Kevin K.

Literally just edited the comment with that info :laughing:

A lot to reply to here :) I don't think we need to be concerned with embedded (bare metal) right now. Maybe one day, but there are many things we...

Thank you for the detailed response! Before I can finish reading the comment and add any thoughts, let me address something real quick as a moderation note: > from reading...

> If we take the case of vec_map, my question is why do we want to make it optional? Shouldn't it be either including it completely or removing it completely?...

> I think using `indexmap` in place of `vec_map` would make the performance worse. And we need `indexmap` for ordered map IIRC. That's entirely possible. I haven't done any investigating...

This might be pretty hairy to implement, especially if we end up trying to detect loops (`A` says display after `B`, and `B` says display after `A`). I'm leaning towards...

> I actually use it. The problem was that it don't play well with structopt flattening, more specifically it doesn't allow to move argument into separate struct while keeping its...

Much appreciated! Just off the top of my head, maybe you'd probably want `unaffected consumers` to require a full version spec? In case the consumer has an update and starts...

This has come up a few times in [`clap`](https://github.com/kbknapp/clap-rs) with an open issue being kbknapp/clap-rs#965

@hoodie I agree the specific test referenced in clap#965 doesn't quite line up with this issue, however just today I ran into an issue where the only change made was...