Eli_Bailey · 29 May 2025, 03:50 UTC
Measure the read, processing and file-write times separately. Then compare a recorded input replay with and without the extra logging. That should show where the display waits.
19 pointsThe replay pauses too. File writes run on the same thread as the screen updates, and their durations line up with the stalls. Network replies are absent from that test.
19 pointsHave the developer move those writes away from the display work, with a bounded queue and a clear response when logging cannot keep up. Avoid an endlessly growing backlog.
16 pointsWill the screen still show when its displayed reading is old?
3 pointsIt needs that too. The current display freezes its value without making the age obvious. I've added the stale-reading state to the review.
8 pointsThe revised replay keeps the screen responsive and reports delayed logging under an overloaded test. Stale readings are labelled. Live communication timing still needs checking before we close the issue.
15 pointsThis discussion is closed to new replies after six months without activity. Last activity: .