
Rules for readable code: short functions, clear names, one level of abstraction.
Controversial in its prescriptiveness, but useful precisely because AI-generated code routinely violates every heuristic in it: long functions, unclear names, mixed levels of abstraction. Reading Martin sharpens the vocabulary for what to fix in a model's output, even if the rules are applied with judgment rather than dogma.
Chosen by shared topics, the entries that hold them, and the people who recommended both.
Lessons from IBM's OS/360: why adding people to a late software project slows it.
“Arguably the only classic book in our field. If you haven't read it, shame on you.” — Jeff Atwood
Source: Recommended Reading for Developers — Coding Horror
Recommended by: Jeff Atwood, Mohnish Pabrai, Paul Graham and 4 others
A programmer's craft: when code is good enough and what to automate.
“Classic book for computer programmers. I read it first in 2003 before I was taking book notes, so I read it again now to take notes. Great wisdom in here. Amazing to see how much of its advice was adopted as norms by Ruby on Rails.” — Derek Sivers
Source: Derek Sivers book notes (primary, rated list) · Rating: 6/10
Recommended by: Derek Sivers, Marc Andreessen
Software's enemy is complexity, which grows through shallow modules and shortcuts.
“Programming wisdom: modular design, keep APIs simple, how to avoid complexity, etc. Great insights from experience.” — Derek Sivers
Source: Derek Sivers book notes (primary, rated list) · Rating: 8/10
Recommended by: Derek Sivers
A catalogue of moves that improve existing code without changing what it does.
MIT's introductory computer science text, taught through Lisp and Scheme.
Recommended by: Bret Victor, Patrick Collison, Paul Graham and 1 other
How open-source software is really made and maintained, often by a few people.
Recommended by: Balaji S. Srinivasan, Delian Asparouhov