Commit e92a1cf4 authored by Eugen Rochko's avatar Eugen Rochko

Add wiki contents to docs/ instead

parent 500278eb
...@@ -25,12 +25,11 @@ If you would like, you can [support the development of this project on Patreon][ ...@@ -25,12 +25,11 @@ If you would like, you can [support the development of this project on Patreon][
## Resources ## Resources
- [List of Mastodon instances]( - [List of Mastodon instances](docs/Using-Mastodon/
- [Use this tool to find Twitter friends on Mastodon]( - [Use this tool to find Twitter friends on Mastodon](
- [API overview]( - [API overview](docs/Using-the-API/
- [How to use the API via cURL/oAuth]( - [Frequently Asked Questions](docs/Using-Mastodon/
- [Frequently Asked Questions]( - [List of apps](docs/Using-Mastodon/
- [List of apps](
## Features ## Features
...@@ -116,7 +115,7 @@ Which will re-create the updated containers, leaving databases and data as is. D ...@@ -116,7 +115,7 @@ Which will re-create the updated containers, leaving databases and data as is. D
## Deployment without Docker ## Deployment without Docker
Docker is great for quickly trying out software, but it has its drawbacks too. If you prefer to run Mastodon without using Docker, refer to the [production guide]( for examples, configuration and instructions. Docker is great for quickly trying out software, but it has its drawbacks too. If you prefer to run Mastodon without using Docker, refer to the [production guide](docs/Running-Mastodon/ for examples, configuration and instructions.
## Deployment on Heroku (experimental) ## Deployment on Heroku (experimental)
These people make the development of Mastodon possible through [Patreon](
**Extra special Patrons**
- [World'sTallestLadder](
- [glocal](
- [Jimmy Tidey](
- [Kurtis Rainbolt-Greene](
- [Kit Redgrave](
**Thank you to the following people**
- [Sophia Park](
- [WelshPixie](
- [John Parker](
- [Christina Hendricks](
- [Jelle](
- [Harris Bomberguy](
- [Martin Tithonium](
- [Edward Saperia](
- [Yoz Grahame](
- [Jenn Kaplan](
- [Natalie Weizenbaum](
- [Matteo De Micheli](
- [BirdMachine](
- [Jessica Hayley](
- [Niels Roesen Abildgaard](
- [Zatnosk](
If you want to localise Mastodon into your language, here is how.
There are two parts to Mastodon, the server and the web client. The translations for the web client are in `app/assets/javascripts/components/locales`. For the server-side, the translations live in `config/locales` and are divided into different files. Here are all the files you’ll need to translate:
| Original file (English) | Location | Description |
| [`en.jsx`](/Gargron/mastodon/tree/master/app/assets/javascripts/components/locales/en.jsx) | `app/assets/javascripts/components/locales/en.jsx` | Strings for the web client |
| [`en.yml`](/Gargron/mastodon/tree/master/config/locales/en.yml) | `config/locales/en.yml` | Strings for general use |
| [`simple_form.en.yml`](/Gargron/mastodon/tree/master/config/locales/simple_form.en.yml) | `config/locales/simple_form.en.yml` | Strings for the settings area |
| [`devise.en.yml`](/Gargron/mastodon/tree/master/config/locales/devise.en.yml) | `config/locales/devise.en.yml` | Generic strings for Devise |
| [`doorkeeper.en.yml`](/Gargron/mastodon/tree/master/config/locales/doorkeeper.en.yml) | `config/locales/doorkeeper.en.yml` | Generic strings for Doorkeeper |
## Translating
If you use Github, first clone the Mastodon repository to your account.
1. Duplicate the files in their folder and replace `en` in the filenames by your language’s standard two-letters code ([ISO 639-1](
For instance `simple_form.en.yml` becomes `` in the Spanish translation.
2. Also replace the language code in the first lines of all the files, and the last line of the `.jsx` file.
3. Translate the right-side values from English to your language. Keep the indentation and punctuation.
Since Devise and Doorkeeper are popular libraries, there may already be translation files for your language available on the Internet.
