@@ -50,11 +50,22 @@ In the test file, we first define the tests: the faulty server should fail and t
5050```haskell
5151spec = describe "example" $ do
5252 it "good server should not fail" $ do
53− fuzz @Api goodServer defaultConfig { coverageThreshold = 0.99 }
53+ fuzz @Api goodServer config
5454 >>= (`shouldSatisfy` isNothing)
5555 it "bad server should fail" $ do
56− fuzz @Api badServer defaultConfig { coverageThreshold = 0.99 }
56+ fuzz @Api badServer config
5757 >>= (`shouldSatisfy` serverFailure)
58+
59+config = defaultConfig
60+ {
61+ -- we expect to be able to cover the api from our starting point:
62+ -- this will fail the test if we don't.
63+ coverageThreshold = 0.99
64+ }
65+
66+-- there are other tweakable things in the config, like maximum runtime, reps,
67+-- per-request healthchecks, and verbose logging. Have a look at
68+-- Roboservant.Types.Config for details.
5869```
5970
6071And unless we want to ship roboservant and all its dependencies to production, we also need