site banner

Small-Scale Question Sunday for May 3, 2026

Do you have a dumb question that you're kind of embarrassed to ask in the main thread? Is there something you're just not sure about?

This is your opportunity to ask questions. No question too simple or too silly.

Culture war topics are accepted, and proposals for a better intro post are appreciated.

2
Jump in the discussion.

No email address required.

What is the right way to talk about the size of a programming project?

I recall someone in a thread a while ago making a dismissive comment about amateurs with AI boasting about how many lines of code they've written, which I think was either referring to me talking about my own vibe-coded project or possibly Garry Tan boasting about 37k LOC per day, which does seem slightly excessive.

I know that less is more, efficient code is better than lengthy code and that AI tends to leave a bunch of comments in there too... but how can you measure the complexity of a project in any quantitative sense besides lines or just outright KB/MB? I'm not a 'real' software engineer but it seems like once you have 134 files and 3.4 MB of code, you can't really count functions in any useful way, what else is there but lines and size?

I usually people hear their projects called "small, medium or large." There's not a hard definition. Small projects can be comfortably worked by one person. Large projects probably need at least a team, maybe multiple. Medium projects probably need a small team.

Probably when you talk about project size you need to be more specific. Larger projects tend to:

  • have more consumers
  • change slowly rather than rapidly
  • be deployed as multiple services
  • have multiple experts who know sub-systems, rather than one expert who knows everything
  • have long release cycles, requiring batched testing approaches

These are not caused by "high LOC." Perhaps I'm describing "(poorly written) enterprise software." But these are still things people sometimes mean when they talk about "large project" or "small project."

Any metric would be opposed because it would allow direct comparison of a human, an LLM, and a centaur. If one's work is evaluated only qualitatively, than there is room for argument, as there is no rigourous dispute resolution criteria.

It is like with teachers, who prefer that results on standarized tests by their stuents not be thought of as valid criteria for evaluating how good of a job they are doing.