MACEDON : Supporting Programmers with Real-Time Multi-Dimensional Code Evaluation and Optimization
summary
MACEDON provides real-time multi-dimensional evaluation of C/C++ code along five metrics—time, space, clarity, redundancy, documentation—using measurements mapped to scores normalized 1–10 derived from the PIE dataset (top 10% = 10). It offers segment-level optimization with clickable edits and auto-generated accompanying changes, plus markdown shareable reports. A controlled study with 24 novices and two real-world case studies shows improved identification of issues, more effective edits, and better understanding of evolving code state.
AI-written summary · about
chapters (6)
- 0:00Introduction to Macedon
- 1:05Challenges in code optimization
- 3:31Five evaluation dimensions
- 6:00Macedon demo walkthrough
- 7:47User study results
- 9:34Q&A session
related talks
22mCodeMend: Assisting Interactive Programming with Bimodal Embedding
46mPanel: AI in the Loop - Left-Shifting Platform DX and Governance
01mThe AI Tools Kent C. Dodds Uses
18mCoding Evals: From Code Snippets to Codebases
21mMeta: Enabling Programming Languages to Learn from the Crowd
45mThe Definitive Answer to Measuring Developer Productivity
51mDoes Your Code Measure Up?
22mCracking the Code: Deciphering Evaluation Essentials for RAG