PLDI Submission
Run this audit before the upload, not after. The numbers below were read from the
PLDI 2026 Call for Papers and track pages on 2026-07-08 (via search renderings of
the official URLs — see resources/official-source-map.md); they describe one
cycle and must be rechecked against the live CFP for the cycle you are entering.
The PLDI 2027 call was not yet posted at that date.
The 2026-cycle mechanics
- One deadline a year. PLDI 2026 papers were due Thursday, November 13, 2025.
There is no second round; a missed November costs twelve months.
- HotCRP, not OpenReview. Submission ran at
pldi2026.hotcrp.com — profile,
conflicts, and topic areas live there, and each edition gets a fresh instance.
- Page budget: at most 20 pages of text, excluding bibliography, in the
single-column ACM
acmsmall layout (10 pt font, 12 pt line spacing, text block
5.478 in x 7.884 in), the format PACMPL publishes in. Build with
sample-acmsmall-conf.tex from the acmart package.
- Summary rejection is the stated penalty for a different format — smaller
fonts, a larger text block, squeezed spacing. The chairs do not negotiate
margins.
- Double-blind, and PLDI has used it for years: reviewers must not learn who
you are from the PDF or anything it links to.
Where PL papers leak identity
Compiler and tool papers leak differently from ML papers. The usual culprits: a
tool name already public on GitHub under your lab, benchmark harness URLs,
\author metadata inside the PDF, grant numbers in acknowledgements, and commit
histories inside supplementary archives.
pdfinfo paper.pdf | grep -Ei 'author|creator'
pdftotext paper.pdf - | grep -niE 'github\.com/[^ ]+|gitlab|acknowledg|grant no' | head
unzip -l supplement.zip | grep -Ei '\.git/|/home/|/Users/' | head
Cite your own prior work in the third person, and if the tool is already public,
discuss with your co-authors whether to rename it in the submission or accept the
risk explicitly — do not discover the collision during reviewing.
Summary-rejection and triage table
| Trigger |
Consequence at PLDI |
Recoverable? |
| Modified format (font, text block, spacing) |
Summary rejection, stated in the CFP |
No, after the deadline |
| Over the text-page cap |
Summary rejection risk |
No |
| Author identity visible in PDF or links |
Desk-level double-blind violation |
No |
| Same work under review elsewhere |
Dual-submission violation under SIGPLAN/ACM policy |
No |
| Evaluation missing baselines or variance |
Survives triage, dies in review |
Only before submission |
| Artifact unplanned |
No submission penalty, but AE starts right after acceptance |
Yes, but plan now |
Final week, in order
- Freeze the claim list; every abstract sentence must have a section that proves it.
- Rebuild the PDF from a clean tree with the unmodified
acmsmall class and
recount text pages against the current cap.
- Run the leak scan above on the PDF and every supplementary archive.
- Verify HotCRP conflicts and topics; a stale conflict list misroutes reviewers.
- Check the CFP's dual-submission and republication wording against anything you
have on arXiv or under review elsewhere.
- Submit a day early. HotCRP deadlines are hard cutoffs.
What this cycle did not settle (待核实)
Whether an abstract-registration step precedes the paper deadline, exact
supplementary-material rules, and any AI-assistance disclosure requirement were
not confirmed in accessible renderings of the 2026 pages — read the live CFP
rather than assuming either way.
Output format
[PLDI readiness] Ready / Needs fixes / Not ready
[Format] acmsmall unmodified? text pages vs cap?
