Remark42 is a self-hosted, lightweight, and simple (yet functional) comment engine, which doesn't spy on users. It can be embedded into blogs, articles or any other place where readers add comments.

• Social login via Google, Facebook, Github and Yandex
• Multi-level nested comments with both tree and plain presentations
• Import from disqus
• Markdown support
• Moderator can remove comments and block users
• Voting, pinning and verification system
• Sortable comments
• Extractor for recent comments, cross-post
• RSS for all comments and each post
• Export data to json with automatic backups
• No external databases, everything embedded in a single data file
• Fully dockerized and can be deployed in a single command
• Clean, lightweight and fully customizable UI
• Multi-site mode from a single instance
• Integration with automatic ssl via nginx-le
• Privacy focused

• copy provided docker-compose.yml and customize for your needs
• prepare user id for container export USER=`id -u $USER`
• make sure you don't keep DEV_PASSWD=something... for any non-development deployments
• pull prepared images from docker hub and start - docker-compose pull && docker-compose up -d
• alternatively compile from sources - docker-compose build && docker-compose up -d

• command line parameters are long form --<key>=value, i.e. --site=https://demo.remark42.com
• multi parameters separated by , in the environment or repeated with command line key, like --site=s1 --site=s2 ...
• required parameters have to be presented in the environment or provided in command line

Most of the parameters have sane defaults and don't require customization. There are only a few parameters user has to define:

1. SECRET - secret key, can be any long and hard-to-guess string.
2. REMARK_URL - url pointing to your remark42 server, i.e. https://demo.reamark42.com
3. At least one pair of AUTH_<PROVIDER>_CID and AUTH_<PROVIDER>_CSEC defining oauth2 provider(s)

The minimal docker-compose.yml has to include all required parameters:

version: '2'

services:
    remark42:
        image: umputun/remark42:latest
        restart: always
        container_name: "remark42"
        environment:
            - REMARK_URL=https://demo.remark42.com  # url pointing to your remark42 server
            - SECRET=abcd-123456-xyz-$%^&           # secret key
            - AUTH_GITHUB_CID=12345667890           # oauth2 client ID
            - AUTH_GITHUB_CSEC=abcdefg12345678      # oauth2 client secret
            - USER=1001                             # UID on the host machine, i.e `id -u`
        volumes:
            - ./var:/srv/var                        # persistent volume to store all remark42 data 

Authentication handled by external providers. You should setup oauth2 for all (or some) of them to allow users to make comments. It is not mandatory to have all of them, but at least one should be correctly configured.

1. Create a new project: https://console.developers.google.com/project
2. Choose the new project from the top right project dropdown (only if another project is selected)
3. In the project Dashboard center pane, choose "API Manager"
4. In the left Nav pane, choose "Credentials"
5. In the center pane, choose "OAuth consent screen" tab. Fill in "Product name shown to users" and hit save.
6. In the center pane, choose "Credentials" tab.
   • Open the "New credentials" drop down
   • Choose "OAuth client ID"
   • Choose "Web application"
   • Application name is freeform, choose something appropriate
   • Authorized origins is your domain ex: https://remark42.mysite.com
   • Authorized redirect URIs is the location of oauth2/callback constructed as domain + /auth/google/callback, ex: https://remark42.mysite.com/auth/google/callback
   • Choose "Create"
7. Take note of the Client ID and Client Secret

instructions for google oauth2 setup borrowed from oauth2_proxy

1. Create a new "OAuth App": https://github.com/settings/developers
2. Fill "Application Name" and "Homepage URL" for your site
3. Under "Authorization callback URL" enter the correct url constructed as domain + /auth/github/callback. ie https://remark42.mysite.com/auth/github/callback
4. Take note of the Client ID and Client Secret

1. From https://developers.facebook.com select "My Apps" / "Add a new App"
2. Set "Display Name" and "Contact email"
3. Choose "Facebook Login" and then "Web"
4. Set "Site URL" to your domain, ex: https://remark42.mysite.com
5. Under "Facebook login" / "Settings" fill "Valid OAuth redirect URIs" with your callback url constructed as domain + /auth/facebook/callback
6. Select "App Review" and turn public flag on. This step may ask you to provide a link to your privacy policy.

