Handbook of Walkthroughs, Inspections, and Technical Reviewswriting

code-reviewsoftware-qualityinspectionspeer-reviewwalkthroughs
1990-01-01 · 1 min read · Edit on Pyrite

Co-authored with daniel-freedman, this book became the definitive practical guide to technical-reviews-and-walkthroughs — the systematic peer review of software artifacts. It went through multiple editions, with an earlier version appearing as the ethnotechnical-review-handbook-1979, and became standard reference material in organizations that took code review seriously.

Core Argument

The book builds on egoless-programming: reviews work only when participants can separate their personal identity from the code under examination. This is not a pious hope but an organizational design problem. The Freedman-Weinberg approach specifies the structural conditions that make ego-free review possible: distinct roles (author, moderator, reviewers), clear rules about what constitutes in-scope and out-of-scope commentary, separation of defect detection from defect correction, and explicit management of the social dynamics of the review room.

Three Forms of Review

The book distinguishes walkthroughs, inspections, and technical reviews as distinct practices with different purposes and dynamics. Walkthroughs are author-led — the author presents the work and the audience asks questions. Inspections follow Fagan's structured methodology — formal roles, checklists, metrics, and a defined process. Technical reviews fall between the two — more structured than walkthroughs, less formal than inspections. The taxonomy matters because organizations that treat all reviews as interchangeable tend to do none of them well.

Significance

The book predates modern code review tools (Gerrit, GitHub pull requests, Phabricator) by decades but established the principles they implement. The standard pull request workflow — submit code, receive line-by-line comments, revise, merge — assumes that developers can give and receive criticism of code without interpersonal breakdown. That assumption was not obvious; Weinberg and Freedman helped build the professional culture in which it became obvious. The "ethnotechnical" framing of the earlier edition reflected Weinberg's anthropological lens — reviews are a cultural practice, and understanding them requires attention to the social dynamics of the review room as much as to the code under examination.