diff --git a/README.md b/README.md
index 6e9d37ba..dc21c1ae 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,6 @@
 # Sci-ModoM
 
-## Production Setup
-
-The recommended way to run the application in production is a container setup. See [Container Setup](docker/CONTAINER_SETUP.md) for details.
-
-## Development Setup
-
-### Overview
+## Overview
 
 The application consists of the following components:
 
@@ -14,11 +8,19 @@ The application consists of the following components:
 - Server: A REST-API backend using Python3 and Flask
 - Client: A Web-GUI using Vue.js
 
-All components can be run locally.
+### General setup
+
+At lauchtime, the app uses tables defined in `SetupService` to populate the database. These tables **must** exist. They are used both in development and production. They **must** be located under `IMPORT_PATH` (defaults to _server/import_, or specify in _.env_ file).
+
+## Production setup
+
+The recommended way to run Sci-ModoM in production is to use podman compose to create, manage, and deploy the application and the database containers, see [Container Setup](docker/CONTAINER_SETUP.md) for details.
+
+## Development setup
 
 ### Database setup
 
-Set up a MariaDB database. One way to do this is to run a MariaDB container image. See the _Development Setup_ section in [Container Setup](docker/CONTAINER_SETUP.md) on how to do this. A local MariaDB database is also possible with similar setup steps. The following steps are required:
+Set up a MariaDB database. One way to do this is to run a MariaDB container image, see the _Development Setup_ section in [Container Setup](docker/CONTAINER_SETUP.md). The following steps are required:
 
 - Create a database
 - Create a database user
@@ -26,7 +28,7 @@ Set up a MariaDB database. One way to do this is to run a MariaDB container imag
 
 ### Server setup
 
-Create a Python3 virtual environment with your preferred method and activate it, _e.g._:
+Create a Python3 virtual environment with your preferred method and activate it _e.g._:
 
 ```bash
 python3 -m venv ~/.virtualenvs/scimodom
@@ -65,22 +67,75 @@ HTTP_PUBLIC_URL=https://sci-modom.my-site.org
 **Important:** If the host name _localhost_ is used in the `DATABASE_URI` the database driver will assume that the database is contacted using a named
 socket. That will not work if a container is used!
 
-Now the database schema can be initialized by executing this command under the _server_ directory:
+In addition to _server/.env/_, a Flask-specific configuration in _server/.flaskenv_ can be used:
+
+```
+FLASK_APP=src/scimodom/app
+FLASK_DEBUG=True
+```
+
+Now the database container must be started under the _docker_ directory:
+
+```bash
+# under scimodom/docker
+docker compose -f docker-compose-db-only.yml up -d
+```
+
+The database schema can then be initialized by executing this command under the _server_ directory:
 
 ```bash
+# under scimodom/server
 alembic upgrade head
 ```
 
 The API backend can be started with:
 
 ```bash
-# under client/server
+# under scimodom/server
 flask run
 ```
 
 Most Python IDEs can run this process in the integrated debugger.
 
-### Development
+### Client setup
+
+To bring up the frontend, go to the _client_ directory and execute:
+
+```bash
+# under scimodom/client
+npm install # first time install
+npm run dev
+```
+
+Now the application is available here:
+
+- http://localhost:5173/
+
+To test the bundled frontend, run under the _client_ directory:
+
+```bash
+npm run build
+```
+
+This will populate the folder _client/dist_ with the bundled static HTML/JavaScript code as it should be deployed in production.
+The server can now also serve this code. The complete application is now also available under the Flask development server URL:
+
+- http://127.0.0.1:5000
+
+#### Local login
+
+Registration is completed via email. These environment variables are required:
+
+```
+SMTP_SERVER=outgoing-email-server.my-site.org
+SMTP_FROM_ADDRESS=sci-modom-admin@my-site.org
+```
+
+**Important:** Make sure to build the frontend beforehand (see above). After receiving the registration link via email, click on the link, but change the default port `5173` to `5000` to point to the Flask development server URL.
+
+Alternatively, you can patch the database (add user).
+
+### Development hints
 
 For development:
 
@@ -112,33 +167,13 @@ To execute the Python tests run under the _server_ directory:
 pytest tests
 ```
 
-### Client setup
-
-To bring up the frontend, go to the _client_ directory and execute:
-
-```bash
-npm install
-npm run dev
-```
-
-Now the application is available here:
-
-- http://localhost:5173/
-
-To test the bundled frontend, run under the _client_ directory:
+To run pytest from command line, adjust the `PYTHONPATH` _e.g._
 
 ```bash
-npm run build
+export PYTHONPATH="${PYTHONPATH}:/path/to/scimodom/server/tests/"
 ```
 
-This will populate the folder _client/dist_ with the bundled static HTML/JavaScript code as it should be deployed in production.
-The server can now also serve this code. The complete application is now also available under the Flask development server URL:
-
-- http://127.0.0.1:5000
-
-## Development Hints
-
-### Database schema updates
+#### Database schema updates
 
 The database schema is tracked using Alembic. Changes to the database must be coded at two locations: