rearvue

A personal social media aggregator / nostalgia engine.

Collates your posts from around the web and brings them home.

Very, very rough and ready.

HTTPS / TLS (certificate bundle)

On hosts where the Python/OpenSSL default store is missing or stale, outbound HTTPS (RSS, APIs, mirroring) can fail verification.

Default behavior: On import, the rearvue package sets SSL_CERT_FILE and REQUESTS_CA_BUNDLE to certifi’s CA bundle only if neither variable is already set. Your environment or process manager can override both.

Production: Prefer the system trust store or your platform’s documented variables, for example:

export SSL_CERT_FILE=/path/to/ca-bundle.crt
# optional; requests/urllib3 also honor:
export REQUESTS_CA_BUNDLE="$SSL_CERT_FILE"

Do not disable TLS verification in application code; fix the CA configuration instead.

---

Application configuration (settings_server)

Runtime settings live in src/rearvue/settings_server.py, which is not committed (see .gitignore). Create it next to settings.py and define at least:

Variable	Purpose
SECRET_KEY	Django secret key
DEBUG	True / False
ALLOWED_HOSTS	List of hostnames the app answers on
DATABASES	Django database config; SQLite may use NAME with __BASE_DIR__ (replaced at runtime)
DATA_STORE	Filesystem root where mirrored media and generated rss.xml files are written (must be writable)
LOG_LOCATION	Path to the rotating log file (non-debug / non-runserver)
STATIC_ROOT, MEDIA_ROOT	Static and media roots for collectstatic / uploads
DEFAULT_DOMAIN_PROTOCOL	"http" or "https" — used when building OAuth callback URLs if the site has no alt_domain
FLICKR_KEY, FLICKR_SECRET	Flickr API key and secret (Flickr App Garden)
INSTAGRAM_KEY, INSTAGRAM_SECRET	Meta app Instagram App ID and Instagram App Secret from the dashboard (Instagram Login / Business Login)
FEEDS_SERVER	Required by django-feed-reader for RSS fetching; set per that package’s deployment docs

Optional overrides (read in rearvue/settings.py via getattr where noted):

Variable	Purpose
INSTAGRAM_REDIRECT_URI	Full OAuth redirect URL if it must differ from the auto-built {protocol}://{host}/rvadmin/instagram_oauth_return/
INSTAGRAM_GRAPH_API_VERSION	Default v22.0
INSTAGRAM_OAUTH_SCOPES	Default instagram_business_basic
FACEBOOK_ACCESS_TOKEN	Optional; not used by the current Instagram Login flow

Install dependencies from src/requirements.txt, run migrations from src/, create a superuser, and collect static files as usual for Django.

---

Sites and domains (RVDomain)

RearVue picks the site from the HTTP Host header (rearvue.utils.page / admin_page). For each hostname you serve:

1. In Django admin, open RV domains (app rvsite) and add a row.
2. Set name to the bare hostname visitors use (e.g. example.com), matching ALLOWED_HOSTS.
3. Set alt_domain (optional) to the full public origin if it differs, e.g. https://www.example.com. Flickr and Instagram OAuth callbacks prefer this when set; if the value has no scheme, DEFAULT_DOMAIN_PROTOCOL is prepended.
4. Set owner to a Django user (typically your superuser).

Wrong or missing name / alt_domain leads to 404 in /rvadmin/ or OAuth redirect mismatches.

---

Admin UIs

• Public site: / (per rvsite URLs).
• Django admin: /admin/ — use this to create and edit RVDomain, RVService, and content models.
• Site admin (RearVue): /rvadmin/ — superuser only; lists services for the current domain and runs connect flows.

---

Configuring each service

RVService rows tie a source to a domain. type must match exactly what the workers expect (lowercase): rss, twitter, flickr, instagram.

RSS (type = rss)

Used for any feed django-feed-reader can poll (blogs, GitHub releases, Mastodon RSS, etc.).

1. In Django admin, add RV services → choose domain, set type to rss, set name as you like.
2. Set auth token to the full feed URL (the code stores the feed URL in this field).
3. Ensure live is checked when you want update_content to crawl it.
4. Run python manage.py update_content (or your cron equivalent) so update_rss runs; feeds are mirrored when enclosures exist.

Pagination depth depends on the feed itself.

Twitter (type = twitter)

Live Twitter API updates are not implemented; import is via X/Twitter archive.

1. In Django admin, add RV services with type twitter, correct domain, and username set to your handle (no @) — used when building status URLs during import.
2. Visit /rvadmin/twitter_connect/<service_id>/ (the id is the primary key of that RVService).
3. Upload the archive tweets.js file using the “Import Twitter Archive” form (the file should start like window.YTD.tweets.part0 = as produced by the archive).

The “Connect” action in the template is a stub (“Broken :(”).

After import, update_content runs mirroring and link discovery for Twitter items when you skip nothing.

Flickr (type = flickr)

1. Configure FLICKR_KEY and FLICKR_SECRET in settings_server.
2. Open /rvadmin/flickr_connect/new/ (POST “Connect”) to create a service and start OAuth, or open /rvadmin/flickr_connect/<id>/ for an existing Flickr service.
3. Flickr redirects back to /rvadmin/flickr_return/?svc=<id>; callback URL is built from alt_domain or DEFAULT_DOMAIN_PROTOCOL + RVDomain.name. Register the same callback URL in the Flickr app settings if required.
4. After authorization, tokens are stored on the RVService row.

Instagram (type = instagram)

Requires a Business or Creator Instagram account and a Meta developer app with Instagram product and Instagram API with Instagram Login (Business Login) configured.

1. In the Meta app, add Valid OAuth Redirect URIs that match what RearVue will send. Typically that is exactly:
   • https://<your-host>/rvadmin/instagram_oauth_return/
     or, if you use INSTAGRAM_REDIRECT_URI, that exact string (no trailing slash mismatch).
2. Set INSTAGRAM_KEY and INSTAGRAM_SECRET in settings_server to the app’s Instagram App ID and Instagram App Secret from the Business Login section of the dashboard.
3. Open /rvadmin/instagram_connect/new/, submit Connect, complete Instagram login, and return to the site while still logged into Django (same browser session).
4. Tokens and IG user id are stored on the service; run update_content to sync. The API returns up to on the order of 10,000 recent media per account; stories are not the same as feed media in the API.

To wipe all Instagram services, items, and mirrored files and start over:

cd src && python manage.py reset_instagram_graph --confirm

---

Refreshing content (cron)

From the src/ directory:

python manage.py update_content

Flags: --skip-rss, --skip-twitter, --skip-instagram, --skip-flickr, --skip-cleanup (skips domain year bounds and per-domain media/<domain>/rss.xml generation).

Typical production use: run without skips on a schedule; ensure DATA_STORE (and DB) are persistent.

---

Supported services (summary)

Service	Imports archive	Live updates
Twitter	✅	❌
Flickr	✅	✅
RSS	✅ (*)	✅
Instagram	✅	✅

* Full history only if the feed exposes it (pagination).

Other sources can be aggregated indirectly via RSS (e.g. GitHub, Mastodon) if they publish a feed URL.