1. Create a new "OAuth App": https://oauth.yandex.com/client/new
2. Fill "App name" for your site
3. Under Platforms select "Web services" and enter "Callback URI #1" constructed as domain + /auth/yandex/callback. ie https://remark42.mysite.com/auth/yandex/callback
4. Select Permissions. You need following permissions only from the "Yandex.Passport API" section:
   • Access to user avatar
   • Access to username, first name and surname, gender
5. Fill out the rest of fields if needed
6. Take note of the ID and Password

For more details refer to Yandex OAuth and Yandex.Passport API documentation.

1. Disqus provides an export of all comments on your site in a g-zipped file. This is found in your Moderation panel at Disqus Admin > Setup > Export. The export will be sent into a queue and then emailed to the address associated with your account once it's ready. Direct link to export will be something like https://<siteud>.disqus.com/admin/discussions/export/. See importing-exporting for more details.
2. Move this file to your remark42 host within ./var and unzip, i.e. gunzip <disqus-export-name>.xml.gz.
3. Run import command - docker-compose exec remark42 /srv/import-disqus.sh <disqus-export-name>.xml <your site id>

Remark42 by default makes daily backup files under ${BACKUP_PATH} (default ./var/backup). Backups kept up to ${MAX_BACKUP_FILES} (default 10). Each backup file contains exported and gzipped content, i.e., all comments. At any point, the user can restore such backup and revert all comments to the desirable state. Note: restore procedure cleans the current data store and replaces all comments with comments from the backup file.

For safety and security reasons restore functionality not exposed outside of your server by default. The recommended way to restore from the backup is to use provided scripts/restore-backup.sh. It can run inside the container:

docker-compose exec remark42 /srv/restore-backup.sh {backup-filename.gz} {your site id}

One special case for backup/restore is schema migration. Some versions or remark42 may extend or change the schema and for such upgrades migration required. Provided migration script scripts/migrate-data.sh makes a fresh backup and then loads it back to your remark42 instance.

docker-compose exec remark42 /srv/migrate-data.sh {your site id}

In addition to automatic backups user can make a backup manually. This command makes userbackup-{site id}-{timestamp}.gz

docker-compose exec remark42 /srv/create-backup.sh {your site id}

Backup file is a text file with all exported comments separated by EOL. Each backup record is a valid json with all key/value unmarshaled from Comment struct (see below).

Admins/moderators should be defined in docker-compose.yml as a list of user IDs or passed in the command line.

    environment:
        - ADMIN=github_ef0f706a79cc24b17bbbb374cd234a691a034128,github_dae9983158e9e5e127ef2b87a411ef13c891e9e5

To get user id just login and click on your username or any other user you want to promote to admins. It will expand login info and show full user ID.

Frontend part is building automatically along with backend if you use docker-compose.

For manual building:

• install Node.js 8 or higher;
• run npm install inside ./web;
• run npm run build there;
• result files will be saved in ./web/public.

For development mode use npm start instead of npm run build. In this case webpack will serve files using webpack-dev-server on localhost:8080.

URLs for development:

• localhost:8080 — page with embedded script from REMARK_URL (default: https://demo.remark42.com);
• localhost:8080/dev.html — page with embedded script from local folder;
• localhost:8080/last-comments.html — page with embedded script for last comments;
• localhost:8080/counter.html — page with embedded script for counter with examples.

It's a main widget which renders list of comments.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID',
    url: 'PAGE_URL', // optional param; if it isn't defined window.location.href will be used
    max_shown_comments: 10, // optional param; if it isn't defined default value (15) will be used 
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/embed.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add this node in the place where you want to see Remark42 widget:

<div id="remark42"></div>

After that widget will be rendered inside this node.

It's a widget which renders list of last comments from your site.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID', 
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/last-comments.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add this node in the place where you want to see last comments widget:

<div class="remark42__last-comments" data-max="50"></div>

data-max sets the max amount of comments (default: 15).

It's a widget which renders a number of comments for the specified page.

