Working in academia, one of the most important things for us when applying for research funding is to be able to evidence how other people benefit from your work.
Using number of downloads is a poor metric for this. It would be much better if we could show how frequently the code is used (i.e. 10 people using something daily is better than 100 downloading but never using).
Is there a way to do this using bash or python scripts? My idea so far has been around using wget or curl pointed at a url (used only for this) which has google analytics attached to it. This way we could monitor number of executions and location (this info would be sufficient).
I'm not quite sure how it would be implemented. Any suggestions?
NOTE: the code will be open source, so they could see what we're doing, we would include an explanation why it's useful for grants (and therefore their benefit), have a switch for them to disable this feature if needs be and the script would still run if there was no internet connection.