Using APIs appropriately
summary
Advice: treat web services as unreliable and never trust an API; monitor dependencies with uptime checks (Pingdom), cache responses, and design for failure so users still get content when an endpoint dies. Handle rate limits (HTTP 429) with exponential backoff, expect malformed bodies and mixed encodings, and avoid hammering after a 200. Use command‑line tooling like jq and backends such as Redis/Elasticsearch for large transforms; require JSON, pagination (avoid 10,000‑record requests that cause 504s) and provide runnable curl examples.
AI-written summary · about
chapters (12)
- 0:00Using APIs effectively
- 1:46Personal API journey
- 4:16Never trust APIs blindly
- 6:24API failures and caching
- 7:06Inconsistent API responses
- 9:13Consistent encoding and formats
- 10:16Respecting API rate limits
- 11:43Successful API responses
- 12:28JQ for JSON processing
- 14:12Large datasets and licensing
- 15:38API provider advice
- 17:02Final thoughts and contact
related talks
47mAPI Tips from the Frontline
49mWeb APIs that Developers Love
52mDesigning and Implementing REST APIs For the Long Haul
40mGive It a REST – Tips for Designing and Consuming Public API’s
49mTop REST API Design Pitfalls
20mPotential of API integrations, common traps and advices
45mAPI on Steroids
49mAdvanced API design: how an awesome API can attract friends...