Skip to content

frans-fuerst/track

Repository files navigation

Track - automatic time tracker for computer work

Track logs the time you're actively working on your desktop computer as well as the applications you're using in order to create a time chart of your working day.

Track does not connect anywhere and shares your information with nobody!

Actually this is why I started this project in the first place..

However at the current development stage Track stores all data unencrypted, so please ensure nobody has access to your Track directory (usually ~/.track).

Unlike KTimeTracker or Hamster Track aims at zero user interaction. Once configured you just (auto)start it and Track runs on it's own.

This is an early screenshot to give an idea of how Track works: recent screenshot

What it does:

  • logs times when your computer is active and which applications are in focus
  • handles a list of regex-rules which assign certain activities to private work

What it does not:

  • manage an abstract task list you would have to maintain
  • send any information to someone

When you should use Track:

  • in case you're working mainly on a computer like software developers
  • if you want to know how long you've been at work today
  • if you want to know how much of your private time you spent on the computer
  • if you want to waste your time with another self profiling tool

The current project state is (still! damn!) very early (see the schedule).

Very basic features are still missing so it might be wise to come back in a week or so. However it's totally save to use the tool and it's providing some interesting information already.

Categories and Rules

In order to get an idea how much of the day you spend for work and how much for private stuff (or how much you spend on project A or project B) track allows you to define special rules which assign each running program to a category.

Right now categories are just numbers (2 for work, 3 for private stuff, 4 for break, 0 for idle and 1 for unassigned programs). In the future I plan to allow arbitrary categories (or category trees), e.g.:

  • work
    • project A
    • project B
    • browsing on github.com
  • private stuff
    • project C
  • procrastinating
    • browsing on reddit.com

The technical approach is very simple: you define standard Python regex rules which are matched against the title of the active window.

For example - in a very simple world - it might be enough to define that browsing the internet using "Firefox" is private stuff (category 3) and everything else is work (category 2). In this case you would just define one rule:

regex = `r".*-Mozilla Firefox.*"` -> category = `3`

This way every program whose title does not match .*-Mozilla Firefox.* would be assigned to the default category 0 and all Firefox browser windows would result in category 1 (which you might be private in your eyes).

Requirements

Basically you Python3 with PyQt5, psutil and zmq installed. Some PyQt5 versions do not behave well and you will need build-essentials etc to make it work. For me installing the following packages worked:

  • Linux with X11 (Wayland had some problems when I last checked it)
  • or Windows - it once worked but currently I have no way to check it
  • Python 3+
  • python3-devel or equivalent via apt, dnf, etc.
  • PyQt5 via pip (v5.14 worked for me)
  • zmq via pip
  • psutil via pip

Try this pip command: pip3 install --user --upgrade psutil zmq PyQt5==5.14

How to run

Clone the repo:

git clone https://github.com/frans-fuerst/track

Run track to start Track client and server. The server will keep running in background if UI gets closed.

To list starting / endings times of recorded days run track-cli list

Know limitations / Shortcomings

  • Currently Track seems to not work well with Wayland, which might be an issue of both Track and Wayland. When your're on Linux consider using X11.
  • I started this project 5 years ago when my Python skills were still very weak. Please do either not look at the code or help me improve it.
  • Track records data but still does not evaluate it. track-cli does some steps in that direction.
  • No check for plausibility: if - for example - your computer wakes up at 3:12 for just a second this incident will be recorded and currently there is no way to remove this stray event and your day will officially start at 3:12
  • Setup / autostart worked once upon a time but doesn't now. But as a Linux Pro you know what to do.
  • Daily note does'nt get cleared on midnight (but you can simply overwrite it)
  • Break (cat 4) cannot be selected in timechart
  • Bug: Gnome Icon not working
  • Categories limited to 0-4
  • Categories as ints

Tests

Here is how I currently "test" track. It's actually more try and look for crashes :)

  • Delete ~/.track/, Try to start track
  • Open all spoilers
  • Delete all rules by pressing delete
  • add rules
  • edit note
  • restart
  • rules still exist?
  • note still exists?

About

personal computer usage tracker

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •  

Languages