My portfolio is for work that has reached a useful public-facing state. The lab is for retaining the technical record that usually disappears on the way there.
That includes research questions, implementation decisions, partial results, validation plans, failed approaches, diagrams, experiments, and revisions. Some entries will eventually become polished projects or papers. Others may remain useful only because they document why a particular approach did not work.
Documentation principle
The site is intentionally not optimized for the shortest possible explanation. When a method depends on an assumption, that assumption should remain visible. When an experiment is provisional, it should be labeled provisional. When a system can be executed, the executable artifact should live next to the document that explains it.
Document types
The working convention is:
- TR — technical report,
- LN — lab note,
- WP — working paper,
- EXP — executable experiment.
The goal is a research archive that can be read later without needing to reconstruct the missing context from memory.