Writing Libraries is Terrible
summary
Library development starts pleasant and becomes a maintenance and funding nightmare: a Python library (Hypothesis) that supports three OSes and many interpreter versions forces constant compatibility work, endless issue triage and documentation, and unpaid customer support. The hard numbers are qualitative—dozens of entitlement bug reports, near-zero cash from bounties—so decisions default to whatever feels doable. Experiments with donations, bounties, services, training and proprietary extensions produce limited income; licensing choices matter (MIT/BSD seen as unhelpful) and ongoing commercial funding is the only scalable remedy.
AI-written summary · about
chapters (4)
related talks
01h 17mDevelopment strategies: You've written a library - now what?
29mMoney, Money, Money...
29mThe Unexpected Demise of Open Source Libraries
26mTrapped in the Open Source Supply Chain
42mThe Public Library As An (Almost) Open Source Institution
14mWhat can we do to make Open Source more sustainable?
46mThe Unexpected Demise of Open Source Libraries
01h 01mKeynote: The Business of Open Source