windows3-4 Exercise 2: Find What Your Patterns Missed ================================================================= Verification: Level 1 (run on the author's machine, read-only, no administrator rights). Prediction (work this out before running the script): Before running: look at the 'Everything else' category and guess which familiar programs are in it. Then predict which category will grow when you add a missing pattern. Commands (workload.ps1 saved in the current folder). Run it as it is, read the last line, then change the Cloud sync pattern to '^(onedrive|onedrive\.sync\.service|dropbox|googledrivefs|box)$' and run it again: .\workload.ps1 Output (real, from the author's machine on 2026-10-24; your numbers will differ): First run (original patterns): Windows itself 120 procs 1,986 MB 28% Browsers 28 procs 1,590 MB 22% Everything else 100 procs 1,327 MB 19% Cloud sync 1 procs 856 MB 12% Editors and dev tools 24 procs 809 MB 11% Security 6 procs 346 MB 5% Embedded browsers 20 procs 189 MB 3% TOTAL 299 procs 7,104 MB Largest in "Everything else": claude 304 MB, lingma 217 MB, slack 176 MB, onedrive.sync.service 111 MB, secure system 101 MB Second run (onedrive.sync.service added to Cloud sync): Windows itself 120 procs 1,988 MB 28% Browsers 28 procs 1,595 MB 22% Everything else 100 procs 1,225 MB 17% Cloud sync 2 procs 967 MB 14% Editors and dev tools 24 procs 811 MB 11% Security 6 procs 344 MB 5% Embedded browsers 20 procs 189 MB 3% TOTAL 300 procs 7,119 MB Largest in "Everything else": claude 300 MB, lingma 207 MB, slack 176 MB, securityservice 104 MB, secure system 101 MB WHY THIS WORKS AS AN ANSWER --------------------------- The first run's last line shows onedrive.sync.service (111 MB) sitting in 'Everything else', a program that belongs with the sync client. After the pattern change the Cloud sync category is larger by about that amount (856 MB, 12%, to 967 MB, 14%) and the Everything else category is smaller by about the same. The other megabyte figures differ between the two runs by tens of megabytes because the machine was running in between: the browsers and the embedded browsers each moved. That is why the exercise asks which category changed, not for exact megabytes, and why comparing runs only makes sense for the category you edited. Common mistakes: judging a pattern by the main process name only; ignoring the catch-all list; and treating a change of 5 or 10 MB between runs in a category you did not edit as meaningful.