## Declaring the language
The locales are mentioned in several other files. To activate your translation, add your language code to the different lists present in these files:
| File | Location | Comment |
| [`index.jsx`](/Gargron/mastodon/tree/master/app/assets/javascripts/components/locales/index.jsx) | `app/assets/javascripts/components/locales/index.jsx` | 2 lines to add |
|[`mastodon.jsx`](/Gargron/mastodon/tree/master/app/assets/javascripts/components/containers/mastodon.jsx) | `app/assets/javascripts/components/containers/mastodon.jsx` | 1 line to add + 1 list to complete |
| [`settings_helper.rb`](/Gargron/mastodon/tree/master/app/helpers/settings_helper.rb) | `app/helpers/settings_helper.rb` | 1 line to add + your language’s name |
| [`application.rb`](/Gargron/mastodon/tree/master/config/application.rb) | `config/application.rb` | 1 list to complete |
## Sending the translation
You can then push the files to git and submit a pull request.
## Testing the translation
Once the pull request is accepted, wait for the code to be deployed on a Mastodon instance. Log-in with your account there, and change the locale in the settings. Browse and use the website. See if everything makes sense in context and if anything seems out of place or breaks the layout. Invite other Mastodon users speaking your language to try it and give feedback. Make changes accordingly and update the translation.
## Updating the translation
Keep an eye on the original English files in `app/assets/javascripts/components/locales` and `config/locales`. When they are updated, pass on the changes to your language files. For new strings, add the new lines to the same position and translate them. Once you’re finished with the updates, you can submit a new pull request.
Development guide
**Don't use Docker to do development**. It's a quick way to get Mastodon running in production, it's **really really inconvenient for development**. Normally in Rails development environment you get hot reloading of backend code and on-the-fly compilation of assets like JS and CSS, but you lose those benefits by compiling a Docker image. If you want to contribute to Mastodon, it is worth it to simply set up a proper development environment.
In fact, all you need is described in the [production guide](, **with the following exceptions**. You **don't** need:
- Nginx
- SystemD
- An `.env.production` file. If you need to set any environment variables, you can use an `.env` file
- To prefix any commands with `RAILS_ENV=production` since the default environment is "development" anyway
- Any cronjobs
The command to install project dependencies does not require any flags, i.e. simply
bundle install
By default the development environment wants to connect to a `mastodon_development` database on localhost using your user/ident to login to Postgres (i.e. not a md5 password)
You can run Mastodon with:
rails s
And open `http://localhost:3000` in your browser. Background jobs run inline (aka synchronously) in the development environment, so you don't need to run a Sidekiq process.
You can run tests with:
You can check localization status with:
i18n-tasks health
You can check code quality with:
## Development tips
You can use a localhost->world tunneling service like ngrok if you want to test federation, **however** that should not be your primary mode of operation. If you want to have a permanently federating server, set up a proper instance on a VPS with a domain name, and simply keep it up to date with your own fork of the project while doing development on localhost.
Ngrok and similar services give you a random domain on each start up. This is good enough to test how the code you're working on handles real-world situations. But as soon as your domain changes, for everybody else concerned you're a different instance than before.
Generally, federation bits are tricky to work on for exactly this reason - it's hard to test. And when you are testing with a disposable instance you are polluting the databases of the real servers you're testing against, usually not a big deal but can be annoying. The way I have handled this so far was thus: I have used ngrok for one session, and recorded the exchanges from its web interface to create fixtures and test suites. From then on I've been working with those rather than live servers.
I advise to study the existing code and the RFCs before trying to implement any federation-related changes. It's not *that* difficult, but I think "here be dragons" applies because it's easy to break.