Add this snippet to the bottom of web page:

<script>
  var remark_config = {
    site_id: 'YOUR_SITE_ID', 
  };

  (function() {
    var d = document, s = d.createElement('script');
    s.src = '/web/counter.js'; // prepends this address with domain where remark42 is placed
    (d.head || d.body).appendChild(s);
  })();
</script>

And then add a node like this in the place where you want to see a number of comments:

<span class="remark42__counter" data-url="https://domain.com/path/to/article/"></span>

You can use as many nodes like this as you need to. The script will found all them by the class remark__counter, and it will use data-url attribute to define the page with comments.

Also script can uses url property from remark_config object, or window.location.href if nothing else is defined.

• GET /auth/{provider}/login?from=http://url&session=1 - perform "social" login with one of supported providers and redirect to url. Presence of session (any non-zero value) change the default cookie expiration and makes them session-only.
• GET /auth/logout - logout

type User struct {
    Name    string `json:"name"`
    ID      string `json:"id"`
    Picture string `json:"picture"`
    Admin   bool   `json:"admin"`
    Blocked bool   `json:"block"`
    Verified bool  `json:"verified"`
}

currently supported providers are google, facebook, github and yandex

• POST /api/v1/comment - add a comment. auth required

type Comment struct {
    ID        string          `json:"id"`      // comment ID, read only
    ParentID  string          `json:"pid"`     // parent ID
    Text      string          `json:"text"`    // comment text, after md processing
    Orig      string          `json:"orig"`    // original comment text
    User      User            `json:"user"`    // user info, read only
    Locator   Locator         `json:"locator"` // post locator
    Score     int             `json:"score"`   // comment score, read only
    Votes     map[string]bool `json:"votes"`   // comment votes, read only
    Timestamp time.Time       `json:"time"`    // time stamp, read only
    Pin       bool            `json:"pin"`     // pinned status, read only
    Delete    bool            `json:"delete"`  // delete status, read only
}

type Locator struct {
    SiteID string `json:"site"`     // site id
    URL    string `json:"url"`      // post url
}

• POST /api/v1/preview - preview comment in html. Body is Comment to render

• GET /api/v1/find?site=site-id&url=post-url&sort=fld&format=tree|plain - find all comments for given post

This is the primary call used by UI to show comments for given post. It can return comments in two formats - plain and tree. In plain format result will be sorted list of Comment. In tree format this is going to be tree-like object with this structure:

type Tree struct {
    Nodes []Node `json:"comments"`
    Info  store.PostInfo `json:"info,omitempty"`
}

type Node struct {
    Comment store.Comment `json:"comment"`
    Replies []Node        `json:"replies,omitempty"`
}

Sort can be time, active or score. Supported sort order with prefix -/+, i.e. -time. For tree mode sort will be applied to top-level comments only and all replies always sorted by time.

• PUT /api/v1/comment/{id}?site=site-id&url=post-url - edit comment, allowed once in 5min since creation

  Content-Type: application/json

  {
    "text": "edit comment blah http://radio-t.com 12345",
    "summary": "fix blah"
  }

• GET /api/v1/last/{max}?site=site-id - get up to {max} last comments

• GET /api/v1/id/{id}?site=site-id - get comment by comment id

• GET /api/v1/comments?site=site-id&user=id&limit=N - get comment by user id, returns response object

  type response struct {
      Comments []store.Comment  `json:"comments"`
      Count    int              `json:"count"`
  }{}

• GET /api/v1/count?site=site-id&url=post-url - get comment's count for {url}

• GET /api/v1/list?site=site-id&limit=5&skip=2 - list commented posts, returns array or PostInfo, limit=0 will return all posts

  type PostInfo struct {
      URL   string      `json:"url"`
      Count int         `json:"count"`
      ReadOnly bool     `json:"read_only,omitempty"`
      FirstTS time.Time `json:"first_time,omitempty"`
      LastTS  time.Time `json:"last_time,omitempty"`
  }

• GET /api/v1/user - get user info, auth required

• PUT /api/v1/vote/{id}?site=site-id&url=post-url&vote=1 - vote for comment. vote=1 will increase score, -1 decrease. auth required

• GET /api/v1/userdata?site=site-id - export all user data to gz stream auth required

