Jimmy Bogard

TBQH I don't think about that scenario really? On Sat, Feb 18, 2017 at 3:47 AM, dotnetshadow wrote: > Firstly a great sample to learn about mediatr > > I...

Also consider adopting the [recommended tags and naming conventions](https://github.com/open-telemetry/opentelemetry-specification/blob/master/specification/trace/semantic_conventions/messaging.md) from OpenTelemetry. And if you need an example of dealing with ActivitySource, I added support for the Trace Context standard and...

I probably should just remove those helper classes, they seem to confuse more than anything. I only test handlers through the mediator, which is why I never hit this.

> Why is it named `Unit.Value`? Especially if it hasn't any value? I don't remember to be honest. The Rx one is called "Unit.Default", which is equally weird to me....

Ah I remember why they can't be public. The overload only differs in the return type. The way to "fix" this would be to make the method public and rename...

> I cannot confirm that, it seems to work just fine Changing the access modifier to public does not work. Nor does removing the explicit interface implementation: ```csharp public abstract...

Honestly, I'd rather just remove these classes than anything else. I don't use them anyway and they just seem to cause more problems than solve.

You're not forced to use a mediator instance, that's not what this issue is. The issue is that you have to cast to get the explicitly implemented interface handler.

Another route would be to only target `netstandard2.1`, that would be OK with me too, especially given the `net5.0` direction.

There would be a separate interface and overload.