Note: This posting is chock full of partly-baked ideas, so YMMV…
The basic notion is to create a LiveBook that provides a set of interactive dashboards for an application suite’s activity. Existing tools such as Observer could be folded in, perhaps with extensions and/or modifications to their presentation.
Once the basic framework for this is in place, it might be useful to extend it in various directions. For example, it could capture runtime information for diagnostic purposes. This data could assist in the generation of typespecs for dialyzer, etc.
Is anything (even vaguely) like this in prospect?
It would be hard to justify creating this sort of tooling for a single use, so my notion is to create a basic LiveBook that could be “hooked up” to any existing application suite.
Customizations could then be added to handle specialized needs. Ideally, at least some of these could be written in a general and/or configurable manner, to promote sharing.
@Adzz has been experimenting in a related area (loading Livebook content that he generated from Hexdocs pages). He discovered significant latency and resource issues with larger items, but I’m pretty sure that these will be resolved in time.
So, I posted a speculative reply, addressing the question “How should a communal project Livebook for Elixir developers be structured and implemented?”. As usual, constructive comments and feedback are welcome…