cgewecke

Ok, that's helpful. Now we have four Ubuntu / Node configurations and three occurrences of the error. | Ubuntu | Node | ok | who | |----|----|----|----| |18.04|14.15.1| ❌ |...

@stevieraykatz If you run `npm list eth-gas-reporter` in your project do you see more than one version?

@pascal1990 To be clear, are you reporting that the box works with `ganache-cli` but not with Geth?

@dapplion It's a bug coming from solidity-parser-antlr which should be fixed in solidity-coverage `0.6.5`. Maybe a lower version is cached here?

@area I feel like this same-line `else` construction is relatively common and Istanbul's report is intuitively more informative. But as you note in #37 - the obvious fix would produce...

Sorry for the delay debugging this. You're seeing zero-coverage because: + solidity-coverage modifies your contracts (injecting trace instrumentation) and recompiles them, saving the result to artifacts + you consume your...

Oh! Just remembered, there's (at least) one more potential coverage issue. SC doesn't handle `hardhat_reset` well, because that method completely re-instantiates the provider. https://github.com/thehubbleproject/hubble-contracts/blob/7d28e78ae646c7a176fa4e8c543d3f20966b36f0/test/fast/frontends.test.ts#L28 It should be possible to achieve...

@i-stam Hi. Out of curiosity, did the codechecks report in CI also stop working?

Leaving some additional info about how performance impact might be evaluated... Have recently benchmarked zeppelinsolidity-contracts tests and measured the number of times they fire the step: | # Tests |...

Thanks for that explanation @davidmurdoch, much appreciated!