>>I would be interested in how he can detect session timeout within a reasonable timeframe. Timing of session timeouts is unreliable unless there's a logout button, and even then it's unreliable. Not being facetious here... I have a serious interest in this because I could benefit from the knowledge.
>
>My guess is Michel looked for "simultaneous" activity from both Kevin and "Grover" (which would mean they're different people), and didn't find any. So he believes they are the same person.
>
>Message#1516095

So, proving !(!(guilt)) instead of proving guilt.