Review algorithms and data structures looking for optimization opportunities. The idea is to prioritize good design, code readability etc but without wasting time/memory unnecessarily.
Measure performance (time/memory) against standard MeTTa interpreter spaces
Eventually, other downstream called classes may be refactored as well. Two known issue are
The test-cases to benchmark can be extracted from MeTTa examples provided in the MeTTa interpreter repo: https://github.com/trueagi-io/hyperon-experimental but it'd be interesting to collect more test MeTTa code from users. There's a channel in Mattermost for MeTTa user group ("Metta Coders" in public World server)
EPIC
Goals:
Eventually, other downstream called classes may be refactored as well. Two known issue are
but there may be others.
The test-cases to benchmark can be extracted from MeTTa examples provided in the MeTTa interpreter repo: https://github.com/trueagi-io/hyperon-experimental but it'd be interesting to collect more test MeTTa code from users. There's a channel in Mattermost for MeTTa user group ("Metta Coders" in public World server)