Tom
AnisseI'll do benchmarks using my own bzcat and comparing that with bzip2's original bzcat. All running on an M1 laptop running Linux. The M1 CPU has both performance and efficiency cores. We'll run the benchmark on both, using hyperfine to do the comparisons. 3/N
Hugo van Kemenade@simon Here's a real-life application of the new tail-calling interpreter in 3.14.0a5, a full build of the CPython docs is 1.26 times faster, saving 7.5 seconds!
Nikhil 🐧just discovered this CLI benchmarking tool and now I have the urge to benchmark EVERYTHING 😆
Jan 
Performance Optimizer Observation Platform - by Andrew Kelley (creator of #ZigLang)
https://github.com/andrewrk/poop
Run your benchmarks in your #CLI!
It's similar to #hyperfine, but with more statistics reported.
⚠️ Linux-only
HishamHow do people #benchmark code speed nowadays? Say I want to compare code branches that have individual runs in the 400ms range.
I'm running them with #hyperfine, but with things like 50 runs the runtime variation while running on a laptop is huge; even when plugged in and in the "performance" mode, the throttling seems to dominate the variation.
Should I just go for really long runs to mitigate the throttling? (like 1 min warmup, plus 5 min worth of executions?)
Choan Gálvez@mrblissett And now #til about #hyperfine. Double win!
Thx :)
Jan de Muijnck-HughesAny #hyperfine users have pointers on plotting the raw data from the output, that doesn’t involve the existing scripts. Trying to find out what others do.
Miguel Madrid-Mencía 🦫More #Rust #CLI #tools:
#navi ⬅️ man+fzf
#hyperfine ⬅️ benchmarking
#czkawka ⬅️ duplicates finder
#mdcat ⬅️ markdown render
#tectonic ⬅️ latex engine
#grex ⬅️ RE generator
#watchexec ⬅️ run commands on file modifications
#xh ⬅️ httpie
#just ⬅️ command runner
#tokei ⬅️ cloc
#sd ⬅️ sed
#amber ⬅️ sed
#dua ⬅️ ncdu
#pueue ⬅️ parallel
#navi ⬅️ man+fzf
#hyperfine ⬅️ benchmarking
#czkawka ⬅️ duplicates finder
#mdcat ⬅️ markdown render
#tectonic ⬅️ latex engine
#grex ⬅️ RE generator
#watchexec ⬅️ run commands on file modifications
#xh ⬅️ httpie
#just ⬅️ command runner
#tokei ⬅️ cloc
#sd ⬅️ sed
#amber ⬅️ sed
#dua ⬅️ ncdu
#pueue ⬅️ parallel
skunksarebetter#hyperfine is a #CLI benchmarking tool.
hyperfine benchmarks CLI programs by running them repeatedly many times, performing statistical analysis on the results to return a final estimate of the runtime. hyperfine can perform warm benchmarking, perform actions before benchmarks, and can insert variable numeric parameters. Results can be output in several formats, including #JSON.
Website 🔗️: https://github.com/sharkdp/hyperfine
apt 📦️: hyperfine
