I published a new performance release of the Egel interpreter named "v.0.1.10: Egel, let's go." I am now at a fork in my thoughts.
Egel successfully shows that an operational semantics where the root of a directed acyclic graph is repeatedly rewritten can serve as the back-end of a small functional scripting language. However, Egel is notably slow, approximately eight times slower than Python according to various benchmarks.
Egel is also rather rudimentary as a scripting language, sufficient for solving Advent of Code puzzles but lacking essential features, presenting an overall unpolished experience. What I want to do is minimally add date/time objects as primitives and incorporate support for applicatives and lenses.
Despite these potential improvements, it's challenging to justify investing significant effort into a tool that will not find widespread use.
No comments:
Post a Comment