foggy-backend 0.1.2

Foggy backend with REST API for the foggy sync iOS app

Foggy

Foggy keeps photos and videos shot with a smartphone in sync between devices. It does so without requiring a separate app to delete, update or add files. On an iPhone you will install the app and keep on using the built-in Photos app to edit and sort your photos and videos. When space gets low on the device you move the photos to an archive on the backend and clear the phone memory knowing that you have copies on the server.

The name foggy is a silly reference to fog as compared to cloud. Fog is something that you can see and almost touch. A cloud is high up in the sky and you have no idea what happens up there.

Test it

At the moment Foggy is much a work in progress and requires more testing and development before it can be trusted as your only handler of photos and videos. But you can try it out with your own server (mostly tested in Arch and Debian).

First we have to setup the launching user to be able to create the Avahi multicast dns file. On most linux distributions you could do this:

sudo setfacl -m u:$USER:rwx /etc/avahi/services

If you do not have acl installed, run these set of commands:

sudo groupadd avahi-config
sudo usermod -a -G $USER avahi-config
sudo chown root:avahi-conf /etc/avahi/services
sudo chmod 775 /etc/avahi/services

Then we install the backend and create a service:

sudo python3 -m pip install foggy-backend
sudo sh -c "cat <<EOF > /etc/systemd/system/foggy@.service
[Unit]
Description=foggy syncronization backend service

[Service]
User=%i
Type=simple
ExecStart=/usr/local/bin/foggy
Restart=on-failure
RestartSec=5s

[Install]
WantedBy=multi-user.target
EOF"

If you do not have the possibility to run pip you can symlink the launch-script like so:

ln -s scripts/run_server /usr/local/bin/
chmod +x scripts/run_server

Finally reload the systemd daemon and start the service:

sudo systemctl daemon-reload
sudo systemctl enable foggy@$USER
sudo systemctl start foggy@$USER

After this you can build the Xcode-project from the ios folder and deploy it to a simulator or physical device. The app should then connect to the server and you should be able to sync the device with the server.

The backend will create a folder in $HOME/pictures/ for each device that is synced. If you want this folder elsewhere it is recommended to symlink it to another directory or set the environment variable FOGGY_ROOT.

Raspbian instruction

With a clean Raspbian installation here is what you do to get going. Login through ssh or directly on the pi and add your user. E.g.:

sudo useradd -m -G sudo mrfog
sudo passwd mrfog

Logout and login with your new user. Run these commands to update and initiate the system:

sudo userdel pi
sudo apt update ; sudo apt upgrade
sudo apt install git

After that you can follow above instructions.

Docker

In case you want to deploy your app with docker, here is an example docker-compose.yaml file:

version: "3"
services:
    backend:
        image: python:3.8
        container_name: foggy
        environment:
            PYTHONPATH: /foggy_backend
        command: python -m foggy
        ports:
            - 21210:21210
        logging:
            driver: journald
            options:
                tag: "foggy"
        volumes:
            - ./backend:/backend
            - /etc/avahi/services/:/etc/avahi/services/

Current limitations

Here are some limitations that needs to be taken into consideration:

• Use only one active backend server on the local network. There is no way to control which backend is connected so you will not know which server gets connected.

Developers documentation

1. Server and app connection

The connection between the server and the app is initiated with a

2. Vendoring

There are two dependencies that have been vendored as compared to creating a virtual environment or using system packages, see backend/foggy/vendor. The reason for choosing this approach is a much simpler configuration. There is really no need for docker or virtualenvs, just run it with python -m....

3. Style and formatting

The code is formatted with black and linted with pylint.

Format all code with black:

black  --target-version py38 --exclude 'vendor/*.?' rednas tests

4. Test packaging

pipenv run python setup.py sdist bdist_wheel
pipenv run twine upload --repository testpypi dist/*

5. Make a release

1. Set the version
2. Build the package:

    pipenv run python setup.py sdist bdist_wheel

3. Test the release:

    pipenv run twine upload dist/* --verbose -r testpypi

4. Make the release

    pipenv run twine upload dist/* --verbose

3. Tag the commit with the release

    git tag 0.1.0
    git push --tags