Should we do database split from zebra state and zebra-db as pat of the zebra-scan MVP ?

[oxarbitrage]

We want to store persistent data (keys and results) as part of the MVP but some of us are unsure if the best way to go first is to do #7926 and #7937 or to just create a database duplicating code as needed inside zebra-scan leaving zebra-state untouched and no new crate creation.

[oxarbitrage]

As far as i know you were already working on splitting the zebra-state code to a new crate, How far are we from that ? If we are close already we might want to do that first.

[teor2345]

I have some draft code based on this suggestion here:
#7975 (reply in thread)

We don't need to split the state straight away, we can give external crates access to the raw database type instead. (DiskDb, or maybe ZebraDb, I'm still working that out.)

Then after we know which parts of the database code we want to use, we can split the state later.

[mpguerra]

if we don't need to split state and it's quicker to give access to the raw database, let's do that. Do we need to create a new issue for this?

[teor2345]

I made a new ticket here: #8001

[oxarbitrage]

Lets close this one, thank you @teor2345