benchmark-memory

benchmark-memory copied to clipboard

Should GC.start be run before reporting results?

7

Coming from https://github.com/michaelherold/benchmark-memory/issues/9. Output was confusing without `GC.start`.

dblock

Ruby 2.7 added the ability to compact the heap during garbage collection so that memory is less fragmented, thereby lowers the necessary memory for a program. I'm wondering if there...

michaelherold

I toyed with the idea of calculating memory "churn", which I define as the ratio of retained memory to allocated memory. This is trivial to calculate but I'm not sure...

michaelherold

Currently, the output looks a little wonky because we're showing object counts and string counts as floating point numbers so they are always displayed with three zeros after the decimal...

michaelherold

Total memory usage can't show the whole picture, so we should measure GC pressure as well. At the very least, we want to know how much time is spent in...

michaelherold

Order comparison results by baseline similar to benchmark-ips

2

When I'm writing benchmarks, I almost always want to compare against an existing result. [benchmark-ips](https://github.com/evanphx/benchmark-ips) supports this with: ``` Benchmark.ips do |x| # ... x.compare! order: :baseline end ``` I'd...

technicalpickles

Allow comparison against a baseline

1

This change introduces a new comparison ordering strategy where the initial report serves as the baseline against which the job compares the other reports. This is useful in cases where...

michaelherold

Bumps [rexml](https://github.com/ruby/rexml) from 3.3.6 to 3.3.9. Release notes Sourced from rexml's releases. REXML 3.3.9 - 2024-10-24 Improvements Improved performance. GH-210 Patch by NAITOH Jun. Fixes Fixed a parse bug for...

dependabot[bot]