Chip Hogg

> Should we support both syntaxes? > > ```c++ > quantity q1 = 1 * m; > quantity q2 = m * 1; > ``` > > Some time ago,...

> > The code below should work fine today: > > ```c++ > > if(num < 1 * Unit / 1'000'000'000'000) { > > ``` > > Unfortunately, I think...

Alternatively, if we wanted to use our own trait, we could use the chrono trait as the default implementation. This might be useful if we found some undesirable behavior in...

Good questions. I used them as motivation to write some docs I had been meaning to write, for our upcoming release: [Conversion Risks discussion](https://aurora-opensource.github.io/au/main/discussion/concepts/conversion_risks/); [Truncation discussion](https://aurora-opensource.github.io/au/main/discussion/concepts/truncation/). Each doc has a...

One thing I was thinking about here: if we simplify this and just delegate to `.as`, then it will affect the results of `std::is_constructible` for various `Quantity` types. The better...

I tried simplifying the constructors and running our tests, and uncovered some concrete use cases that depend on the status quo. - Basically all of `chrono_policy_validation_test.cc` uses this to test...

Ah! I understand better now. I was confused because I had explicitly deleted the constructor for disallowed conversions, which _should_ give a nice, clean error message. The real problem was...

Hmm... this still isn't a complete solution, because it still has the problem with making the `clamp` calls ambiguous. 🤔 Not yet sure where to go next.

Fascinating! I strove to make Au's interfaces a kind of "composable grammar", so people could learn the meaning of words like `as` and `in` and apply them consistently (e.g., `round_as`...

Something we'll need to consider: how will this interact with policy arguments? I think we want the default to include safety checks, so we'll want a policy argument version. But...