We must talk about String
summary
Most code treats string as an all‑accepting type so unexpected values slip through; claim is that using the type system to tag domain strings and enforce validation prevents those bugs. The talk categorizes strings into free form text symbols data and code (Hillel Wayne) and recommends caps on length, regex or canonical parsers for validation, explicit trust boundaries at I/O, and manipulating structured representations (URL objects, parsed JSON) instead of raw concatenation to avoid injection and corruption.
AI-written summary · about
chapters (6)
- 0:00Start
- 0:55static typing
- 1:14what's the purpose of types?
- 5:02what assumptions?
- 5:32characters
- 13:32be careful mixing string types
related talks
30mStrings are Evil: Methods to hide the use of primitive types
37mOvercoming Our Primitive Obsessions
50mCracking the Code to Secure Software
39mOf representation and interpretation: A unified theory
52mSecure by Design, Revisited - Insights & Pitfalls
28mThe Fellowship of the String
01h 34mThe Data Abstraction Talk
01h 02mBasics Designs and How We Got Them Wrong