

                  In 2026 you can have a fast and secure mail server of your own, easy to admin & update without a compsci degree. One that is tightly-secured, supports all mail clients, has top delivery & encrypted storage. All under your control, & your domain.

                  If that sounds interesting, sign up:

                  courses.nikau.io/convoy

                  We'll pick a date as a group, & I'll guide you all through the build.

                  The title image for the convoy training, featuring Canadian Geese flying against a pale sky above the word convoy in caps.

                  Alt...The title image for the convoy training, featuring Canadian Geese flying against a pale sky above the word convoy in caps.

                    [?]Stefano Marinelli » 🌐
                    @stefano@mastodon.bsd.cafe

                    I've always appreciated the simplicity of the BSDs. Power, yes, but without unnecessary complexity.

                    BastilleBSD and vm-bhyve have been my trusted companions for many years. But whenever I used Linux, I missed them.

                    That's why, years ago, I started putting together a collection of shell scripts (and some Python, although to a much lesser extent) to create something somewhere in between.

                    A bit of Bastille, a bit of vm-bhyve, a bit of Proxmox: unprivileged Linux system containers and virtual machines, managed from the command line.

                    Live backups. And live restore: bringing a workload back online before all its data has finished copying from the backup. Multiple storage backends. Encryption at the storage or individual workload level. Different networks. Isolation of each workload (a bit like illumos zones), with additional layers of protection.

                    Each container has its own UID/GID mapping. VMs run under separate unprivileged identities, with namespaces, seccomp and resource limits adding further layers of protection.

                    Pull-based backups from external hosts, with different levels and types of encryption. With raw encrypted ZFS replication, the backup server doesn’t even need the decryption keys. Backups are pulled over SSH by an external server. The workload host holds no credentials to log in to that server.

                    It also runs on a 1 GB Orange Pi PC 2, where I’ve tested both containers and native ARM64 KVM guests. And on a Orange PI One, armv7 and 512MB, it can run some lxc containers without problems, from the SD, using f2fs.

                    Simple disaster recovery. Because after those 142 servers went up in flames in Strasbourg, disaster recovery became something I care a lot about.

                    Currently supported and tested on Alpine Linux, Debian 13 and Armbian. On Debian, LXC containers also run under enforcing AppArmor profiles, adding another layer of protection.

                    Few commands. Lots of power. And it's called "Jail It!"

                    I'm not going to release it, at least not for now. I need to focus on my other projects, and Jail It! still needs some cleaning up.

                    But it works. And it does its job very well.

                    For containers, there's absolutely nothing running in the background. For VMs, just a small Python helper. No continuously running services. You connect via SSH, run your commands, and that's it.

                    Still, for anyone curious enough to have a look, I've put together a short demo on Asciinema:

                    asciinema.org/a/BbwGs99U9UJDLA

                    Well, "short" might be a bit of an exaggeration. It's quite long, but I think it's worth watching all the way to the end.

                      [?]Elena Rossini 🌈 » 🌐
                      @_elena@mastodon.social

                      There has never been a better time to introduce young French people to the wonderful world of decentralized social networks and - optional - .

                      An accidental ad for the , exhibit A: politico.eu/article/french-gov

                        #netbsd boosted

                        [?]jbz » 🌐
                        @jbz@indieweb.social

                        [?]Eggs now in different baskets. » 🌐
                        @the_wub@mastodon.social

                        1/n OK about to start dipping my toes into the world of .

                        Reasons.
                        My disappointment with the new Mastodon web UI just reminds me of the many issues of using the "cloud", that is having your stuff on someone else's server.

                        My disappointment with those working on the new UI sound just like any other UX developer in a big oligarch owned company who are not going to go back or not make any significant improvements to their beloved concept.

                        As no-one is going to force them to.

                          [?]viq [he/him] » 🌐
                          @viq@social.hackerspace.pl

                          Sigh, is slop, and dropping features.
                          I've looked at rexify.org/ and I failed to find how it deals with secrets.
                          I guess I either switch to by way of , or unify my systems on (and swallow linux sloppification). Sigh.

                            [?]Justine Smithies [She / Her] » 🌐
                            @justine@snac.smithies.me.uk

                            Seriously contemplating migrating to just because I am starting to out grow in terms of size(followers/following). Don't get me wrong Snac is absolutely FANTASTIC and I'd 100% recommend it to anyone starting out their own fedi instance but Littlefedi is a lot snappier for me in terms of account size and it doesn't use near as much disk access as Snac. I haven't 100% made up my mind yet and something or someone may change it but the Littlefedi option is really looking promising.

