What Code Review Looked Like
The team chose to use a modified Fagan-style code review. Each code review had a moderator and several reader-reviewers, and the author was the scribe. In a true Fagan code review, the reader-reviewers would each read the code individually in advance, and read the code out loud at the review. But this team chose not to read the code out loud during the review. The moderator would ask, "Does anyone have any issues on page 1?" Everyone would respond with "No," or "Yes," and explain his issue. Similar to a Fagan code review, the reviewers would hand their small issues to the author at the beginning of the review.
One of the reasons for not reading the code aloud was to save time. If the developers had chosen to read every line of code aloud, the team would have had to schedule many more code reviews, because each one would have taken much longer. With these code reviews, the team could review 50 pages of code in one sitting.
The team realized they had to trust each other to explain at the beginning of the review if someone hadn't finished reading the code. In those cases, the review stopped where everyone had finished reading and another review was scheduled for the next day.
If the author had concerns about how he was supposed to fix a problem in the code, the author and the moderator managed those concerns together.
Get Beautiful Teams now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
