Closed ndokos closed 1 year ago
The failures are ... weird. But they make me wonder if we need to cherry-pick my Flask/Werkzeug requirements.txt
changes back to b0.72. 😦 Although the signature is bizarre and somewhat different, it's possible that Flask 2.3.1 did partially resolve the issues causing a less direct failure mode.
OK, now I don't know what to think. I don't see these on a local branch pulled from b0.72, with Flask 2.3.1 and Werkzeug 2.3.1. Since your build however used Werkzeug 2.3.0 ... it's possible they just published a fix. I'm going to re-submit the build.
The working case used Werkzeug==2.3.1
while the failing case used Werkzeug==2.3.0
(all other versions looked the same)...so perhaps we can uncap Flask and Werkzeug on main
now?
The working case used
Werkzeug==2.3.1
while the failing case usedWerkzeug==2.3.0
(all other versions looked the same)...so perhaps we can uncap Flask and Werkzeug onmain
now?
You might take a look at #3409, which has been up for an hour. 😁
You might take a look at #3409, which has been up for an hour. 😁
Just did. (I'm working in chronological sequence...relative things like "an hour" are meaningless.... 🙃)
Fixes #3404
Add pbench-tools-kill to the click commands in the agent Makefile.