coherent level of dictionary

[mizunashi-mana]

coherence ~ global uniqueness

• no coherence (but non-ambiguity)
• unchecked coherence
  • compiler does not check
  • compiler rely coherence to optimize
• checked coherence

References

1. White, L., Bour, F., & Yallop, J. (2015). Modular implicits. Electronic Proceedings in Theoretical Computer Science, 198, 22–63. https://doi.org/10.4204/EPTCS.198.2
   • The OCaml case
2. SCHRIJVERS, T., OLIVEIRA, B. C. D. S., WADLER, P., & MARNTIROSIAN, K. (2019). COCHIS: Stable and coherent implicits. Journal of Functional Programming, 29, e3. https://doi.org/10.1017/S0956796818000242
   • The calculus with implicit system focusing coherence
3. Edward Z. Yang. Type classes: confluence, coherence and global uniqueness. http://blog.ezyang.com/2014/07/type-classes-confluence-coherence-global-uniqueness/
   • The summary of Haskell's type classes
4. Winant, T., & Devriese, D. (2018). Coherent explicit dictionary application for Haskell. ACM SIGPLAN Notices, 53(7), 81–93. https://doi.org/10.1145/3299711.3242752
   • An extensions of Haskell for explicit dictionary application with coherence
5. Dotty's Issue. What kind of typeclass coherence should we support?. https://github.com/lampepfl/dotty/issues/4234
   • A discussion for Dotty (Scala 3)
6. Rust's dicussion. Orphan rules. https://internals.rust-lang.org/t/orphan-rules/1322
   • A diccusion of orphan instances for Rust

[mizunashi-mana]

• Dotty's discussion. Allow Typeclasses to Declare Themselves Coherent. https://github.com/lampepfl/dotty-feature-requests/issues/4
  • A proposal of coherent typeclasses for Scala
  • The motivation by Odersky is summarized of coherence problem simply.