Production guide
## Nginx
Regardless of whether you go with the Docker approach or not, here is an example Nginx server configuration:
map $http_upgrade $connection_upgrade {
default upgrade;
'' close;
server {
listen 443 ssl;
ssl_certificate /etc/letsencrypt/live/;
ssl_certificate_key /etc/letsencrypt/live/;
keepalive_timeout 70;
sendfile on;
client_max_body_size 0;
gzip off;
root /home/mastodon/live/public;
add_header Strict-Transport-Security "max-age=31536000; includeSubDomains";
location / {
try_files $uri @proxy;
location @proxy {
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto https;
proxy_pass_header Server;
proxy_pass http://localhost:3000;
proxy_buffering off;
proxy_redirect off;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection $connection_upgrade;
tcp_nodelay on;
error_page 500 501 502 503 504 /500.html;
## Running in production without Docker
It is recommended to create a special user for mastodon on the server (you could call the user `mastodon`), though remember to disable outside login for it. You should only be able to get into that user through `sudo su - mastodon`.
## General dependencies
curl -sL | sudo bash -
sudo apt-get install imagemagick ffmpeg libpq-dev libxml2-dev libxslt1-dev nodejs
sudo npm install -g yarn
## Redis
sudo apt-get install redis-server redis-tools
## Postgres
sudo apt-get install postgresql postgresql-contrib
## Rbenv
It is recommended to use rbenv (exclusively from the `mastodon` user) to install the desired Ruby version. Follow the guides to [install rbenv][1] and [rbenv-build][2] (I recommend checking the [prerequisites][3] for your system on the rbenv-build project and installing them beforehand, obviously outside the unprivileged `mastodon` user)
Then once `rbenv` is ready, run `rbenv install 2.3.1` to install the Ruby version for Mastodon.
## Git
You need the `git-core` package installed on your system. If it is so, from the `mastodon` user:
cd ~
git clone live
cd live
Then you can proceed to install project dependencies:
gem install bundler
bundle install --deployment --without development test
yarn install
## Configuration
Then you have to configure your instance:
cp .env.production.sample .env.production
nano .env.production
Fill in the important data, like host/port of the redis database, host/port/username/password of the postgres database, your domain name, SMTP details (e.g. from Mailgun or equivalent transactional e-mail service, many have free tiers), whether you intend to use SSL, etc. If you need to generate secrets, you can use:
rake secret
To get a random string.
## Setup
And setup the database for the first time, this will create the tables and basic data:
RAILS_ENV=production bundle exec rails db:setup
Finally, pre-compile all CSS and JavaScript files:
RAILS_ENV=production bundle exec rails assets:precompile
## Systemd
Example systemd configuration for the web workers, to be placed in `/etc/systemd/system/mastodon-web.service`:
ExecStart=/home/mastodon/.rbenv/shims/bundle exec puma -C config/puma.rb
Example systemd configuration for the background workers, to be placed in `/etc/systemd/system/mastodon-sidekiq.service`:
ExecStart=/home/mastodon/.rbenv/shims/bundle exec sidekiq -c 5 -q default -q mailers -q push
This allows you to `sudo systemctl enable mastodon-*.service` and `sudo systemctl start mastodon-*.service` to get things going.
## Cronjobs
I recommend creating a couple cronjobs for the following tasks:
- `RAILS_ENV=production bundle exec rake mastodon:media:clear`
- `RAILS_ENV=production bundle exec rake mastodon:push:refresh`
- `RAILS_ENV=production bundle exec rake mastodon:feeds:clear`
You may want to run `which bundle` first and copypaste that full path instead of simply `bundle` in the above commands because cronjobs usually don't have all the paths set. The time and intervals of when to run these jobs are up to you, but once every day should be enough for all.
You can edit the cronjob file for the `mastodon` user by running `sudo crontab -e mastodon` (outside of the mastodon user).
## Things to look out for when upgrading Mastodon
You can upgrade Mastodon with a `git pull` from the repository directory. You may need to run:
- `RAILS_ENV=production bundle exec rails db:migrate`
- `RAILS_ENV=production bundle exec rails assets:precompile`
Depending on which files changed, e.g. if anything in the `/db/` or `/app/assets` directory changed, respectively. Also, Mastodon runs in memory, so you need to restart it before you see any changes. If you're using systemd, that would be:
sudo systemctl restart mastodon-*.service
Specs and RFCs used
* [OStatus](
* [Salmon](
* [Portable Contacts](
* [Atom](
* [Atom ActivityStreams](
* [Atom Threading](
* [PubSubHubbub](
* [Webfinger](
* [Link-based Resource Descriptor Discovery](
List of apps
Some people have started working on apps for the Mastodon API. Here is a list of them:
|Tooter|Chrome extension|<>|[](|
If you have a project like this, let me know so I can add it to the list!
