I understand and appreciate the analogy the author made. I wish more people thought about code complexity from first principles. That being said, I doubt chrom

I understand and appreciate the analogy the author made. I wish more people thought about code complexity from first principles. That being said, I doubt chromatic number is the right measure for software dependency graphs. Especially when there are hundreds of graph complexity metrics (tree-width, centrality measures, measures based on cuts and flows, etc.). Even just now for the first time I saw this thing called cyclomatic complexity: https://en.wikipedia.org/wiki/Cyclomatic_complexity https://en.wikipedia.org/wiki/Cyclomatic_complexity I'd be interested to hear from someone who has experience applying graph metrics to static code analysis in practice. What's useful and informative?