Performance test that simulates a long conversation.

[jacob314]

For the performance test we should load a 60+MB conversation with at least 1400 items.

Things to time:
Time to load the conversation.
Latency after typing each key in the input prompt after the conversation loaded. Must be under 30ms.
Time to execute a simple command such as echo hi
We should execute the command N times so we measure speed to execute as well as time for first shell command to run.

Measure time intentionally scrolling the conversation history (only applies to terminalBuffer and alternateBuffer rendering.