Publish zebra crates to crates.io

[mpguerra]

Motivation

Once the code is audited and we do our first stable release, we will want to publish zebra to crates.io.

Instructions

https://doc.rust-lang.org/cargo/reference/publishing.html#before-publishing-a-new-crate

In particular, we'll want to include the Zcash Foundation in the crate description and homepage, and "Zcash" in the crate keywords.

Crates that need to be published

• tower-batch-control (was renamed in Rename tower-batch crate #6202)
• tower-fallback
• zebra-chain
• zebra-consensus
• zebra-network
• zebra-node-services
• zebra-rpc
• zebra-script
• zebra-state
• zebrad

Testing and developer tools:

• zebra-utils
• zebra-test

Reserved Crate Names

We have already reserved these crate names on crates.io.

We can, to just get the crate names, do local renames of zebra-cli to each one, publish, and so forth, for now.

These placeholders are now co-owned by @ZcashFoundation/owners (https://crates.io/teams/github:zcashfoundation:owners?sort=new):

• tower-batch-cpu (https://crates.io/crates/tower-batch-cpu)
• tower-fallback
• zebra-chain
• zebra-cli
• zebra-consensus
• zebra-network
• zebra-node-services
• zebra-rpc
• zebra-script
• zebra-state
• zebra-test
• zebra-utils
• zebrad (and https://crates.io/crates/zebra)

[mpguerra]

Do we want to squat and publish the other crates, e.g zebra-network, etc... to crates.io?

[teor2345]

Do we want to squat and publish the other crates, e.g zebra-network, etc... to crates.io?

We have to publish all our production crates to crates.io, because it won't accept a crate unless its dependencies are also published:

• We will need to rename tower-batch, because it already exists, and does something different (file writing batches, not cryptography batches)

I don't mind if we squat or publish our test crates, it's probably easier just to publish them.

zebra-client is currently a placeholder, but we should add a README saying that.

[teor2345]

Do we also want to publish a placeholder for zebra-cli? Or are we going to use zebrad as the zebra-cli instead?

(It seems simpler, but if we publish a placeholder, we can leave that decision for later.)

[teor2345]

We might not be able to publish zebra-test due to the 10 MB crates.io size limit (our test data is close to that size):
https://doc.rust-lang.org/cargo/reference/publishing.html#packaging-a-crate

If we can't, we should still publish a placeholder.

[mpguerra]

@dconnolly has been working on this. I'm hoping we can finish this off during next sprint (2023 Sprint 8)

[teor2345]

@mpguerra should this be blocked by renaming tower-batch in #6202?

[mpguerra]

Now that we have the placeholders we can wait to publish the actual crates until we tag the stable release

[mpguerra]

Hey team! Please add your planning poker estimate with Zenhub @arya2 @conradoplg @dconnolly @oxarbitrage @teor2345 @upbqdn

[teor2345]

@dconnolly I split up the list of crates into crates that need to be published for the zebrad release, and everything else.

I think for our first step we might just want to publish the zebrad crates.

[teor2345]

I removed the empty crates from the list because PR #6726 deletes them.