brycewang-stanford ICSE Camera-Ready
Acceptance converts the paper from an argument into a permanent record — ICSE proceedings live in the ACM Digital Library and IEEE Xplore, and the camera-ready is the version every future reader, replicator, and artifact evaluator holds against your claims. The window is short and lands amid artifact-evaluation prep, so run it as a checklist, not a rewrite.
What actually changes
The 2027 call (read 2026-07-08) grants accepted papers one extra page of main text for the camera-ready — 11 + 2 instead of 10 + 2. Spend it where review friction was, in this order of value:
- Changes promised in the author response or Major Revision ledger that were squeezed by space — reviewers who see their request honored in the final version include your artifact evaluators and future citers.
- The threats-to-validity or study-design detail that reviews called thin.
- Restored per-subject detail that summarize-and-point compression removed.
Do not spend it on a new contribution: the paper was accepted for what it evidenced, and content that never faced review is a correctness risk with your name permanently on it.
De-anonymization order
Reverse the double-anonymous transformations systematically — grep, don't recall:
# Residual-anonymity sweep on the camera-ready sources
grep -rn "anonym\|blinded\|double-blind\|under review" *.tex # stale phrases
grep -rn "anonymous.4open\|anonymized-link\|REDACTED" *.tex # placeholder URLs
grep -n "our prior work\|the authors of \[" *.tex # third-person self-cites to naturalize
pdfinfo camera-ready.pdf | grep -Ei 'author|title' # metadata now SHOULD carry authors
Restore: the author block with affiliations, emails, and ORCIDs;
acknowledgments and funding with grant numbers; self-citations rephrased into
natural first person where it improves clarity; and — most consequentially —
the availability statement's links, from anonymized mirrors to the
permanent DOI-issuing archive (icse-artifact-evaluation covers minting it).
A camera-ready that still points at a review-time anonymous mirror will
outlive the mirror.
Production-quality gate
| Check | Detail |
|---|---|
| Template | Current IEEE conference class, unmodified; production vendors reject tampering the PC tolerated |
| Metadata block | Copyright/DOI stamps per the acceptance instructions; correct year and conference string |
| Author records | Name spellings, affiliations, ORCIDs consistent with the submission system — indexing errors are near-impossible to fix post-publication |
| Fonts and figures | Fonts embedded, figures vector where possible, colorblind-legible; print-grayscale still distinguishable |
| References | Every entry re-verified against dblp/DL records; camera-ready is the last chance to fix a wrong venue or year |
| Availability statement | DOI links live; artifact contents match final table/figure numbering |
| Title/abstract sync | Final title and abstract match the submission system's metadata fields exactly |
Consistency with the accepted version
Chairs expect the camera-ready to be the accepted paper plus the required changes — not a quiet second revision. Keep a change log from accepted PDF to camera-ready; if a real error surfaced after acceptance (a bug in an analysis script, a corrected number), fix it and flag it to the track chairs rather than hoping nobody diffs. Silent number changes discovered later are an integrity problem; disclosed corrections are routine science.
Obligations around the paper — verify, don't assume
Several items were not published for the 2027 cycle at check time and are 待核实 on the live pages once acceptance emails go out:
- Camera-ready deadline and the exact submission system/vendor instructions.
- Registration requirements tied to publication (one author registering at a qualifying rate is the common conference pattern — confirm ICSE 2027's wording before budgeting).
- Presentation format and whether remote presentation exists; plan for in-person in Dublin (Apr 25 – May 1, 2027; core days Apr 28–30).
- Page-purchase options beyond the granted extra page.
- Video, poster, or preprint requirements attached to the final submission.
Book Dublin logistics early regardless: late-April conference housing in a capital city is its own deadline.
The December–January squeeze
For Major-Revision papers the final decision (December 18 in the 2027 cycle) lands directly against the artifact-evaluation window (early January in recent cycles) and, soon after, camera-ready production. Sequence deliberately: freeze the artifact container first (it feeds the badge deadline), then finalize the availability statement's DOI (the camera-ready needs it), then spend the extra page — because page-11 content that cites package paths must cite the final package. Groups that treat these as three independent deadlines do the DOI work twice.
Publication-week hygiene
Post the accepted-version preprint per the copyright terms you signed (arXiv or institutional repository — the open-science culture expects it), update the artifact's citation file with the final DOI, and make the paper's page, package, and preprint agree on one canonical title and author order. Future citation graphs are built from whichever version spreads first.
Output format
[Extra page] spent on: <review-friction items>; new-contribution check: none added
[De-anonymization] sweep results; availability links now DOI-permanent
[Production gate] table above -> pass/fail per row
[Change log] accepted -> camera-ready diffs; disclosed items
[Obligations] deadline / registration / presentation — verified or 待核实 with owner