brycewang-stanford OOPSLA Topic Selection
OOPSLA is the broadest seat at the SIGPLAN table: its published scope covers
programming languages "from design to implementation and from mathematical
formalisms to empirical studies," and its verified exemplars include a
benchmark suite, a measurement-methodology critique, a corpus/survey study,
a soundness counterexample, a language experience report, and a formalized
runtime mechanism (resources/exemplars/library.md). The routing question
is therefore rarely "is this PL?" but "is OOPSLA's breadth the asset, or
would a narrower sibling reward this paper's center of mass more?"
The center-of-mass test
State the one sentence a reviewer would repeat to the committee. Route on what that sentence is about:
| The sentence is about... | First-choice venue | OOPSLA case exists when... |
|---|---|---|
| A metatheoretic result, calculus, logic | POPL | The formalism explains a deployed mechanism or design choice |
| Compiler/runtime performance engineering | PLDI | The insight is the design trade-off, not the speedup |
| Typed/functional language technique | ICFP | The audience is OO/imperative/multi-paradigm practice |
| A provocative vision without full evidence | Onward! (same SPLASH building) | Evidence has matured past position-paper strength |
| OO language design, European community | ECOOP (AITO, not PACMPL) | PACMPL journal publication or the SPLASH audience matters |
| Testing/analysis practice at SE scale | ISSTA / ICSE / FSE | The finding changes how languages are designed, not just tools |
| Human factors of programming | CHI or OOPSLA's human-aspects lane | The methodology meets study-design scrutiny (oopsla-experiments) |
Signals the project is genuinely OOPSLA-shaped
- The contribution couples layers — a semantics plus its implementation consequences, a design plus the empirical study of its use — rather than living purely in one.
- The paper would still matter if a competing implementation shipped: the claim is about the design space, not the artifact's existence.
- Experience and applications carry analysis: OOPSLA has a real lane for "we built and deployed it; here is what generalizes" (the Julia paper is the model), which PLDI/POPL reward less.
- Empirical questions about programs and programmers at scale — adoption, API misuse, bug taxonomies — with instrument-grade methodology.
Signals to route away
- The evaluation is the contribution and it's a speedup: PLDI's benchmark-first culture is the better courtroom.
- The theorem stands alone without an implementation story: POPL.
- The idea needs permission to be speculative: Onward!, and return to OOPSLA a volume later with evidence — a deliberate SPLASH-internal pipeline.
- The novelty is a product description: no SIGPLAN venue; consider an industry track or experience venue outside the family.
Two-round consequence for routing
Because OOPSLA offers two doors a year with a revision system, it is often
the fastest family member from "solid evidence" to "published journal
article" — a legitimate tiebreaker when a paper fits two venues. It is a bad
reason to submit work whose center of mass is elsewhere: family reviewer
pools overlap heavily, and a misrouted paper burns goodwill you will meet
again (oopsla-review-process).
Decision procedure
1. Write the one-sentence claim; identify its center of mass (table above)
2. Check exemplar precedent: which library entry shape matches? none → doubt
3. Ask the coupling question: does the paper bridge ≥2 layers? no → sibling
4. Verify current scope text on the live track page (topics drift by volume)
5. Commit to a round (oopsla-workflow) or a re-route with a named venue
Output format
[Center of mass] <one sentence + layer(s) it lives in>
[Routing] OOPSLA / POPL / PLDI / ICFP / Onward! / ECOOP / SE venue / other
[Precedent] exemplar shape matched, or none
[Fit risk] the objection the committee would raise
[Commitment] target round or re-route action
Source: brycewang-stanford/Awesome-Journal-Skills → OOPSLA-Skills/skills/oopsla-topic-selection/SKILL.md