I have a thought about new API, autonomi
.
It’s a new layer put on top of the old API, sn_client
, instead of replacing it. It is leaving unnecessary code, while adding new, and more code is more possibilities of bugs. Old code is still not refactored, which was agreed is needed. So, we have two APIs now to be maintained, and one of them is a mess.
Moreover, tools like CLI, Auditor, Faucet etc. still depend on sn_client
, which means there is no official app based on autonomi
, so it’s still untested by the team on a real app. I think it would be a good move to base official tools on a new API as a proof of it’s stability and that it’s ready for developers` use.
In its current form, it’s an architectural and organizational error for me. I’ll probably stick to the sn_client
API until official tools start to use the new API.