@@ -456,19 +456,31 @@ with the moment when they are added to the `Span`.
456456
457457An ` Event ` is defined by the following properties:
458458
459- - (Required) Name of the event.
460- - (Optional) [ ` Attributes ` ] ( ../common/common.md#attributes ) .
461- - (Optional) Timestamp for the event. If not provided, the current time when the event is added MUST be used.
459+ - Name of the event.
460+ - A timestamp for the event. Either the time at which the event was
461+ added or a custom timestamp provided by the user.
462+ - [ ` Attributes ` ] ( ../common/common.md#attributes ) further describing the event.
462463
463464The ` Event ` SHOULD be an immutable type.
464465
465466The Span interface MUST provide:
466467
467468- An API to record a single ` Event ` where the ` Event ` properties are passed as
468469 arguments. This MAY be called ` AddEvent ` .
469-
470- Events SHOULD preserve the order in which they're set. This will typically match
471- the ordering of the events' timestamps.
470+ This API takes the name of the event, optional ` Attributes ` and an optional
471+ ` Timestamp ` which can be used to specify the time at which the event occurred.
472+ If no custom timestamp is provided by the user, the implementation automatically
473+ sets the time at which this API is called on the event.
474+
475+ Events SHOULD preserve the order in which they are recorded.
476+ This will typically match the ordering of the events' timestamps,
477+ but events may be recorded out-of-order using custom timestamps.
478+
479+ Consumers should be aware that an event's timestamp might be before the start or
480+ after the end of the span if custom timestamps were provided by the user for the
481+ event or when starting or ending the span.
482+ The specification does not require any normalization if provided timestamps are
483+ out of range.
472484
473485Note that the OpenTelemetry project documents certain [ "standard event names and
474486keys"] ( semantic_conventions/README.md ) which have prescribed semantic meanings.
0 commit comments