Frequently Asked Questions
#### What is a Mastodon?
A prehistoric animal, predecessor of the mammoth.
#### Why the name Mastodon?
There's a progressive metal band with the same name that I'm a fan of that brought the animal to my attention. I thought it's a pretty cool name/animal.
#### How exactly is it decentralized?
There are different ways in which something can be decentralized; in this case, Mastodon is the "federated" kind. Think e-mail, not BitTorrent. There are different servers (instances), users have an account on one of them, but can interact and follow each other regardless of where their account is.
#### Technically, how does the federation work?
We are using the OStatus suite of protocols:
1. Webfinger for user-on-domain lookup
2. Atom feeds with ActivityStreams, Portable Contacts, Threads extensions for the actual content
3. PubSubHubbub for subscribing to Atom feeds
4. Salmon for delivering certain items from the Atom feeds to interested parties such as the mentioned user, author of the status being replied to, person being followed, etc
#### What is
The "flagship" instance of Mastodon, aka the server I run myself with the latest code. It's not supposed to be the only instance in the end.
#### What else is part of the federated network?
Let's call it the "fediverse". It has existed for a longer while, populated by GNU social servers, Friendica, Hubzilla, Diaspora etc. Not every one of those servers is fully compatible with every other. Mastodon strives to be fully standards-compliant and compatibility with GNU social is higher in priority than the others.
#### I tried logging into a GNU social client app with Mastodon and it didn't work, why?
While Mastodon is compatible with GNU social in terms of server to server communication, the client to server API (aka how you access Mastodon) is different. Therefore, client apps that were made for specifically GNU social will not work with Mastodon. The reason for this is half technical, half ideological.
Because Mastodon has been created from a blank slate, it is much simpler to have the API mirror internal structures as closely as possible, rather than build an emulation layer. Secondly, the GNU social client API is actually a half-way implementation of the legacy Twitter API - that's the reason why it works with some older Twitter client apps. However, many of those apps are not maintained anymore, the GNU social API does not actually keep up with the real Twitter API and never fully implemented all its features; at the same time, the Twitter API was never meant for a federated service and so obscures some of the functionality.
#### How is Mastodon funded?
Development of Mastodon and hosting of is funded through my [Patreon (also BTC/PayPal donations)]( Beyond that, I am not interested in VC funding, monetizing, advertising, or anything of that sort. I could offer setup/maintenance services on demand.
The software is free and open source and communities should host their own servers if they can, that way the costs are more or less distributed. Obviously it'd be hard for me to pay the bills if literally everyone decided to use the instance only.
\ No newline at end of file
List of Mastodon instances
* [](
* [](
* [](
* [](
* [](
* [](
* [](
Let me know if you start running one so I can add it to the list!
# Contents
- [Available libraries](#available-libraries)
- [Notes](#notes)
- [Methods](#methods)
- Posting a status
- Uploading media
- Retrieving a timeline
- Retrieving notifications
- Following a remote user
- Fetching data
- Deleting a status
- Reblogging a status
- Favouriting a status
- Threads (status context)
- Who reblogged/favourited a status
- Following/unfollowing accounts
- Blocking/unblocking accounts
- Creating OAuth apps
- [Entities](#entities)
- Status
- Account
- [Pagination](#pagination)
# Available libraries
- [For Ruby](
- [For Python](
- [For JavaScript](
- [For JavaScript (Node.js)](
# Notes
When an array parameter is mentioned, the Rails convention of specifying array parameters in query strings is meant. For example, a ruby array like `foo = [1, 2, 3]` can be encoded in the params as `foo[]=1&foo[]=2&foo[]=3`. Square brackets can be indexed but can also be empty.
