Emil Lundberg

Thanks for the review! I'll try to fix these things up in the next couple of days.

Alright, I think I've addressed all your comments now. > Looking at this, I think to me it might make more sense to do the parsing of the aliased command...

I took a stab at it and quickly ran into a couple of nontrivial issues. :slightly_smiling_face: For starters you'll have to pass the whole `Shell` as a mutable parameter to...

Hm... oh yeah, you're probably right. I figured I'd need to construct a `Simple` instance directly at some point, which has the io handles in its fields, but yeah, you...

Okay, [here](https://github.com/emlun/rush/pull/1) are the changes I ended up with to make that happen. What do you think, should I pull that in here?

That all sounds like good points and good ideas, thanks! I'll look into that.

No, I argue this is in fact always true, but that the _client device_ can also act as a [roaming authenticator](https://w3c.github.io/webauthn/#roaming-authenticators) in some contexts. Whether a given authenticator is a...

That example is described in [§6.2.1. Authenticator Attachment Modality](https://w3c.github.io/webauthn/#sctn-authenticator-attachment-modality): >Some [platform authenticators](https://w3c.github.io/webauthn/#platform-authenticators) could possibly also act as [roaming authenticators](https://w3c.github.io/webauthn/#roaming-authenticators) depending on context. For example, a [platform authenticator](https://w3c.github.io/webauthn/#platform-authenticators) integrated into a...

Ok, those are fair counterexamples.

It doesn't appear to be explicitly documented in #1276, but I think it seems likely that backwards compatibility is the reason we made it not `required` - RPs cannot rely...