10.2.1 release preparation

[neilmayhew]

Update CHaP index and cardano-node version

[mgmeier]

Maybe some context to making ForwardingVersionData an extra argument:
Its definition is just a wrapper newtype ForwardingVersionData = ForwardingVersionData{networkMagic :: NetworkMagic}. This means that no function who has the networkId in scope will ever need ForwardingVersionData as a seperate argument.

EDIT: If this is motivated by any change in simpleSingletonVersions - would it be possible to change such that we see
[\_ ->] SomeResponderApplication $ const $ forwarderApp [...]? Which would describe IMHO much better what's actually going on?

[carbolymer]

👍🏻 on chairman, submit-api and testnet

[neilmayhew]

@crocodile-dentist @coot:

The review feedback here is on changes that came from your PRs (#6107 #6114) that were targeting release/10.2.x. I merged those and cherry-picked them onto this branch because we wanted to release from master and it's not possible to rebase a release branch because of branch protections.

Feel free to push changes to this branch, or to create a new PR targeting master for your previous PRs. I'll then rebase this branch onto master after they're merged.

[coot]

@mgmeier ouroboros-network-framework's simpleSingletonsVersions function takes now a callback (the vdata -> r in the link above) with an argument - the negotiated version data. This allows to customise the application based on negotiated data.

[mgmeier]

@coot Thanks for confirming this is indeed motivated by simpleSingletonVersions. As I pointed out, the forwarding application does not require the extra complexity of making behaviour dependent on negotiated version data. Moreover, the way version data definition is currently implemented, it's conceptually impossible to do so - which makes sense, at this has not been an option in the past.

I'd like the code to be very discriptive and reflect that situation, ideally also in its type signatures. I'll add a commit to make the forwarderApp a constant function.

[mgmeier]

Adressed my concerns in an additional commit.