Engineering discipline applied to code, as distinct from computer science.
12 topics · 21 curated works
No prior grounding assumed.
10+ Deploys Per Day: Dev and Ops Cooperation at Flickr
John Allspaw & Paul Hammond · 2009
Argues that deploying code many times a day is safer than deploying rarely, because it forces developers and operations to share tools and incentives…
+4 more at this level
Assumes you know the vocabulary.
Managing the Development of Large Software Systems
Winston W. Royce · 1970
Lays out a sequential design-then-build process for large software projects and, in the same paper, argues it fails without iteration and feedback…
+6 more at this level
Primary sources and full treatments.
The Protection of Information in Computer Systems
Jerome H. Saltzer & Michael D. Schroeder · 1975
Sets out design principles for building secure systems — least privilege, fail-safe defaults, economy of mechanism — arguing security has to be built…
+8 more at this level
12 of 21 works
John Allspaw & Paul Hammond
Argues that deploying code many times a day is safer than deploying rarely, because it forces developers and operations to share tools and incentives instead of treating a release as a handoff between adversaries — the talk that gave the DevOps movement its founding example.
Adam Wiggins
Distils a dozen concrete rules — config in the environment, logs as event streams, one codebase per app — for building software-as-a-service applications that scale and deploy cleanly, from patterns observed across hundreds of Heroku apps.
link checked 17 Sept 2026Marijn Haverbeke
Teaches JavaScript as a real programming language with its own idioms — closures, higher-order functions, the DOM as a tree to be manipulated — rather than a collection of snippets to copy.
link checked 17 Sept 2026Chad Fowler
Argues a server should never be patched or reconfigured in place, only ever replaced wholesale from a known-good image, because a long-lived server accumulates undocumented state that no one can reason about.
Martin Fowler
Argues that once server configuration is expressed as version-controlled code rather than manual steps, the whole discipline of testing, review and reproducibility that software already has can be applied to infrastructure.
link checked 17 Sept 2026Winston W. Royce
Lays out a sequential design-then-build process for large software projects and, in the same paper, argues it fails without iteration and feedback loops back to earlier phases — the passage later readers dropped when they turned it into 'waterfall'.
Richard I. Cook
Argues catastrophic failure in a complex system always requires the simultaneous occurrence of several small, individually harmless faults, so that any single root cause found after the fact is a retrospective simplification.
link checked 17 Sept 2026IEEE Computer Society
Organises software engineering into a fixed set of knowledge areas that a practitioner is expected to know, arguing the discipline is broad and mature enough to warrant a defined common body of knowledge like older engineering fields.
Joshua Bloch
Argues a good API should be as small as possible but no smaller, should fail fast, and should make the caller's mistakes hard to make — and that these properties matter more over an API's lifetime than any single clever feature.
link checked 17 Sept 2026Michael Armbrust et al.
Argues cloud computing's real novelty is economic, not technical — the illusion of infinite, instantly available capacity billed by the hour — and works out what that does to how systems should be architected.
link checked 17 Sept 2026Addy Osmani
Works through the classic object-oriented design patterns as they actually get used in JavaScript, arguing the language's prototypal, dynamically-typed nature changes which patterns are useful and how they should be implemented.
Jay Kreps
Argues the append-only log, not the database table, is the right shared abstraction for moving data between the many systems a modern company runs, and traces how that insight became Kafka.
link checked 17 Sept 2026