`class Quantum`

Represents the smallest possible time interval that can be measured given the current system overheads.

Definitions

`def self.resolve`

Resolve the quantum for the current system.

Signature

returns　Float: The quantum for the current system in seconds.

Implementation

def self.resolve
	self.new.to_f
end

`def to_f`

Signature

returns　Float: the quantum for the current system in seconds.

Implementation

def to_f
	precision
end

`def precision`

Lazily compute the precision of the current system.

Signature

returns　Float: the quantum for the current system in seconds.

Implementation

def precision
	@precision ||= self.measure_host_precision
end

`def measure_host_precision(repeats: 1000, duration: 0.1, factor: 2.0)`

Measure the precision of the current system.

Signature

parameter　repeats　Integer: The number of times to repeat the measurement.
parameter　duration　Float: The duration of the measurement in seconds.
parameter　factor　Float: A scaling factor to apply to the measured overhead.
returns　Float: the quantum for the current system in seconds.

Implementation

def measure_host_precision(repeats: 1000, duration: 0.1, factor: 2.0)
	step = duration / repeats
	
	start_time = self.now
	repeats.times do
		sleep(step)
	end
	end_time = self.now
	
	actual_duration = end_time - start_time
	
	if actual_duration < duration
		warn "Expected to sleep for #{duration}+ seconds, but only slept for #{actual_duration} seconds."
		error = 0.1
	else
		# Total execution time, minus the expected execution time = overhead.
		error = actual_duration - duration
	end
	
	# It's not like this is a precise measurement anyway...
	return error * factor
end