[Eng] use a proper logging framework (ideally Open Telemetry)

[ohadschn]

print just doesn't cut it, here is a quick AI summary explaining why:

https://share.google/aimode/xdCXo7IZOIh0SGrKY
Moving from print to a logging library is like upgrading from a megaphone to a professional broadcast system.

• Provider Versatility (Sinks): Unlike print, which only targets the terminal, libraries like Loguru or structlog allow you to send data to multiple destinations simultaneously. You can show color-coded logs in the console, write rotating archives to local files, and ship JSON payloads to a database or cloud provider—all with one call.
• Granular Verbosity: Logging uses levels (DEBUG, INFO, WARNING, ERROR, CRITICAL). You can keep your console clean by showing only INFO while simultaneously recording deep DEBUG traces in a hidden log file for troubleshooting.
• Structured Metadata: Loggers automatically inject context that print ignores, such as timestamps, file names, line numbers, and even the specific thread or process ID where the event occurred.
• Production Readiness: You can globally toggle or mute logs without hunting through your codebase to delete individual statements. This ensures your production environment stays performant and your terminal isn't cluttered with "debugging noise."
  you want things like multiple log providers (console, file, DB etc.)

Specifically for a desktop application development like this one:

• Visibility & Support: While print output is invisible to users once the app is bundled, a logging library allows you to create an "Export Logs" button, making it easy for users to provide diagnostic data to your support team.
• Crash Recovery: print data is often lost during a system crash. Logging libraries can be configured to flush to disk immediately, ensuring the final moments before a "hard crash" are preserved for debugging.
• Remote Troubleshooting: Instead of relying on manual user reports, loggers can be integrated with automated telemetry (like Sentry) to report critical errors from the field as they happen.
• Disk Space Hygiene: Unlike print redirected to a file—which can grow indefinitely—logging libraries manage rotation and retention, automatically deleting old files to prevent the app from consuming the user's hard drive.
• Live GUI Feedback: Loggers can feed a built-in diagnostic console within your app's UI, allowing advanced users to troubleshoot issues like connection timeouts without needing external tools.

Why OpenTelemetry is the "Gold Standard"
Unlike a standard logging library that just writes text to a file, OpenTelemetry focuses on correlation across your entire system.

• The "Three Pillars": OTel unifies Logs (what happened), Traces (the path of a request across services), and Metrics (numerical health data like CPU/RAM) into a single, standardized data model.
• Trace Correlation: If your app crashes, OTel can link that specific error log to a "Trace ID." You can then see every step the app took across different threads or microservices leading up to that crash.
• Vendor Neutrality: You instrument your code once using the OpenTelemetry SDK. You can then send that data to any backend—like Grafana, Datadog, Jaeger, or Sentry—simply by changing a configuration file, without touching your Python code.