# Surprising test timings

**URL:** <https://discourse.julialang.org/t/surprising-test-timings/129437>\
**Category:** Tooling\
**Tags:** testing\
**Created:** [May 29, 2025, 1:53pm UTC](https://discourse.julialang.org/t/surprising-test-timings/129437 "2025-05-29T13:53:28Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![TimG](https://avatars.discourse-cdn.com/v4/letter/t/82dd89/32.png) [@TimG](https://discourse.julialang.org/u/TimG)\
**Post date:** [May 30, 2025, 8:04am UTC](https://discourse.julialang.org/t/surprising-test-timings/129437/4 "2025-05-30T08:04:17Z")

</div>

I learned yesterday ([here](https://discourse.julialang.org/t/type-instability-with-type-t-selectors-in-keyword-arguments/129439/5)):

> Keyword input types _are_ specialized for compilation (this is different from the “method specialization” where you _write_ multiple methods with varied positional arguments)

The functions in the `Conditional Formats` tests make extensive use of a mixed set of keywords to control the formatting options. There are very many combinations of different keywords possible. I guess this means that almost every invocation of each function will need compiling separately. I modelled this (very loosely) on my experience with Makie.jl.

Maybe there is a more efficient way?

---

_[View the full topic](https://discourse.julialang.org/t/surprising-test-timings/129437)._