PyMeetup, Technical debt - how to measure it and best practices

Ljubljana PyMeetup, 22. June 2017

Transcript

1. Technical debt – how to measure it and best practices jernej.makovsek@bitstamp.net
2. Kinds of debts in a SW organization: • Process debt • Organisational debt • Techical debt
3. Technical debt • Similar as financial debt: • You borrowed money from the future • You have to pay it off in the long run • Not always a bad thing: • Creating a MVP • A/B testing • Sometimes necessary to keep a customer • When the code piece will not be touched often
4. Technical debt But mostly you don’t want to have it in your project because of this:
5. Technical debt And as a consequence, this:
6. Technical debt – why bother with measurement • Can be a selling argument – customers get an idea of how risky the software is • Makes investment decisions easy • Determines affordability and maintenance costs
7. Technical debt - measurement • Sonar with technical debt plugin • Todo and Fixme in the code as a comment • Add complexity score e.g. # todo:5: bla
8. Technical debt – strategies and tools • Jenkins Blue Ocean
9. Technical debt – strategies and tools • Pytest
10. Technical debt – strategies and tools • Hypothesis
11. Technical debt – strategies and tools • Upgrading your libraries: • Tox + Docker