It Will Never Work in Theory (Online, April 2022)
2022-04-27
-- online --
who it's for
software engineers, engineering managers, technical leads, test engineers, developer experience researchers, platform engineers
topics
file order in code review; code review smells and bias; real-world TDD practice; variable naming patterns; enhanced compiler error messages; code coverage for test guidance; low-bandwidth and offline defaults; GitHub Actions adoption effects; expert debugging practices; automated database fuzz testing; remote onboarding practices; mining build and CI systems
| title | author | length |
|---|---|---|
| How code review works (and doesn't) in the real world | Alberto Bacchelli | 10m |
Empirical results show that file order in code review changes what gets inspected and what bugs get found. By instrumenting a code review tool and by analyzing… | ||
| Value and waste in software engineering | Matthias Galster | 11m |
Presents a concrete definition of waste with nine categories of waste and empirical causes, catalogs specific review smells — “looks good to me,” review… | ||
| The hidden costs and benefits of TDD | Davide Fucci | 08m |
A controlled study uses an instrumented IDE plugin and a real legacy system with external APIs to record how 60+ professional developers actually practice TDD… | ||
| How patterns in variable names can make code easier to read | Felienne Hermans | 09m |
Shows that linguistic code smells measurably raise cognitive load using brain scanners, defines categories like methods that do more or less than their name… | ||
| Automatically enhancing error messages | Christoph Treude | 08m |
A system called Pycee automatically queries Stack Overflow for matching answers, rewrites variable names, and summarizes the accepted responses into enhanced… | ||
| How code coverage can be used and abused to guide testing. | Maurício Aniche | 07m |
Argues that code coverage is useful when it augments test suites rather than serving as a hard target, and that running a code coverage tool to find uncovered… | ||
| Challenges and opportunities for software engineering in Papua New Guinea. | Sebastian Baltes | 11m |
Workshop uses surveys and focus groups and finds infrastructure and education shape what software can actually do in Papua New Guinea: 30% internet access… | ||
| How people really use GitHub Actions | Marieli Wessel | 09m |
Study of 3000 projects finds 700 actions across 20 categories and shows typical actions are added twice and then rarely changed, turning automation into a… | ||
| How expert programmers think about errors | Marian Petre | 11m |
Experts treat errors as information: they wait instead of immediately swatting a bug, build errors into code to test the test harness, and use unexpected… | ||
| How to test software without writing tests | Manuel Rigger | 10m |
Claims an automated tester called SQLancer that pairs random query generation (it reuses tools like SQLsmith) with a metamorphic oracle named ternary logic… | ||
| Programmer by day, tester by night | Andy Zaidman | 08m |
A plugin for popular IDEs collects 150 person-years of activity from 2442 developers and shows self-reports of ~50/50 test versus production time but measured… | ||
| What we've learned about remote onboarding during the pandemic | Paige Rodeghero | 09m |
A study at Microsoft surveys new hires during early pandemic hiring, sending ~1,000 emails and collecting 267 responses, and distills seven simple rules that… | ||
| Bias in evaluating code contributions | Mei Nagappan | 10m |
Claim: inferred race and ethnicity from names using NamePrism on about 2.5 million pull requests across 46,000 GitHub projects yields concrete disparities. The… | ||
| The unintended consequences of mining software build systems | Shane McIntosh | 10m |
A tool called Build Medic consumes build failure logs, classifies failures, selects repair strategies and rewrites build scripts (Maven examples, Log4j fixes)… | ||
| Causal testing: understanding the root causes of defects | Brittany Johnson-Matthews | 11m |
Claim: automated causal experiments take existing failing tests, use fuzzing and automated test generation to produce minimally-different passing and failing… | ||
| Fostering a learning culture in coding teams | Cat Hicks | 09m |
Interviews with 25 developers and 30+ hours of think‑aloud coding sessions produce a concrete claim: teams create learning debt when onboarding and review… | ||
| Negotiation and padding in software project estimates | Igor Steinmacher | 11m |
A study claims negotiation and padding routinely reshape software estimates. Researchers spend three months observing four teams and conduct followup… | ||
| The effects of destructive criticism in code review | Kelly Blincoe | 09m |
Study uses an online questionnaire with vignette-style code-review scenarios and compares constructive versus destructive wording. Main claim: destructive… | ||
| Online community and safety in software engineering | Denae Ford Robinson | 12m |
Survey data show harassment is common — Pew: 4 in 10 Americans report online harassment — and that open source projects for social good face polarized… | ||
| To search or not to search: it depends on the question | Katie Stolee | 10m |
Developers run code search about 12 sessions per day, each session averaging three to four queries, and a study categorizes searches as how (≈33%), what… | ||
| What does 'productivity' actually mean for developers? | Margaret-Anne Storey | 11m |
Argues that developer productivity is multidimensional and not reducible to simple telemetry the core claim is the SPACE framework that breaks productivity… | ||
| nothing matches | ||