Log files are in chronological order and that's how merge-logs displays them, but during debugging we're typically interested in the most recent events (as those are closer to when the problem occurred).

For large datasets, this can mean waiting for longer stretches until the end of the buffer renders.

An option (perhaps even set as default) to parse the logs in reverse chronological order seems useful, though it might be weird to mentally switch back and forth if merge-logs does not become our one and only way to look at the logs (I personally think it will be my one and only way).

Jira issue: CRDB-6350

Epic CRDB-20791