Fix UTF-8 character rendering in logs

This commit fixes a bug where multi-byte UTF-8 characters in the log view were being displayed as a sequence of individual, incorrect characters.

The issue was caused by improperly decoding the base64-encoded log data from the server. The `atob()` function was used directly, which decodes to a raw byte string, not a UTF-8 string. This broke the encoding for any character outside the basic ASCII range.

The fix introduces a proper UTF-8 decoding step after the base64 decoding. It uses the `decodeURIComponent(escape(...))` idiom to correctly interpret the byte stream as UTF-8, ensuring all characters render correctly in the browser.