brycewang-stanford EDBT Supplementary
Use this when assembling EDBT supplementary material. The governing rule is simple and strict: the paper must be judgeable from the reviewed pages alone. The artifact and any appendix support the paper; they do not hold the argument. Reviewers open the artifact at their discretion, so anything the decision depends on lives in the body.
What goes where
| Content | Body (within page budget) | Artifact / appendix |
|---|---|---|
| The problem, contribution, and what it enables | Yes | — |
| Core mechanism / system design | Yes | Full parameter grids, extra configs |
| The headline results and the scope/overhead | Yes | Full result tables, secondary sweeps |
| The evaluation setup a claim depends on | Named in the body | The complete harness and logs |
| Workload / dataset summary + derivation | Summary + criterion | The full derivation scripts and data |
| Raw logs, build scripts, containers | — | Yes |
| Reproduction instructions | A pointer | The README and run scripts |
If a reviewer would need to open the artifact to know whether to accept, the paper is mis-partitioned — move that evidence into the body.
The page-budget discipline (per shape)
EDBT's budget depends on the paper shape (EDBT 2022 CFP; reverify per cycle): Regular and Experiments & Analysis papers ≤12 pages; Vision papers ≤6; references are unlimited and do not count. Consequences:
- References do not eat your budget — cite generously; the constraint is body text, figures, tables, and in-body appendices.
- Keep the headline result tables and the scope/overhead discussion in the body — they are decision-critical.
- Long parameter sweeps, per-configuration breakdowns, and full harness output go to the artifact with explicit forward references.
- Do not use the artifact to smuggle in a result that would not fit — an argument that only closes with material outside the reviewed pages reads as unreviewable.
Reproducibility-forward supplementary rules
[Runnable] the artifact builds and runs from its README on a clean machine; a demo run succeeds
[Provenance] dataset versions, workload derivation, engine build/commit, and seeds are pinned
[Consistency] every body number traces to a script in the package
[Clean archive] no stray credentials, caches, or huge irrelevant files; (if double-blind) no
identity strings, owner paths, or cluster names
[Opens clean] verify the archive unpacks and the README orients a reader in one minute
Appendix architecture (when the template allows in-body appendices)
- Order appendix sections to mirror the evaluation's structure; reviewers navigate by result, not page number.
- Keep each appendix section referenced from the body at least once — orphaned material is invisible under discretionary review.
- Confirm on the current call whether in-body appendices count against the page budget (they usually do) versus supplementary material attached separately; this changes the split.
Vignette: splitting a distributed-systems paper
A paper contributing a query-processing operator and its evaluation: the body keeps the problem, the mechanism, the headline latency/scalability results with variance, the overhead-on-skew-free case, and a scope subsection; the artifact holds the container/build recipe, the workload-derivation scripts with pinned dataset versions, the full parameter sweeps, the raw logs, and the analysis notebooks. Nothing decision-critical lives only in the artifact, because artifact inspection is discretionary and the reviewers judge from the pages.
Output format
[Supplement status] ready / needs fixes / not ready
[Shape + budget] Regular | Experiments-&-Analysis | Vision — body within limit?
[Partition check] anything decision-critical outside the body? <none / move: what>
[Reproducibility] artifact runnable + provenance pinned + consistent with paper? yes/no
[Clean archive] (if double-blind) identity-free; no stray files? passed/issues
[Body dependency] <what a reviewer can decide without opening the artifact>