v3.3.0

• Removed the odd default insertion of an extra zero bucket only for duration histogram bucket definitions, this was causing confusion for users

v3.2.0

• Add instrument package with Call that provides convenient instrumentation for a given function in terms of a success and error Counter and latency Timer for successful calls

v3.1.0

• Add ability to sanitize metric identifiers, both name, tags and values by using SanitizeOptions field of ScopeOptions when constructing a scope

v3.0.2

• Fix for m3 reporter sending on a closed channel when metrics are reported to it after it Close() is called on it

v3.0.1

• Fix for Close() on the closer returned from NewRootScope(...) (and other constructors) not emitting any final buffered values before closing; it will now flush the last of its buffered values to the reporter immediately before attempting to also close the reporter

v3.0.0

• Fix Snapshot() Snapshot on TestScope not providing the tag name and values as part of the key for the different maps returned for counters, timers, gauges and histograms
• Add Histograms() map[string]HistogramSnapshot to Snapshot to provide the ability to inspect histogram values when using a TestScope

v2.1.0

• Update stopwatches implementation to use a zero allocation Stopwatch for both timers and histograms
• Fix possible flush after close semantics for scope

v2.0.0

• Add native histogram support with support for all reporters
• Consolidate NewRootScope and NewCachedRootScope to a single NewRootScope method that takes ScopeOptions to avoid future breaking changes to the constructor
• Update documentation with current interfaces

v1.1.0

• Fix duplicate scope.Tagged(...) scopes creating new scopes instead of using existing scopes
• Add multi reporter module for reporting a scope to multiple backends
• Cleanup prometheus reporter module and add ability to use histograms or summaries for timers
• Add m3 reporter module for reporting to a m3 collector

v1.0.0

v1.0.0