[Anonymity] PDF metadata / links / tool name / supplement
[Policy] dual submission / arXiv posture / AI disclosure (待核实)
[Fix order] <ordered list before the November deadline>
1---2name: pldi-submission3description: Use when auditing a PLDI submission for HotCRP readiness — the single annual November deadline, the 20-pages-of-text-excluding-bibliography cap in single-column acmsmall format, double-blind hygiene across tool names and repositories, dual-submission rules, and summary-rejection triggers before the cutoff.4---56# PLDI Submission78Run this audit before the upload, not after. The numbers below were read from the9PLDI 2026 Call for Papers and track pages on 2026-07-08 (via search renderings of10the official URLs — see `resources/official-source-map.md`); they describe one11cycle and must be rechecked against the live CFP for the cycle you are entering.12The PLDI 2027 call was not yet posted at that date.1314## The 2026-cycle mechanics1516- **One deadline a year.** PLDI 2026 papers were due Thursday, November 13, 2025.17 There is no second round; a missed November costs twelve months.18- **HotCRP, not OpenReview.** Submission ran at `pldi2026.hotcrp.com` — profile,19 conflicts, and topic areas live there, and each edition gets a fresh instance.20- **Page budget: at most 20 pages of *text*, excluding bibliography**, in the21 single-column ACM `acmsmall` layout (10 pt font, 12 pt line spacing, text block22 5.478 in x 7.884 in), the format PACMPL publishes in. Build with23 `sample-acmsmall-conf.tex` from the acmart package.24- **Summary rejection is the stated penalty** for a different format — smaller25 fonts, a larger text block, squeezed spacing. The chairs do not negotiate26 margins.27- **Double-blind**, and PLDI has used it for years: reviewers must not learn who28 you are from the PDF or anything it links to.2930## Where PL papers leak identity3132Compiler and tool papers leak differently from ML papers. The usual culprits: a33tool name already public on GitHub under your lab, benchmark harness URLs,34`\author` metadata inside the PDF, grant numbers in acknowledgements, and commit35histories inside supplementary archives.3637```bash38pdfinfo paper.pdf | grep -Ei 'author|creator'39pdftotext paper.pdf - | grep -niE 'github\.com/[^ ]+|gitlab|acknowledg|grant no' | head40unzip -l supplement.zip | grep -Ei '\.git/|/home/|/Users/' | head41```4243Cite your own prior work in the third person, and if the tool is already public,44discuss with your co-authors whether to rename it in the submission or accept the45risk explicitly — do not discover the collision during reviewing.4647## Summary-rejection and triage table4849| Trigger | Consequence at PLDI | Recoverable? |50|---|---|---|51| Modified format (font, text block, spacing) | Summary rejection, stated in the CFP | No, after the deadline |52| Over the text-page cap | Summary rejection risk | No |53| Author identity visible in PDF or links | Desk-level double-blind violation | No |54| Same work under review elsewhere | Dual-submission violation under SIGPLAN/ACM policy | No |55| Evaluation missing baselines or variance | Survives triage, dies in review | Only before submission |56| Artifact unplanned | No submission penalty, but AE starts right after acceptance | Yes, but plan now |5758## Final week, in order59601. Freeze the claim list; every abstract sentence must have a section that proves it.612. Rebuild the PDF from a clean tree with the unmodified `acmsmall` class and62 recount text pages against the current cap.633. Run the leak scan above on the PDF and every supplementary archive.644. Verify HotCRP conflicts and topics; a stale conflict list misroutes reviewers.655. Check the CFP's dual-submission and republication wording against anything you66 have on arXiv or under review elsewhere.676. Submit a day early. HotCRP deadlines are hard cutoffs.6869## What this cycle did not settle (待核实)7071Whether an abstract-registration step precedes the paper deadline, exact72supplementary-material rules, and any AI-assistance disclosure requirement were73not confirmed in accessible renderings of the 2026 pages — read the live CFP74rather than assuming either way.7576## Output format7778```text79[PLDI readiness] Ready / Needs fixes / Not ready80[Format] acmsmall unmodified? text pages vs cap?81[Anonymity] PDF metadata / links / tool name / supplement82[Policy] dual submission / arXiv posture / AI disclosure (待核实)83[Fix order] <ordered list before the November deadline>84```
brycewang-stanford