atc-topic-selection
GitHub辅助判断计算机系统项目是否适合投稿至ATC。基于其广泛实用、重视工程实现与诚实评估的特点，对比OSDI/SOSP等顶会及NSDI/FAST等专业会议，结合广度测试与已部署系统赛道进行路由决策。
Trigger Scenarios
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill atc-topic-selection -g -y
SKILL.md
Frontmatter
{
"name": "atc-topic-selection",
"description": "Use when deciding whether a computer-systems project belongs at ATC (ACM SIGOPS Annual Technical Conference, formerly USENIX ATC) or should be routed to OSDI, SOSP, NSDI, EuroSys, FAST, or a workshop like HotOS\/HotStorage — distinguishing ATC by its broad practical scope, its Deployed Systems \/ experience lane, the breadth-vs-selectivity trade, and the November\/June calendar."
}
ATC Topic Selection
Decide the venue before drafting. ATC — the ACM SIGOPS Annual Technical Conference, for about fifty years the USENIX Annual Technical Conference — is the systems community's broad, practical venue. It emphasizes implementations and experimental results across the whole systems stack and takes a wider slice than the more selective OSDI/SOSP flagships. The decisive question is rarely "is this systems?" but "is this a solid, useful, measured systems result — and is ATC's breadth the right home versus a more specialized or more selective venue?"
What ATC rewards
- Built and measured systems across a broad scope: operating systems, runtime systems, parallel and distributed systems, storage, networking, security and privacy, virtualization, hardware-software interaction, performance and workload characterization, reliability, availability, scalability, energy, and bug-finding / tracing / troubleshooting — "of all scales, from embedded devices to data centers and clouds."
- Useful over merely novel. A solid, well-engineered, carefully measured system that a flagship might call incremental is squarely in scope if it is real and the evaluation is honest.
- Deployed-systems / experience papers. ATC has a lane for papers about real-world deployed systems that need not present new ideas — they are judged on practical insight. (Continuity of the named track under SIGOPS is 待核实, but the appetite for experience papers is long-standing.)
Sibling-venue routing table
| Signal in your project | Better home | Why |
|---|---|---|
| Solid, useful, measured systems work of broad interest; usefulness over novelty | ATC | The broad practical-systems venue; welcomes strong engineering + honest measurement |
| A large, novel systems contribution aimed at the flagship bar | OSDI / SOSP | More selective, novelty-forward; different calendars and no extended-abstract gate |
| The contribution is a networked-systems design (dataplane, protocols, RDMA, congestion) | NSDI | Networked-systems center; its own multi-deadline model |
| Broad systems, but EuroSys's cycle lands sooner or fits the community | EuroSys | The other broad SIGOPS-family systems conference; choose by calendar/pull |
| The core is file and storage systems depth | FAST | Storage's specialized home |
| A provocative idea or early design with argument but no built system | HotOS / HotStorage / workshop | Position/early-work venues; ATC wants a built, measured system |
| A real deployed system with practical lessons but no new idea | ATC (Deployed Systems lane) | ATC explicitly values experience papers |
The breadth and deployed-systems tests
Two quick tests sharpen a borderline verdict:
- Breadth-vs-selectivity test: if the paper's strength is careful engineering and honest measurement of something genuinely useful — rather than a headline-novel mechanism — ATC's broad, practical bar fits better than OSDI/SOSP's selectivity. Do not withhold a solid, useful system from ATC because it is "not flagship-novel."
- Deployed-systems test: if the contribution is the experience of running a real system at scale — what broke, what you learned, what others should copy — ATC's experience lane is a real home, where "no new idea" is not a rejection reason.
Evidence maturity, without the ladder cliché
Fit is necessary but not sufficient: the same idea sits at different doors depending on how far the system has come. An idea with an argument but no implementation is a HotOS-style position paper; a prototype measured only on inputs you chose needs a real testbed and workloads before the research track; a networked-systems design whose contribution is the network belongs at NSDI. Submitting one step early earns a "promising, but build and measure it" and costs a cycle.
Cheap reconnaissance before committing
[Scope] scan the last two ATC programs (dblp: USENIX ATC; the SIGOPS ATC page) for your subarea
-> several recent papers = a reviewer pool exists; none = opening or mismatch
[Citations] is your bibliography majority systems venues (ATC/OSDI/SOSP/NSDI/EuroSys/FAST)?
-> majority elsewhere => reviewers read you as a visitor; naturalize the intro first
[Calendar] compare ATC's June deadline / November conference with OSDI/NSDI/EuroSys/FAST dates
-> route to the nearest honest fit; ATC 2026 is Hong Kong, Nov 15-18
[Bar] is the paper "useful and solid" (ATC) or "novel and large" (OSDI/SOSP)? be honest
Decision procedure
[Audience] who acts differently if the system works? -> systems builders/operators/researchers?
[Claim type] built-system / technique+implementation / architecture / experience / measurement study
[ATC vs flagship] solid+useful -> ATC; novel+large -> OSDI/SOSP
[Specialist check] network core -> NSDI; storage core -> FAST; idea-only -> HotOS/workshop
[Verdict] ATC research track / Deployed Systems lane / sibling venue / workshop, with a one-line reason
Run this before the writing skills; a wrong venue decision wastes every later step. When the verdict
is ATC, continue with atc-workflow for the June-anchored calendar and atc-writing-style for the
self-standing extended abstract and the systems-paper shape.
Version History
- 9f86f09 Current 2026-07-19 14:20
