Log Mark I

• Displays a graph of logged measures from my (or your) Withings Body Scale

• An anagram of "kilogram"

If you have a Withings wireless scale and you're not impressed with the Healthmate web ui, you might find this a bit more usable.

Features

• Displays weight measures and trend line based on exponential moving average

• Displays fat percentage

• Zoom and pan using standard gestures (drag, scroll wheel zoom, pinch-to-zoom on mobile, etc)

Installation

It's not in any kind of a "finished" state, but it works well enough for me to dogfood. Send feedback on these instructions, especially if you know Python (I don't, much) and they seem unaccountably weird. Because it's probably not intentional.

Requires Python 3.x where I don't know what x is (3.8 works for me).

Register a Withings developer application by visiting https://developer.withings.com/oauth2/#tag/getting-started. Note down the Client Id and Consumer Secret they provide you with, and create a config.json file

{
    "WITHINGS_CLIENT_ID": "58cd5442c76439c754e8172e452d184a7c69a2ef2c5d8e71c3db3815f7f4245e",
    "WITHINGS_CONSUMER_SECRET": "ebbe0a47dd30efdb7aa8e6692b65b0610318523a649b8b28465502d8aa5a"
    "WITHINGS_CALLBACK": "https://log-mark-i.example.com/withings/callback",
}

with virtualenv

If you are installing onto a machine (or container) that has other things on it, you probably want to be using a virtualenv

git clone https://github.com/telent/log-mark-i.git
cd log-mark-i
python -m venv venv
. venv/bin/activate
python -m pip install .
# copy config.json into ./instance
gunicorn 'logmarki:create_app()'

system-wide

If you're doing Docker or something, probably you can just do a system python install

You should be able to run

python -m pip install git+https://github.com/telent/log-mark-i.git
# copy config.json into /usr/var/logmarki-instance/config.json
gunicorn 'logmarki:create_app()'

With Nix (if you already understand Nix)

default.nix builds log-mark-i as a Python module and creates a two-line shell script to start the server. Howver it won't run usefully unless you set the INSTANCE_PATH environment variable to the name of a directory which (1) contains config.json; (2) is writable by the user that runs the script.

I install it using a NixOS module roughly along the lines of nixos-module.example.nix

Regenerating the Nix dependencies

The Nix packaging depends on the generated files requirements.nix and requirements_override.nix, which if you're lucky you can regenerate as follows:

(As of July 2020 this procedure is said to require an unstable version of Nixpkgs, which is why I have the custom NIX_PATH settings)

NIX_PATH=nixpkgs=../nixpkgs nix run nixpkgs.python38 -c bash << EOF
   python -m venv venv
   . venv/bin/activate
   python -m pip install -r requirements.txt
   python -m pip freeze | sed -e '/logmarki/d' -e '/log-mark-i/d' > frozen.txt
   echo 'logmarki=0.1.0' >> frozen.txt
EOF
NIX_PATH=nixpkgs=../nixpkgs nix run nixpkgs.pypi2nix -c pypi2nix -r frozen.txt

HTTPS

You will need to create a proxy pointing at it so that you can access it with HTTPS: I did this with Nginx, but ngrok or something else would work just as well. If you don't/won't/can't use HTTPS you need to edit the call to response.set_cookie to remove secure=True

Development

If you're actively hacking on it,

FLASK_APP='logmarki:create_app()' FLASK_ENV=development python -m flask run --host=0.0.0.0 --port=5007

This is my first time writing Python and my first time using d3.js, so the code style may be weird, nonidiomatic, inconsistent, or just flat-out wrong in places. Constructive criticism gratefully received.