Eric Snow

Hmm... ``` Execute: 'C:\Users\RUNNER~1\AppData\Local\Temp\tmpg12c4haj\Scripts\python.exe' -m azdev setup --cli 'D:\a\pyperformance\pyperformance\pyperformance\benchmarks\data\azure-cli' Fatal Python error: _Py_HashRandomization_Init: failed to get random numbers to initialize Python Python runtime state: preinitialized ```

FYI, using azure-cli on 3.9+ is currently blocked: https://github.com/Azure/azure-cli/issues/16611.

This is a follow-up to #155.

Great observation! tl;dr +1, but add an option to disable `--no-binary`. The kinds of benchmarks where dependencies matter are typically workload-oriented (macro benchmarks). We want to be able to communicate...

Also, presumably all the dependencies build on all our tested versions, based on the CI results.

FYI, I've merged my PR that allows running benchmarks that aren't part of pyperformance. I also have a PR up against the Pyston benchmarks repo to allow them to be...

+1 to making `pyperformance run` match `pyperformance compile` as closely as possible, including calling `pyperf system tune`. > My other pet peeve about it (that maybe we could fix while...

> Well, this branch is already getting pretty big. Pull a thread, unravel a sweater. How much work would it be to move the `imp` removal to a separate PR?...

Thanks for taking another look! I know what a pain it can be to split up a big PR.

IIRC, @gpshead & @Yhg1s have had some interest in this ifdef in the past.