Musings on the representation, propagation and handling of errors in existing programming languages, as well as thoughts on future systems. (15K+ words)

Target audience: Practitioners interested in programming language design and familiar with representations of errors in at least a few different languages such as error codes, checked/unchecked exceptions, tagged unions, polymorphic variants etc.

Estimated reading time: 60 to 90 mins.

  • KissakiEnglish
    1·
    17 hours ago

    Quoting the conclusion:

    The fundamental nature of errors is that they are often multi-faceted, and complex to reason about by virtue of being multi-faceted. The wide variety of socio-technical contexts that software is created, delivered and used in add further complexity.

    The linked The Error Model article is very exhaustive and good.

    They end with more concrete experience/findings/conclusions - given the context of a more concrete project and scope and goals.