Decide how to represent objects and values in log messages

[lawrence-forooghian]

Background

The log messages emitted by the SDK frequently wish to include objects and other non-primitive values. For example, the inputs and outputs described in #1623, as well as objects used to make the internal decisions described in that issue.

For example, these might be:

• arguments of methods
• objects such as ProtocolMessages received from or sent to Ably
• objects like ARTRetryAttempt which describe a decision taken by the SDK

We need to decide the process by which these objects should be serialized to a form that can be logged, and what factors might affect this process. This will affect the internal logging API (#1617) and potentially the public logging API (#1628).

Factors to consider:

• Whether to log a whole object or a subset of its properties (e.g. unique ID, or memory address) — this will depend on factors such as whether we’ve already logged the full object at some point or not, or how crucial for debugging it is to have access to all of the properties of the object
• Whether any data should be redacted (see Understand SDK users’ requirements for log output #1618)
• Whether to emit objects in a machine-parsable format (e.g. JSON) so that the emitted logs are easy to later parse and analyse (see Understand Ably SDK developers’ requirements for log output #1625)

What factors (e.g. configuration, log level) might affect the above decisions?

┆Issue is synchronized with this Jira Task by Unito

[sync-by-unito]

➤ Automation for Jira commented:

The link to the corresponding Jira issue is https://ably.atlassian.net/browse/SDK-3519