-
Notifications
You must be signed in to change notification settings - Fork 148
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use monotonic clock. #150
base: master
Are you sure you want to change the base?
Use monotonic clock. #150
Conversation
Thanks for this! It's going to take me a little bit to review it carefully, but I will do so as soon as I get a chance. |
monotonic_time=None, | ||
sleep=None, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Random dude passing through here, but IMO it would be cleaner to just declare monotonic_time=time.monotonic, sleep=asyncio.sleep
and then just use monotonic_time()
and sleep(n)
everywhere instead of the (monotonic_time or time.monotonic)()
and (sleep or asyncio.sleep)(seconds)
calls that lack legibility and clarity.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I definitely agree with you that it would be cleaner. If no one has issues with having the interface be explicit about the default functions that are being used I'm happy with changing that.
If we want to keep having None
to indicate "use whatever default the implementation thinks is best" we could have something like sleep_function = time.sleep if sleep is None else sleep
somewhere and use that.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'd go with the Zen of Python on this: Explicit is better than implicit. Users shouldn't have to guess what the default is when it can be right there in the definition.
The only reason to use None
as a default argument and override it with the actual default in the body of the function is if your default is mutable and could lead to unexpected behaviour, such as a dict
or list
. That's the only case I know of where it's standard to go arg=None
and then set the actual default in the body:
if arg is None:
arg = {}
Unless I'm missing some specifics here of course.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looking at changing this I remember why i did it like this. The unit tests monkeypatch time.sleep
in a lot of places, and if you pass it as a default value to the functions then it makes monkey patching much harder because it needs to be done before the module loads instead of simply before calling the function that is going to use time.sleep
.
I am going to have to refactor most unit tests to pass the sleep
argument instead of monkey patching.
Fix to allow trio event loop sleep.
Fix to allow trio event loop sleep.
This fixes #149 and #125.
This pull request proposes the use of
time.monotonic
by default instead ofdatatime.now
. A monotonic clock has the garantee of never going back in time, which is what we want when computing elapsed time.This also has the benefit of cleaning up tests, because we can easily patch both
time.sleep
andtime.monotonic
so that the latter moves forward by exactly the amount of seconds that we pretend-slept for.This found a "bug" in one of the test which I changed, it requires special attention when reviewing this code. The amount of expect calls to
sleep
didn't seem logical to me and I believe tthey where the result of time imprecisions.