技能创建器
skill-creator
创建、改进和评估自定义技能,把专业流程转化为可重复使用的 AI 能力。
oopsla-camera-ready【数据与研究】Use when converting an accepted OOPSLA paper into its PACMPL journal article — final acmsmall formatting within the 25-page revision cap, de-anonymization and Data-Availability Statement updates, ACM open-access and rights steps, hitting the OOPSLA1 (April) or OOPSLA2 (October) issue, and planning the SPLASH talk.
独立 GitHub 快照 · 自动分类 · 以原作者文档为准
An accepted OOPSLA paper does not go into "conference proceedings" — it becomes a journal article in PACMPL (Proceedings of the ACM on Programming Languages), Gold Open Access, in the issue matching its acceptance round: Round 1 acceptances of the 2026 volume appeared as Vol. 10, Issue OOPSLA1 (published no earlier than April 1, 2026); Round 2 acceptances form Issue OOPSLA2 (no earlier than October 1, 2026). The talk then happens at SPLASH — for 2026, Oakland, with OOPSLA sessions October 5–7. Production deadlines arrive by email from ACM/the chairs and were not publicly posted at verification time (待核实): treat the acceptance letter as the calendar.
| Item | Submission state | Published-article state |
|---|---|---|
| Class options | acmsmall,screen,review,anonymous | acmsmall,screen per production instructions |
| Authors | Hidden | Full names, affiliations, ORCIDs, CCS concepts |
| Acknowledgments | Absent | Funders, grant numbers, contributors |
| Data-Availability Statement | Anonymized archive | Real archive — ideally the DOI-stamped Zenodo deposit from artifact evaluation |
| Citations to your own PACMPL-era work | Third person | May remain third person; only add identity where required |
| Page budget | 23 pages | Up to 25 pages (revision cap), same exclusions |
For papers that passed through Minor Revision, the final version is checked against the concern list, so build the reconciliation table before touching LaTeX:
| # | Reviewer requirement | Where satisfied (section/page) | Evidence |
|---|----------------------|--------------------------------|----------|
| 1 | Add baseline B to Table 3 | §5.2, p.14 | artifact run 2026-xx-xx |
| 2 | Clarify Thm 2 side condition | §3.4, p.8 | rewritten statement |
...every row filled before upload; unfillable rows escalated to the chairs.
Silently dropping a commitment is the single fastest way to sour the same committee that will see your group's next submission.
dl.acm.org/journal/pacmpl/open-access before advising anyone on
cost, and route institutional questions to ACM Open.PACMPL 10(OOPSLA1)
— and that is how others must cite it (oopsla-related-work). Get title
capitalization and author-name forms right in the ACM form, not just the
PDF, because the DL record is generated from the form.oopsla-artifact-evaluation) so the deposit DOI exists before the
camera-ready statement references it.Accepted papers present in person at SPLASH; for 2026 that meant registering for Oakland (October 4–9) and preparing for a PL audience that has already had access to the published April article for months in the OOPSLA1 case — the talk should assume the paper is read, and spend its minutes on the mechanism and the demo, not the related-work slide. An accepted-paper poster is additionally encouraged at SPLASH and is cheap conference-hallway leverage.
[Issue target] OOPSLA1 / OOPSLA2, volume <n>
[Reconciliation] <rows total / rows satisfied / escalations>
[De-anonymization] complete: names, ORCID, acks, statement, links
[Rights] form status + license line + APC/waiver status (待核实 items flagged)
[Talk] registration, slot, poster decision