Étienne Fortin

In theory yes. It use another IL opcode that calls the method instead of a virtual call to the Invoke() method behind. However in practice I wonder if the JIT...

I'm looking at using Nullable Reference Types as well. From what I understand though it's a compiler only thing. It will not validate at runtime. So it's only usable at...

Maybe at the end of porting the whole thing to generics, the current genericless classes should inherit from the SomeUnit one. And in theory every users of the types fall...

I did a double double implementation. Maybe it could be added to UnitsNet.

You are right. Before going forward with such an extensive change we should make sure it has benefits accross the board. For me it just make sense to have that...

Now that Net 7 preview 5 is 9ut generic math are a lot more advanced in their development. There exist an IFloatingPoint interface that constraint numbers to floating point types...

I don't know if it's here that we should discuss breaks in v5 but here I go. Yesterday I upgraded to v5 alpha 3 to test it on one of...

> @ebfortin Thanks, could you please report back with the exact error and what you would expect instead? There are many breaking changes in v5, but if you found something...

Hmmm ok I see. At first I wrote that we should keep == and IEquatable since we have == for doubles and when you use it you know there are...

Do not throw ArgumentException on NaN, Inf+ or Inf- or give a way to configure the behavior globally. Context: https://github.com/angularsen/UnitsNet/issues/1268#issuecomment-1630856242