The goal of this HotOS’21 panel to explore and highlight future avenues for research in the context of the shell, informed by both industry and academia—across various academic communities including systems, programming languages, software engineering, security, and human-computer interaction. We envision this panel as a round-table discussion between panelists and participants, which include many authors of shell-related technologies — e.g., popular and niche shells, static analysis for the shell, secure shell scripting languages, automated shell script parallelization, shell semantics, etc. We anticipate minimal distinction between panelists and participants in the discussion: panelists will provide brief statements (~2min) about the future and/or challenges of the shell to get the discussion started, and will facilitate the discussion in various ways. We’ll have some iterated discussion in a few themed rooms, shuffling once or twice to let people explore different topics.
Attendees will have the ability to participate in “webinar” format, some way for them to ask questions, get them upvoted or whatever. These questions and the discussion will be moderated by the organizers. To attend, please register with HotOS.
You can read more in the official panel description.
Background, Aim, and Scope
The Unix shell is the de facto choice for specifying succinct and simple scripts for system orchestration, automation, and data ingestion. Ops trends in containerization and the cloud have brought shell scripting back to prominence, and thus motivate research of both theoretical and practical significance. The aim of this HotOS’21 Panel is to provide a forum for the discussion of new, renewed, and continued research in this area.
The focus will be on topics related to shell scripts and their runtime environments, and includes:
- Implementations and performance
- Parallelism and distribution
- Containerization, serverless, and microservices
- Tool support for the shell: analysis, transformation, and synthesis
- Annotations and type systems for tool support
- Mechanization and correctness
- Benchmarks and tools
The concrete topics of the panel will be informed by the interests of the panel’s participants. We’ll produce a summary of the event: participants will be able to contribute their perspectives and visions to the summary.
The 90-minute event (on June 3rd 2021 from 1–2:30pm ET) has a small panel of experts on shell-related topics. The event itself is discussion based.
|Panelists summarize statement||10min|
|Breakout in three topic-oriented rooms||30min|
|More breakout, attendees shuffle rooms||15min|
|Panelists summarize discussions for all attendees||15min|
The panelists will produce single-page statements of their interest in and vision for the shell, offering one or more challenge problems. We will distribute these statements in advance to participants.
Our panelists will be:
- Arjun Guha (Northeastern University)
- Deepti Raghavan (Stanford University)
- Chet Ramey (Case Western Reserve University)
- Diomidis Spinellis (Athens University of Economics and Business; Delft University of Technology)
We have confirmed 16 participants (updated May 18th):
- Carol Alexandru (University of Zurich)
- Andy Chu (Oil shell)
- Jürgen Cito (TU Wien)
- Tom Van Cutsem (Bell Labs)
- Ting Dai (IBM Research)
- Christos Dimoulas (Northeastern University)
- Elena Glassman (Harvard University)
- Vidar Holen (Facebook)
- Michał J Gajda (Migamake)
- David Karger (MIT)
- Julia Lawall (Inria)
- Doug McIlroy (Dartmouth College)
- Ruzica Piskac (Yale University)
- Jiasi Shen (MIT)
- Keith Winstein (Stanford University)
- Tianyin Xu (UIUC)
Organizers / Contact
- Michael Greenberg (Pomona College)
- Konstantinos Kallas (University of Pennsylvania)
- Nikos Vasilakis (MIT)
If you want to participate, please email the future-shell list.