Lines of code
Also known as LOC, SLOC
Definition
Lines of code is a count of source lines in a file, change, or codebase. Depending on the tool, the count may include blank lines, comments, generated files, or only executable statements, so the definition must be stated.
A size measure, not a value measure
Lines of code can describe how much text changed. It cannot, by itself, tell you whether the change solved an important problem, reduced risk, or made the system easier to maintain.
An illustrative refactor may remove hundreds of duplicated lines while improving behavior. A generated client can add thousands of lines without representing thousands of units of engineering judgment. Counting both changes as comparable output would misread the work.
Where line count helps
Code size can provide a rough review-scope signal. A very large change may deserve decomposition, extra review time, or a clearer explanation. Repository size can also affect build and maintenance planning.
The useful question is what the count is being used to decide. If it is a prompt to inspect complexity or generated content, it can be helpful. If it is a quota or ranking, it encourages changes to the count rather than improvements to the software.
Use richer evidence
Read line counts alongside developer productivity, review time, tests, rework, defects, and delivered outcomes. Compare similar work and keep generated files separate when they distort the picture.
The amount of code written is sometimes relevant. It is never a complete description of the contribution that produced it.
How this relates to Weave
Weave analyzes the substance and estimated complexity of code changes rather than treating line count as a unit of engineering value. Code size can still help describe review scope or repository growth, but it should be interpreted with quality and outcome evidence.
Explore Engineering intelligence