                              [?]Grunkle Moose » 🌐
                              @RootMoose@mastodon.bsd.cafe

                              Finally turned attention to getting Immich up and running tonight. I like it, seems snappy enough with tens of thousands of photos imported.

                              The Immich website has a handy form driven page for generating a docker compose file so it took like 30 seconds to get it running once mutated to follow my way of doing things on my server (persistent storage, naming, cert proxy, etc).

                              I struggled a bit with using the Immich command line tool to import my photo collection and in the process of trying to figure out why --recursive didn't work, stumbled upon the immich-go project on github. It's also a command line tool and works very well; immediately was able to recursively import photos, etc. Compiles cleanly on Alpine Linux (ofc, it's just golang).

                              Immich-go has a migrate function that allows you to connect to one Immich server and copy the files to another Immich server (but no 'delete source' as far as I could tell).

                              I set up a second Immich instance on a different port and started to use the migrate function to finally start moving all the photos of my computer electronics and vehicle technical projects out of the whole household's photo collection. My wife will appreciate that. Ha. Managed to split out 2026-2025, only ~25 more years to go. Ugh.

                              Anyway, Immich (the server) and Immich-go (the CLI tool) are cool. Recommended.

                              Fun!

                                [?]Jan Wildeboer 😷:krulorange: » 🌐
                                @jwildeboer@social.wildeboer.net

                                I just had a thought. Some people talk about and as some prepper like thing for weird people. I want to change that, as it certainly doesn't reflect how I feel about my mailserver, my various web servers and other services I run myself.

                                And that's where the thought comes in. From now on I will call it:

                                instead (or again, depending on your age :)

                                @homelab

                                  [?]GreyUnix » 🌐
                                  @greylinux@mylittlefedi.capella-ember.uk

                                  Hi everyone from my littlefedi instance, this is my first test post with a media file.

                                  This is one of my dogs when she was very young

                                  #LittleFedi
                                  #SelfHosting

                                  a medium size black dog lying in a white and blue bed. Inside the bed is some of her toys, the one in the foreground is a little orange fox. The wall behind is white and the floor is a light coloured.wood

                                  Alt...a medium size black dog lying in a white and blue bed. Inside the bed is some of her toys, the one in the foreground is a little orange fox. The wall behind is white and the floor is a light coloured.wood

                                    [?]El Perro Negro » 🌐
                                    @elperronegro@khumbu.social

                                    This is an #Intro post from a test account on LittleFedi running from home on a Pi. Let's see how this goes before we migrate all my other accounts here.

                                    I'm Richard, retired adventurer/author living in #Spain. El Perro Negro named after first dog. Bit of a grumpy old man but try to be good company. Will post about #hiking, #mountaineering, #photography, #technology, #selfhosting and #privacy issues. Interested in human generated #RSS content. Not got time for Politicians nor Big Tech.

                                    Would love to meet up on the #Fediverse with like minded folk. Photo from last Novembers trip to El Chalten in Patagonia

                                    A person walks along a rocky ridge. Behind lies the peaks of Cerro Torre and My Fitzroy, El Chalten, Patagonia

                                    Alt...A person walks along a rocky ridge. Behind lies the peaks of Cerro Torre and My Fitzroy, El Chalten, Patagonia

                                      dave boosted

                                      [?]Daniel Gultsch » 🌐
                                      @daniel@gultsch.social

                                      RE: sciences.social/@annasdtc/1172

                                      Not everyone can or should self-host, but the technology that enables it—federation, open standards, and vendor independence—also allows cooperatives, municipalities, and community groups to run their own infrastructure.

                                        [?]Paco Hope [He/Him] » 🌐
                                        @paco@infosec.exchange

                                        I love getting in print, but this time they’ve gone too far. I don’t know how they broke into my basement but I did not give them permission to use photos of my stuff in their mock ads.

                                        Photo of a printed coupon in the onion. It pretends to be from Best Buy and shows a big cardboard box full of random computer cords like USB and power cables. The coupon says “good for one free pull from the Tub O’ Cords”

                                        Alt...Photo of a printed coupon in the onion. It pretends to be from Best Buy and shows a big cardboard box full of random computer cords like USB and power cables. The coupon says “good for one free pull from the Tub O’ Cords”

                                          [?]Brad Taunt :openbsd: :netbsd: » 🌐
                                          @bt@mastodon.bsd.cafe

                                          I think I'm slowly going to make my way back to self-hosting my repositories (via cgit, most likely). Most providers now either have too strict of rules or none at all...

                                          Probably better to have full control anyway 🤷

                                            Back to top - More...