The defect https://scan8.coverity.com:8443/reports.htm#v21582/p10801/fileInstanceId=3240428&defectInstanceId=1396798&mergedDefectId=96018 found by Coverity is actually a false positive since the locking was not in place to give public thread safety but only to guarantee correct implementation, given the internal timer thread, but reading the code, it was difficult to tell. I considered suppressing the warning but then I realised making the type threadsafe was actually a trivial change and removes ambiguity for the reader so I thought to go down that route instead.