Please don't take me the wrong way. I am excited to see where this is going.
I am particularly curious about how chatty you can get with these agents before latency suffers. I am also curious to know how much memory allocation is going on behind the scenes to services the agents.
If I was overly critical in my first comment, it comes from prior experience where the examples made a language/tool look really simple. Then as soon as you apply it to a real world problem, the kludge, scaffolding, and other gotchas show their head. Haskell
being my most recent disappointment.
I don't mean to suggest that this will be the case with Axum. I am simply curious to see something less trivial.
I don't have the brainpower to know who's right on the static typing debate, but as someone who's written a great deal of code in static and dynamic languages for production use, I was shocked at the suggestion that static typing does not aid reliability.
To my thinking, reliability based on code that will error at compile time is far more desirable than code that depends on developer discipline. Not because it makes coding easier, but because it frees my limited mind to be creative about far more interesting