gh-115637 clarify that time.perf_counter cannot go backwards - #115638
gh-115637 clarify that time.perf_counter cannot go backwards#115638anarcat wants to merge 2 commits into
Conversation
Sorry, something went wrong.
|
All commit authors signed the Contributor License Agreement. |
Sorry, something went wrong.
I'm not completely sure about this, but it seems to me this counter is *also* designed to avoid jumping backwards on (say) daylight saving or clock changes. It's kind of hinted to in the current description, but reading PEP 418 it's very clear that it's part of the intention, as it falls back to time.monotonic() apart from on `nt`.
|
@arnarcat Thanks for the draft PR. Given your discussion with @vstinner on the issue, do you want to move this out of draft or close the PR? |
Sorry, something went wrong.
|
i'm not sure. i guess i'm waiting on other folks to weigh in on the issue first... |
Sorry, something went wrong.
|
This PR is stale because it has been open for 30 days with no activity. |
Sorry, something went wrong.
|
|
||
| .. impl-detail:: | ||
|
|
||
| On CPython, use the same clock as :func:`time.monotonic` and is a |
There was a problem hiding this comment.
This is documented here.
Sorry, something went wrong.
There was a problem hiding this comment.
but that's different from the patch I was suggesting though. below is an "implementation detail" and above is more of a specification...
Sorry, something went wrong.
|
Closing as per note on issue. |
Sorry, something went wrong.
I'm not completely sure about this, but it seems to me this counter is also designed to avoid jumping backwards on (say) daylight saving or clock changes. It's kind of hinted to in the current description, but reading PEP 418 it's very clear that it's part of the intention, as it falls back to time.monotonic() apart from on nt.
📚 Documentation preview 📚: https://cpython-previews--115638.org.readthedocs.build/