Andreas Abel

Revision landed: https://hackage.haskell.org/package/ekg-core-0.1.1.7/revisions/

To get things rolling, I revised the bounds on hackage: https://hackage.haskell.org/package/ekg-core-0.1.1.7/revisions/

I revised the bounds on hackage: https://hackage.haskell.org/package/ekg-core-0.1.1.7/revisions/

Would it be possible to supply a MWE that builds out-of-the-box and demonstrates the problem?

Over to @tibbe and @hvr ...

Sorry for the slow reaction. I rebased this onto `master` to get CI to run. The testsuite has (trivial) problems with the legacy GHCs (< 7.10). ``` tests/UnitTests.hs:485:30: Not in...

> Adding 'pure' to imports fixed the problem I moved this to the conditional import of `Control.Applicative` to avoid the _unused import_ warning on new GHCs.

I see your parser supports untagged unions via ``. Backtracking can be costly, so I do not know if this is in the spirit of this package: https://github.com/haskell-hvr/cassava/blob/fd168df89805fb0f66d9a8aab92abeb157a300c3/README.md?plain=1#L78-L86 I added...

I agree to your argumentation. Banning untagged unions altogether just because there are problematic cases would throw out the baby with the bathwater. So, let's include them.

@qnikst > I'll investigate and see if builders in the `bytestring` should be improved. What was the conclusion of this investigation? I think it would make sense to add QuickCheck...