Formalizing the purpose of Arabica

[evan-forbes]

The purpose of the mocha series of testnets has been relatively clear, however the purpose of the arabica series of networks has changed over time and sometimes in an adhoc way. From my understanding, the arabica's are supposed to be a bleeding edge version of core/node, but they've rarely been used in that way. Instead, they've been used as purpose built networks for rollup testnets or backup networks incase mocha goes down. More importantly I don't think we have communicated the purposes of these testnets to users and builders.

It could be useful to formalize (and then document) its purpose now mainnet is live. @cmwaters and I have discussed having independent arabica's per breaking set of features. The reason being that if there are features that users need to start building around, they could get a head start by using that network in an isolated way. Then we can simply shutdown the old ones as needed. This also means that the arabica's would likely not use the normal upgrade mechanism, although they technically could