Make hydra-cluster --devnet configurable

By default it will now open a head and wait again and only when given
`--busy`, the head is busy respending the same head.

This allows to verify interactions with a open head interactively
quickly by just doing:

```
cabal build hydra-cluster
hydra_cluster_datadir=hydra-cluster cabal exec hydra-cluster -- --devnet --publish-hydra-scripts
```

This is particularly handy when writing the user manual.

CHANGELOG.md

@@ -32,6 +32,9 @@ changes.
   `HandshakeFailure` event will be recorded in the logs and sent as a server
   output on the API.
 
+- Make `hydra-cluster --devnet` more configurable
+  - Now it is idle by default again and a `--busy` will make it busy respending the same UTxO.
+
 ## [0.16.0] - 2024-04-03
 
 - Tested with `cardano-node 8.9.0`, `cardano-cli 8.20.3.0` and `mithril 2412.0`.

hydra-cluster/exe/hydra-cluster/Main.hs

@@ -6,7 +6,7 @@ import CardanoNode (findRunningCardanoNode, waitForFullySynchronized, withCardan
 import Hydra.Cluster.Faucet (publishHydraScriptsAs)
 import Hydra.Cluster.Fixture (Actor (Faucet))
 import Hydra.Cluster.Mithril (downloadLatestSnapshotTo)
-import Hydra.Cluster.Options (Options (..), PublishOrReuse (Publish, Reuse), UseMithril (UseMithril), parseOptions)
+import Hydra.Cluster.Options (Options (..), PublishOrReuse (Publish, Reuse), Scenario (..), UseMithril (UseMithril), parseOptions)
 import Hydra.Cluster.Scenarios (EndToEndLog (..), respendUTxO, singlePartyHeadFullLifeCycle, singlePartyOpenAHead)
 import Hydra.Logging (Verbosity (Verbose), traceWith, withTracer)
 import Options.Applicative (ParserInfo, execParser, fullDesc, header, helper, info, progDesc)
@@ -34,11 +34,14 @@ run options =
           withCardanoNodeDevnet fromCardanoNode workDir $ \node -> do
             txId <- publishOrReuseHydraScripts tracer node
             singlePartyOpenAHead tracer workDir node txId $ \client walletSk -> do
-              -- Start respending the same UTxO with a 100ms delay.
-              -- XXX: Should make this configurable
-              respendUTxO client walletSk 0.1
+              case scenario of
+                Idle -> forever $ pure ()
+                RespendUTxO -> do
+                  -- Start respending the same UTxO with a 100ms delay.
+                  -- XXX: Should make this configurable
+                  respendUTxO client walletSk 0.1
  where
-  Options{knownNetwork, stateDirectory, publishHydraScripts, useMithril} = options
+  Options{knownNetwork, stateDirectory, publishHydraScripts, useMithril, scenario} = options
 
   withRunningCardanoNode tracer workDir network action =
     findRunningCardanoNode (contramap FromCardanoNode tracer) workDir network >>= \case

hydra-cluster/src/Hydra/Cluster/Options.hs

@@ -12,6 +12,7 @@ data Options = Options
   , stateDirectory :: Maybe FilePath
   , publishHydraScripts :: PublishOrReuse
   , useMithril :: UseMithril
+  , scenario :: Scenario
   }
   deriving stock (Show, Eq, Generic)
   deriving anyclass (ToJSON, FromJSON)
@@ -24,13 +25,18 @@ data UseMithril = NotUseMithril | UseMithril
   deriving stock (Show, Eq, Generic)
   deriving anyclass (ToJSON, FromJSON)
 
+data Scenario = Idle | RespendUTxO
+  deriving stock (Show, Eq, Generic)
+  deriving anyclass (ToJSON, FromJSON)
+
 parseOptions :: Parser Options
 parseOptions =
   Options
     <$> parseKnownNetwork
     <*> parseStateDirectory
     <*> parsePublishHydraScripts
     <*> parseUseMithril
+    <*> parseScenario
  where
   parseKnownNetwork =
     flag' (Just Preview) (long "preview" <> help "The preview testnet")
@@ -86,3 +92,11 @@ parseOptions =
             \If not set, the cardano-node will synchronize the network given the current \
             \cardano-node state in --state-directory."
       )
+
+  parseScenario =
+    flag
+      Idle
+      RespendUTxO
+      ( long "busy"
+          <> help "Start respending the same UTxO with a 100ms delay (only for devnet)."
+      )