But behind the scenes, there’s a lot more going on. I’ll describe some of what happens, and (much much more importantly!) explain some tools you can use to see what’s going on behind the scenes yourself. We’ll use readelf, strace, ldd, debugfs, /proc, ltrace, dd, and stat. I won’t talk about the Python-specific parts at all – just what happens when you run any dynamically linked executable.

Analyzing the characteristics of real-world applications is difficult. We often make generalizations about applications we see in the wild, both anecdotal and statistical: "Single-Page Applications are slower than multipage" or "apps with low TTI loaded fast". However, the extent to which these generalizations hold for the performance and architectural characteristics we care about varies. I believe one of the primary determinants of this variability are a product's features and design constraints, and classifying applications based on their features & constraints can enable more targeted and impactful solutions to the problems faced by each.