When a file parameter is mentioned, a form-encoded upload is expected.
# Methods
## Posting a new status
**POST /api/v1/statuses**
Form data:
- `status`: The text of the status
- `in_reply_to_id` (optional): local ID of the status you want to reply to
- `media_ids` (optional): array of media IDs to attach to the status (maximum 4)
- `sensitive` (optional): set this to mark the media of the status as NSFW
- `visibility` (optional): either `private`, `unlisted` or `public`
Returns the new status.
**POST /api/v1/media**
Form data:
- `file`: Image to be uploaded
Returns a media object with an ID that can be attached when creating a status (see above).
## Retrieving a timeline
**GET /api/v1/timelines/home**
**GET /api/v1/timelines/mentions**
**GET /api/v1/timelines/public**
**GET /api/v1/timelines/tag/:hashtag**
Returns statuses, most recent ones first. Home timeline is statuses from people you follow, mentions timeline is all statuses that mention you. Public timeline is "whole known network", and the last is the hashtag timeline.
Query parameters:
- `max_id` (optional): Skip statuses younger than ID (e.g. navigate backwards in time)
- `since_id` (optional): Skip statuses older than ID (e.g. check for updates)
## Notifications
**GET /api/v1/notifications**
Returns notifications for the authenticated user. Each notification has an `id`, a `type` (mention, reblog, favourite, follow), an `account` which it came *from*, and in case of mention, reblog and favourite also a `status`.
## Following a remote user
**POST /api/v1/follows**
Form data:
- uri: username@domain of the person you want to follow
Returns the local representation of the followed account.
## Fetching data
**GET /api/v1/statuses/:id**
Returns status.
**GET /api/v1/accounts/:id**
Returns account.
**GET /api/v1/accounts/verify_credentials**
Returns authenticated user's account.
**GET /api/v1/accounts/:id/statuses**
Returns statuses by user. Same options as timeline are permitted.
**GET /api/v1/accounts/:id/following**
Returns users the given user is following.
**GET /api/v1/accounts/:id/followers**
Returns users the given user is followed by.
**GET /api/v1/accounts/relationships**
Returns relationships (`following`, `followed_by`, `blocking`) of the current user to a list of given accounts.
Query parameters:
- `id` (can be array): Account IDs
**GET /api/v1/accounts/search**
Returns matching accounts. Will lookup an account remotely if the search term is in the username@domain format and not yet in the database.
Query parameters:
- `q`: what to search for
- `limit`: maximum number of matching accounts to return
**GET /api/v1/blocks**
Returns accounts blocked by authenticated user.
**GET /api/v1/favourites**
Returns statuses favourited by authenticated user.
## Deleting a status
**DELETE /api/v1/statuses/:id**
Returns an empty object.
## Reblogging a status
**POST /api/v1/statuses/:id/reblog**
Returns a new status that wraps around the reblogged one.
## Unreblogging a status
**POST /api/v1/statuses/:id/unreblog**
Returns the status that used to be reblogged.
## Favouriting a status
**POST /api/v1/statuses/:id/favourite**
Returns the target status.
## Unfavouriting a status
**POST /api/v1/statuses/:id/unfavourite**
Returns the target status.
## Threads
**GET /api/v1/statuses/:id/context**
Returns `ancestors` and `descendants` of the status.
## Who reblogged/favourited a status
**GET /api/v1/statuses/:id/reblogged_by**
**GET /api/v1/statuses/:id/favourited_by**
Returns list of accounts.
## Following and unfollowing users
**POST /api/v1/accounts/:id/follow**
**POST /api/v1/accounts/:id/unfollow**
Returns the updated relationship to the user.
## Blocking and unblocking users
**POST /api/v1/accounts/:id/block**
**POST /api/v1/accounts/:id/unblock**
Returns the updated relationship to the user.
