Skip to content
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

Point to culprit library in error message for subnormals-are-zero #3458

Closed
Zac-HD opened this issue Sep 17, 2022 · 0 comments · Fixed by #3472
Closed

Point to culprit library in error message for subnormals-are-zero #3458

Zac-HD opened this issue Sep 17, 2022 · 0 comments · Fixed by #3472
Labels
legibility make errors helpful and Hypothesis grokable

Comments

@Zac-HD
Copy link
Member

Zac-HD commented Sep 17, 2022

We added a new warning in #3239, and it turns out that the underlying problem is much more common than I expected. In Someone’s Been Messing With My Subnormals!, Brendan Dolan-Gavitt found 49 packages compiled with -Ofast (most notably gevent/gevent#1864), with around 2.5k dependents between them.

My first thought was to intersect that set of of 49 known-ever-bad packages with sys.modules. My second was that we could actually just check each of the imported packages (using a subprocess) directly, which would be trickier and more expensive but always up-to-date.

Whichever we go for, I'd add a note on a command like pipdeptree -rp BTrees,gevent,quiskit-aer to the error message so that users can tell why they're using a library that they might never have heard of.

@Zac-HD Zac-HD added the legibility make errors helpful and Hypothesis grokable label Sep 17, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
legibility make errors helpful and Hypothesis grokable
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant