docs: Release the 0.22 docs

4 files changed, 111 insertions, 0 deletions

diff --git a/docs/versioned_docs/version-v0.22.0/07-Development/01-setup.md b/docs/versioned_docs/version-v0.22.0/07-Development/01-setup.md
new file mode 100644
index 00000000..6d359234
--- /dev/null
+++ b/docs/versioned_docs/version-v0.22.0/07-Development/01-setup.md
@@ -0,0 +1,63 @@
+# Setup
+
+## Manual Setup
+### First Setup
+
+- You'll need to prepare the environment variables for the dev env.
+- Easiest would be to set it up once in the root of the repo and then symlink it in each app directory (e.g. `/apps/web`, `/apps/workers`) and also `/packages/db`.
+- Start by copying the template by `cp .env.sample .env`.
+- The most important env variables to set are:
+  - `DATA_DIR`: Where the database and assets will be stored. This is the only required env variable. You can use an absolute path so that all apps point to the same dir.
+  - `NEXTAUTH_SECRET`: Random string used to sign the JWT tokens. Generate one with `openssl rand -base64 36`. Logging in will not work if this is missing!
+  - `MEILI_ADDR`: If not set, search will be disabled. You can set it to `http://127.0.0.1:7700` if you run meilisearch using the command below.
+  - `OPENAI_API_KEY`: If you want to enable auto tag inference in the dev env.
+- run `pnpm run db:migrate` in the root of the repo to set up the database.
+
+### Dependencies
+
+#### Meilisearch
+
+Meilisearch is the provider for the full text search. You can get it running with `docker run -p 7700:7700 getmeili/meilisearch:v1.11.1`.
+
+Mount persistent volume if you want to keep index data across restarts. You can trigger a re-index for the entire items collection in the admin panel in the web app.
+
+#### Chrome
+
+The worker app will automatically start headless chrome on startup for crawling pages. You don't need to do anything there.
+
+### Web App
+
+- Run `pnpm web` in the root of the repo.
+- Go to `http://localhost:3000`.
+
+> NOTE: The web app kinda works without any dependencies. However, search won't work unless meilisearch is running. Also, new items added won't get crawled/indexed unless workers are running.
+
+### Workers
+
+- Run `pnpm workers` in the root of the repo.
+
+### iOS Mobile App
+
+- `cd apps/mobile`
+- `pnpm exec expo prebuild --no-install` to build the app.
+- Start the ios simulator.
+- `pnpm exec expo run:ios`
+- The app will be installed and started in the simulator.
+
+Changing the code will hot reload the app. However, installing new packages requires restarting the expo server.
+
+### Browser Extension
+
+- `cd apps/browser-extension`
+- `pnpm dev`
+- This will generate a `dist` package
+- Go to extension settings in chrome and enable developer mode.
+- Press `Load unpacked` and point it to the `dist` directory.
+- The plugin will pop up in the plugin list.
+
+In dev mode, opening and closing the plugin menu should reload the code.
+
+
+## Docker Dev Env
+
+If the manual setup is too much hassle for you. You can use a docker based dev environment by running `docker compose -f docker/docker-compose.dev.yml up` in the root of the repo. This setup wasn't super reliable for me though.
diff --git a/docs/versioned_docs/version-v0.22.0/07-Development/02-directories.md b/docs/versioned_docs/version-v0.22.0/07-Development/02-directories.md
new file mode 100644
index 00000000..54552402
--- /dev/null
+++ b/docs/versioned_docs/version-v0.22.0/07-Development/02-directories.md
@@ -0,0 +1,28 @@
+# Directory Structure
+
+## Apps
+
+| Directory                | Description                                            |
+| ------------------------ | ------------------------------------------------------ |
+| `apps/web`               | The main web app                                       |
+| `apps/workers`           | The background workers logic                           |
+| `apps/mobile`            | The react native based mobile app                      |
+| `apps/browser-extension` | The browser extension                                  |
+| `apps/landing`           | The landing page of [hoarder.app](https://hoarder.app) |
+
+## Shared Packages
+
+| Directory         | Description                                                                  |
+| ----------------- | ---------------------------------------------------------------------------- |
+| `packages/db`     | The database schema and migrations                                           |
+| `packages/trpc`   | Where most of the business logic lies built as TRPC routes                   |
+| `packages/shared` | Some shared code between the different apps (e.g. loggers, configs, assetdb) |
+
+## Toolings
+
+| Directory            | Description             |
+| -------------------- | ----------------------- |
+| `tooling/typescript` | The shared tsconfigs    |
+| `tooling/eslint`     | ESlint configs          |
+| `tooling/prettier`   | Prettier configs        |
+| `tooling/tailwind`   | Shared tailwind configs |
diff --git a/docs/versioned_docs/version-v0.22.0/07-Development/03-database.md b/docs/versioned_docs/version-v0.22.0/07-Development/03-database.md
new file mode 100644
index 00000000..40e2d164
--- /dev/null
+++ b/docs/versioned_docs/version-v0.22.0/07-Development/03-database.md
@@ -0,0 +1,11 @@
+# Database Migrations
+
+- The database schema lives in `packages/db/schema.ts`.
+- Changing the schema, requires a migration.
+- You can generate the migration by running `pnpm drizzle-kit generate:sqlite` in the `packages/db` dir.
+- You can then apply the migration by running `pnpm run migrate`.
+
+
+## Drizzle Studio
+
+You can start the drizzle studio by running `pnpm db:studio` in the root of the repo.
diff --git a/docs/versioned_docs/version-v0.22.0/07-Development/04-architecture.md b/docs/versioned_docs/version-v0.22.0/07-Development/04-architecture.md
new file mode 100644
index 00000000..5a864034
--- /dev/null
+++ b/docs/versioned_docs/version-v0.22.0/07-Development/04-architecture.md
@@ -0,0 +1,9 @@
+# Architecture
+
+![Architecture Diagram](/img/architecture/arch.png)
+
+- Webapp: NextJS based using sqlite for data storage.
+- Workers: Consume the jobs from sqlite based job queue and executes them, there are three job types:
+  1. Crawling: Fetches the content of links using a headless chrome browser running in the workers container.
+  2. OpenAI: Uses OpenAI APIs to infer the tags of the content.
+  3. Indexing: Indexes the content in meilisearch for faster retrieval during search.