POPL's seat in the SIGPLAN year is fixed and unforgiving: submit in July, hear in early October, revise and package artifacts in the fall, present in January. For POPL 2027 (verified 2026-07-08): deadline July 9, 2026 AoE; notification October 5, 2026; symposium January 10-16, 2027, Mexico City. There is no abstract deadline and no second call. POPL is a conference — no editor-in-chief, no submission fee; the chairs rotate per edition (2027 names 待核实) and publication costs run through PACMPL's open-access model, not author-facing page charges in the normal case.
| Weeks before deadline | Theory-paper milestone |
|---|---|
| 16+ | Core calculus stable; main theorem conjectured precisely |
| 12 | Hard lemmas proved on paper; mechanization of the kernel started |
| 8 | Main theorem proved; decide mechanize-vs-hand-prove split per theorem |
| 6 | Case studies or prototype evidence frozen (popl-experiments) |
| 4 | Full draft in acmsmall; informal-to-formal ramp reviewed by an outsider |
| 2 | Mock review by someone who tries to break the theorem, not the prose |
| 1 | Anonymity sweep of paper + proof archive; correspondence table regenerated |
| 0 | HotCRP upload with hours of AoE margin, not minutes |
The distinctive scheduling truth: theorems do not compress. An experiment can be rerun overnight; a broken lemma two weeks out cannot be fixed by working weekends. Freeze the formal core earlier than feels necessary.
October notification opens three parallel tracks that all land before January:
popl-camera-ready).popl-artifact-evaluation).All four SIGPLAN flagships publish in the same journal, so retargeting costs reviewer culture, not archival quality. From POPL's July seat, the recent pattern (verify each venue's current call — dates rotate):
POPL (July deadline, January conference) <- you are here
-> OOPSLA: next PACMPL deadline after July in recent cycles
-> PLDI: November-ish deadline, June conference
-> ICFP: late-winter/spring deadline, fall conference
Retarget rule: move to the venue whose reviewers match your EVIDENCE,
not the nearest date — see popl-topic-selection.
Never park the same paper at two of them at once; dual submission burns the pair.
[Current stage] framing / proving / mechanizing / drafting / submitted / response / revision / production
[Next hard date] <event + date + source URL, or 待核实>
[Critical path] <the three tasks that gate the next date>
[Frozen vs fluid] <what may still change without cascading into proofs>
[Fallback] <retarget venue + its verified deadline, if July is unrealistic>