• POST /api/v1/deleteme?site=site-id - request deletion of user data. auth required

• GET /api/v1/config?site=site-id - returns configuration (parameters) for given site

  type config struct {
      Version       string   `json:"version"`
      EditDuration  int      `json:"edit_duration"` // seconds
      Admins        []string `json:"admins"`
      Auth          []string `json:"auth_providers"`
      LowScore      int      `json:"low_score"`
      CriticalScore int      `json:"critical_score"`
  }

• GET /api/v1/info?site=site-idd&url=post-ur - returns PostInfo for site and url

• GET /api/v1/rss/post?site=site-id&url=post-url - rss feed for a post
• GET /api/v1/rss/site?site=site-id - rss feed for given site

• DELETE /api/v1/admin/comment/{id}?site=site-id&url=post-url - delete comment by id.
• PUT /api/v1/admin/user/{userid}?site=site-id&block=1 - block or unblock user.
• GET api/v1/admin/blocked&site=site-id - list of blocked user ids.

  type BlockedUser struct {
      ID        string    `json:"id"`
      Name      string    `json:"name"`
      Timestamp time.Time `json:"time"`
  }

• GET /api/v1/admin/export?site=side-id&mode=[stream|file] - export all comments to json stream or gz file.
• POST /api/v1/admin/import?site=side-id - import comments from the backup.
• PUT /api/v1/admin/pin/{id}?site=site-id&url=post-url&pin=1 - pin or unpin comment.
• GET /api/v1/admin/user/{userid}?site=site-id - get user's info.
• DELETE /api/v1/admin/user/{userid}?site=site-id - delete all user's comments.
• PUT /api/v1/admin/readonly?site=site-id&url=post-url&ro=1 - set read-only status
• PUT /api/v1/admin/verify/{userid}?site=site-id&verified=1 - set verified status
• GET /api/v1/admin/deleteme?token=token - process deleteme user's request

all admin calls require auth and admin privilege

• Remark42 is trying to be very sensitive to any private or semi-private information.
• Authentication requesting the lowest (minimal) possible scope from providers. All extra information returned by them dropped immediately and not stored in any form.
• Generally remark42 keeps user id, username and avatar link only. None of these fields exposed directly - id and name hashed, avatar proxied.
• There is no tracking of any sort.
• Login mechanic uses JWT stored in a cookie (httpOnly, secured). The second cookie (XSRF_TOKEN) is a random id preventing Cross-Site Request Forgery
• There is no cross-site login, i.e., user's behavior can't be analyzed across independent sites running remark42.
• There are no third-party analytic services involved.
• User can request all information remark42 knows about and export to gz file.
• Supported complete cleanup of all information related to user activity on demand.
• Cookie lifespan can be restricted to session-only.
• All potentially sensitive data stored by remark42 hashed and encrypted.

• Data stored in boltdb (embedded key/value database) files under STORE_BOLT_PATH
• Each site stored in a separate boltbd file.
• In order to migrate/move remark42 to another host boltbd files as well as avatars directory AVATAR_FS_PATH should be transferred.
• Automatic backup process runs every 24h and exports all content in json-like format to backup-remark-YYYYMMDD.gz.
• Authentication implemented with jwt stored in a cookie. It uses HttpOnly, secure cookies.
• All heavy REST calls cached internally in LRU cache limited by CACHE_MAX_ITEMS and CACHE_MAX_SIZE.
• User's activity throttled globally (up to 1000 simultaneous requests) and limited locally (per user, usually up to 10 req/sec)
• Request timeout set to 60sec
• Development mode (--dev-password set) allows to test remark42 without social login and with admin privileges. Adds basic-auth for username: dev, password: ${DEV_PASSWD}. should not be used in production deployment
• User can vote for the comment multiple times but only to change the vote. Double-voting not allowed.
• User can edit comments in 5 mins window after creation.
• User ID hashed and prefixed by oauth provider name to avoid collisions and potential abuse.
• All avatars resized and cached locally to prevent rate limiters from google/github/facebook/yandex.
• Images can be proxied (IMG_PROXY=true) to prevent mixed http/https.
• Docker build uses publicly available base images.