Consider switching to substrate-lite for client functionality

[insipx]

Substrate-Archive uses a 'custom' client implementation based on implementing all the traits from Substrate to provide functionality required for collecting blocks/metadata/etc. A new library substrate-lite might be able to do all this but wrapped in a better more modular API, resulting in cleaner / more performant code all around. However this requires more investigation/research and testing

[emostov]

Some questions I had:

• Do we know the long term maintenance plans for substrate-lite? If maintenance stops at some point would we still be able to use it?
• Is substrate-lite compatible with any valid substrate runtime?

[insipx]

Good questions! Not sure if I know the correct answer. I think the ultimate answers to these questions will come with time, though.

Some questions I had:

* Do we know the long term maintenance plans for substrate-lite? If maintenance stops at some point would we still be able to use it?

Don't know/don't think there is. I think if substrate-lite becomes does not have a maintenance plan i'd rather stick with the substrate "official" client. However I think if there is a significant enough userbase substrate-lite will be maintained in one way or the other.

* Is substrate-lite compatible with _any_ valid substrate runtime?

I believe so. Especially since substrate-lite is using only wasm execution, then it probably accepts a Runtime generic not unlike how the Archive Client currently works.