Chip Hogg

This should help catch a wider variety of warnings.

Right now, all of our rounding functions (the `round`, `floor`, and `ceil` families) are designed as "unit-aware" analogues of their standard library counterparts. This has the galling consequence that for...

Provide more flexible alternative to integer_quotient()

1

The point of `integer_quotient()` is to make it extremely obvious that we are about to do a possibly-truncating integer division. We have seen that this is very error prone when...

This breaks the 1:1 correspondence between target and test, so we split up our helper accordingly into two. We make `au` the single mega-target that holds everything that doesn't depend...

Many things are easy for Magnitude based types that are virtually impossible for conventional numeric types --- for example, they can represent vast quantities without overflowing, and can even perform...

Switch from `#pragma once` to include guards

3

This would be a more robust solution. `#pragma once` is fine for first-party code in a project, but for third party users, there will always be edge cases.

Try reducing new units when incorporating into common units

1

See [this example](https://godbolt.org/z/1GPnq4v6j): ```cpp int main(int, char**) { std::cout

See: - DoozyX/clang-format-lint-action#73 - DoozyX/clang-format-lint-action#75 - DoozyX/clang-format-lint-action#78 For now, we will just disable this job. I think this is a fine intermediate solution, because I am the one making most...

Very impressive project! :slightly_smiling_face: Nice work. I noticed that you're treating radians and steradians as separate and completely independent quantities. This isn't right, though; the steradian equals one radian squared....

It appears that every release is associated with the same commit, which is also the latest commit on the main branch:  This means that when you click the `Compare`...