akkoma/README.md

6.5 KiB

What is This?

This is a fork of AkkomaGang/akkoma, with an opinionated Docker setup.

The differences between the upstream repository are described below:

  • Uses the official Elixir image from Docker Hub
  • Automatically builds the Akkoma backend when building the Docker image, instead of requiring complex commands after pulling/building images
  • Files/directories to mount are minified and completely separated from the repository files, allowing better control over file permissions
  • Has an "actual" Docker entry point that works the way Docker entry points are supposed to work
  • Supports Podman quite well

How to Setup (Podman Rootless)

0. Requirements

  • A decent CPU
  • 2GB of memory (200MB runtime)
  • 500MB free disk space
  • Podman runs as non-root user

1. Build your image

This will take 10+ minutes.

cd /YOUR/GIT/DIRECTORY/akkoma/
podman build -t akkoma .

2. Create a Podman Pod for Akkoma

podman pod create \
  --publish YOUR_PORT:4000 \
  akkoma

You can use pasta(1) if you want to preserve incoming IPs and easily access the host machine through the network gateway.

podman pod create \
  --network pasta:--map-gw,-a,10.0.2.0,-n,24,-g,10.0.2.1 \
  --add-host gateway:10.0.2.1 \
  --publish YOUR_PORT:4000 \
  akkoma

3. Setting up the database

podman run -d \
  --pod akkoma \
  --restart unless-stopped \
  --name akkoma-db \
  -e POSTGRES_DB=YOUR_DB_NAME \
  -e POSTGRES_USER=YOUR_DB_USER \
  -e POSTGRES_PASSWORD=YOUR_DB_PASS \
  -v ${PWD}/docker-db/:/var/lib/postgresql/data/:Z \
  postgres:14-alpine

4. Setting up the backend

podman run -d \
  --pod akkoma-pod \
  --restart unless-stopped \
  --name akkoma-web \
  -e DB_NAME=YOUR_DB_NAME \
  -e DB_USER=YOUR_DB_USER \
  -e DB_PASS=YOUR_DB_PASS \
  -v ${PWD}/static/:/var/lib/akkoma/static/:Z \
  -v ${PWD}/uploads/:/var/lib/akkoma/uploads/:Z \
  -v ${PWD}/etc/:/etc/akkoma/:Z \
  akkoma

podman exec -it akkoma-web \
  pleroma_ctl \
  instance gen \
  --output /etc/akkoma/config.exs

You will be asked a few questions. Make sure the database password is not [autogenerated].

5. Test your setup

podman restart akkoma-web
curl -I http://localhost:YOUR_PORT/

You should see HTTP/1.1 200 OK.

6. Setting up the frontend

podman exec -it akkoma-web \
  pleroma_ctl \
  frontend install \
  pleroma-fe
podman exec -it akkoma-web \
  pleroma_ctl \
  frontend install \
  admin-fe

7. Create your first user

podman exec -it akkoma-web \
  pleroma_ctl \
  user new \
  YOUR_NAME \
  YOUR@EMAIL.EXAMPLE \
  --admin

Open the printed link to set your password.


Akkoma

a smallish microblogging platform, aka the cooler pleroma

English OK 日本語OK

About

This is a fork of Pleroma, which is a microblogging server software that can federate (= exchange messages with) other servers that support ActivityPub. What that means is that you can host a server for yourself or your friends and stay in control of your online identity, but still exchange messages with people on larger servers. Akkoma will federate with all servers that implement ActivityPub, like Friendica, GNU Social, Hubzilla, Mastodon, Misskey, Peertube, and Pixelfed.

Akkoma is written in Elixir and uses PostgreSQL for data storage.

For clients it supports the Mastodon client API with Pleroma extensions (see the API section on https://docs.akkoma.dev/stable/).

Differences with Pleroma

Akkoma is a faster-paced fork, it has a varied and potentially experimental feature set tailored specifically to the corner of the fediverse inhabited by the project creator and contributors.

This should not be considered a one-for-one match with pleroma; it is more opinionated in many ways, and has a smaller community (which is good or bad depending on your view)

For example, Akkoma has:

  • Custom Emoji reactions (compatible with misskey)
  • Misskey-flavoured markdown support
  • Elasticsearch and Meilisearch support for search
  • Mastodon frontend (Glitch-Soc and Fedibird flavours) support
  • Automatic post translation via DeepL or LibreTranslate
  • A multitude of heavy modifications to the Pleroma Frontend (Pleroma-FE)
  • The "bubble" concept, in which instance administrators can choose closely-related instances to make a "community of communities", so to say

And takes a more opinionated stance on issues like Domain blocks, which are enforced far more on Akkoma.

Take a look at the Changelog if you want a full list of recent changes, everything since 3.0 has been Akkoma.

Installation

If you are running Linux (glibc or musl) on x86, the recommended way to install Akkoma is by using OTP releases. OTP releases are as close as you can get to binary releases with Erlang/Elixir. The release is self-contained, and provides everything needed to boot it. The installation instructions are available here.

From Source

If your platform is not supported, or you just want to be able to edit the source code easily, you may install Akkoma from source.

Docker

Docker installation is supported via this setup

Packages

Akkoma is packaged for YunoHost and can be found and installed from the YunoHost app catalogue.

Compilation Troubleshooting

If you ever encounter compilation issues during the updating of Akkoma, you can try these commands and see if they fix things:

  • mix deps.clean --all
  • mix local.rebar
  • mix local.hex
  • rm -r _build

Documentation