인터페이스는 현지화되어 있으며, Skill 소개와 설명 문서는 선택한 언어로 AI가 자동 번역합니다. 코드·명령어·경로·전문 용어는 원문을 유지하며 원작자 문서가 우선합니다.
Skills 둘러보기
데이터 및 연구 · SKILL.md

oopsla-submission

oopsla-submission

Use when preparing or auditing an OOPSLA submission — picking between the two yearly PACMPL rounds, meeting the 23-page acmsmall anonymous format, writing the required Data-Availability Statement, clearing double-anonymous and concurrent-submission checks, and filing correctly on HotCRP before a firm AoE deadline.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill 문서

GitHub

name: oopsla-submission description: Use when preparing or auditing an OOPSLA submission — picking between the two yearly PACMPL rounds, meeting the 23-page acmsmall anonymous format, writing the required Data-Availability Statement, clearing double-anonymous and concurrent-submission checks, and filing correctly on HotCRP before a firm AoE deadline.

OOPSLA Submission

OOPSLA accepts new papers into either of two yearly rounds feeding the PACMPL OOPSLA volume; for the 2026 volume the rounds closed October 10, 2025 and March 17, 2026, both 11:59:59 PM AoE and explicitly firm, via oopsla26.hotcrp.com. Before running this audit, confirm which round is actually open on the current SPLASH track page — as of 2026-07-08 both 2026 rounds had closed and the 2027 Round 1 call was not yet posted (待核实).

The three questions that precede any checklist

  1. Which round? A round is a strategic choice, not just a date. Filing in the earlier round leaves the later round of the same volume available for a Major-Revision resubmission; filing in the later round pushes any Major Revision into next year's volume. See oopsla-workflow for the calendar math.
  2. New paper or invited revision? A Major-Revision invitation must come back as a revision in the designated round, answering the reviewers' expectation list — never re-filed as a fresh submission. Mixing these up forfeits reviewer continuity.
  3. Is the evidence finished? The two-round model gives reviewers a revision lever; an evaluation you plan to "complete during review" invites exactly that lever to be pulled against you.

Format contract (2026 volume)

RuleValueConsequence of violating
Class optionsacmsmall,screen,review,anonymous on acmartWrong template reads as non-compliant at triage
Initial submission cap23 pages, excluding references, required statements, appendicesOver-cap papers risk rejection without review
Invited-revision cap25 pages, same exclusionsExtra two pages exist to answer reviewers
Data-Availability StatementRequired, placed before the references, not countedMissing statement is an avoidable flag
AnonymityDouble-anonymous, multi-stageIdentity leaks void the review contract

Anonymity sweep beyond the obvious

  • Tool and prototype names that are already public: rename or cite in the third person, and scrub GitHub organization strings from listings.
  • The Data-Availability Statement itself must stay anonymous — point to an anonymized archive, not a lab server whose hostname names your group.
  • PDF metadata (pdfinfo shows author fields LaTeX quietly embeds), embedded figure paths, and acknowledgment stubs left from a prior journal draft.
  • Self-citation grammar: "we extended X [12]" de-anonymizes; "this work builds on X [12]" does not.

Concurrent-submission and prior-work posture

Declare any overlapping manuscript honestly. The dangerous case at OOPSLA is the sibling-venue loop: a rejected PLDI or POPL version resubmitted with the same evaluation weaknesses — reviewer pools overlap, and the two-round model means the same people may see the paper twice within one year. Fix what the last venue's reviews established before re-entering the family.

Filing-day sequence

T-3 days   Freeze claims; regenerate all numbers from the artifact
T-2 days   Anonymity sweep (text, metadata, statement, supplement)
T-1 day    HotCRP record: title/abstract match PDF, topics, conflicts,
           all authors listed with institutions in the form only
T-0        Upload PDF + supplement; re-download and open both;
           confirm timestamp shows AoE margin, not local-time optimism

Deadlines being firm means the plan needs no heroics on the final night; if the paper is not assembled at T-3, target the next round instead — with two rounds a year, the cost of waiting is months, not a year.

Output format

[Round] <target round + deadline, from the live track page>
[Filing type] new submission / invited revision
[Format] pass / fail per contract row above
[Anonymity] clean / leaks found: <list>
[Statement] Data-Availability Statement present and anonymous: yes/no
[Verdict] file now / fix list / wait for next round
저장한 Skills