## OAuth apps
**POST /api/v1/apps**
Form data:
- `client_name`: Name of your application
- `redirect_uris`: Where the user should be redirected after authorization (for no redirect, use `urn:ietf:wg:oauth:2.0:oob`)
- `scopes`: This can be a space-separated list of the following items: "read", "write" and "follow" (see [this page]( for details on what the scopes do)
- `website`: (optional) URL to the homepage of your app
Creates a new OAuth app. Returns `id`, `client_id` and `client_secret` which can be used with [OAuth authentication in your 3rd party app](
# Entities
## Status
| Attribute | Description |
| `id` ||
| `uri` | fediverse-unique resource ID |
| `url` | URL to the status page (can be remote) |
| `account` | Account |
| `in_reply_to_id` | null or ID of status it replies to |
| `reblog` | null or Status|
| `content` | Body of the status. This will contain HTML (remote HTML already sanitized) |
| `created_at` ||
| `reblogs_count` ||
| `favourites_count` ||
| `reblogged` | Boolean for authenticated user |
| `favourited` | Boolean for authenticated user |
| `media_attachments` | array of MediaAttachments |
| `mentions` | array of Mentions |
| `application` | Application from which the status was posted |
Media Attachment:
| Attribute | Description |
| `url` | URL of the original image (can be remote) |
| `preview_url` | URL of the preview image |
| `type` | Image or video |
| Attribute | Description |
| `url` | URL of user's profile (can be remote) |
| `acct` | Username for local or username@domain for remote users |
| `id` | Account ID |
| Attribute | Description |
| `name` | Name of the app |
| `website` | Homepage URL of the app |
## Account
| Attribute | Description |
| `id` ||
| `username` ||
| `acct` | Equals username for local users, includes @domain for remote ones |
| `display_name` ||
| `note` | Biography of user |
| `url` | URL of the user's profile page (can be remote) |
| `avatar` | URL to the avatar image |
| `header` | URL to the header image |
| `followers_count` ||
| `following_count` ||
| `statuses_count` ||
# Pagination
API methods that return collections of items can return a `Link` header containing URLs for the `next` and `prev` pages. [Link header RFC](
We use the [Doorkeeper gem for OAuth](, so you can refer to their docs on specifics of the end-points.
The API is divided up into access scopes:
- `read`: Read data
- `write`: Post statuses and upload media for statuses
- `follow`: Follow, unfollow, block, unblock
Multiple scopes can be requested during the authorization phase with the `scope` query param (space-separate the scopes).
\ No newline at end of file
Mastodon builds around the idea of being a server first, rather than a client itself. Similarly to how a XMPP chat server communicates with others and with its own clients, Mastodon takes care of federation to other networks, like other Mastodon or GNU Social instances. So Mastodon provides a REST API, and a 3rd-party app system for using it via OAuth2.
You can get a client ID and client secret required for OAuth [via an API end-point](
From these two, you will need to acquire an access token. It is possible to do using your account's e-mail and password like this:
curl -X POST -d "client_id=CLIENT_ID_HERE&client_secret=CLIENT_SECRET_HERE&grant_type=password&username=YOUR_EMAIL&password=YOUR_PASSWORD" -Ss
The response will be a JSON object containing the key `access_token`. Use that token in any API requests by setting a header like this:
curl --header "Authorization: Bearer ACCESS_TOKEN_HERE" -sS
Please note that the password-based approach is not recommended especially if you're dealing with other user's accounts and not just your own. Usually you would use the authorization grant approach where you redirect the user to a web page on the original site where they can login and authorize the application and are then redirected back to your application with an access code.
# Authentication
Make sure that you allow your users to specify the domain they want to connect to before login. Use that domain to acquire a client id/secret for OAuth2 and then proceed with normal OAuth2 also using that domain to build the URLs.
In my opinion it is easier for people to understand what is being asked of them if you ask for a `username@domain` type input, since it looks like an e-mail address. Though the username part is not required for anything in the OAuth2 process. Once the user is logged in, you get information about the logged in user from `/api/v1/accounts/verify_credentials`
# Usernames
Make sure that you make it possible to see the `acct` of any user in your app (since it includes the domain part for remote users), people must be able to tell apart users from different domains with the same username.