Re: Testing some singleton implementations

[Jonathan Bettencourt]

Is it just me or does the implementation of atomic.d look grossly 
inefficient and badly in need of a rewrite?