Closed mrob95 closed 1 year ago
I have mainly been testing this with the austin
profiler.
@benfred Do you think this could be adapted to work with py-spy as well? It looks to me like py-spy removes a lot of the info which would allow us to tell the difference between user code and library code, but it would be great if we could make it work.
Base: 90.21% // Head: 90.33% // Increases project coverage by +0.11%
:tada:
Coverage data is based on head (
08a4194
) compared to base (8c0a28c
). Patch coverage: 98.41% of modified lines in pull request are covered.
:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
Was it intentional that you also made changes to the flamegraph
submodule?
Was it intentional that you also made changes to the
flamegraph
submodule?
Whoops, nope. Reverted that. To be honest I had forgotten about this PR!
Heh, yeah, sorry, I've been falling really behind on OSS stuff!
BTW, you may be glad to hear that inferno helped us track down and report an O(n^2) in boto3
:)
Published in 0.11.11 :tada:
Adds a new colour palette for python. Aiming to distinguish four cases:
I've tested it out in a few scenarios and it definitely makes the output more understandable at a glance: