Fix issue #216

[zarthross]

Removing the Lazy on the ResultMatcher (which is only barely recursive on situations of F[ResultMatcher] seems to fix the issue in #216 with circe Encoder causing a compilation issue.

Reproduction requires this dependency:
"io.circe" %% "circe-core" % "0.10.0"

Minimum reproduction.

import cats.effect.Effect
import io.circe.Encoder
import org.http4s._
import org.http4s.rho.RhoRoutes

trait Auto

class Test[F[+_]: Effect] extends RhoRoutes[F] {
  implicit def autoEncoder[A <: Auto : Encoder]: EntityEncoder[F, A] = ???

  GET / "get" |>> ({ () => // Didn't work.  Used Lazy[ResultMatcher]
    Ok(())
  })

  GET / "get2" |>> ({ // Worked.  Does not use Lazy[ResultMatcher]
    Ok(())
  })
}

I'm still wondering how I can make a test, I haven't isolated what it is about the circe Encoder that causes the root issue.

[zarthross]

Lazy was introduced here #93 and here #110, so summoning the original reviewer @bryce-anderson and author @shengc for comments.

[danxmoran]

@zarthross is there a snapshot artifact of this branch which I could pull into my build?

[zarthross]

No, but I'm going to go ahead and release this as part of the next milestone, which will be either tonight or tomorrow.

…

On Wed, Feb 20, 2019, 12:39 PM Daniel Moran ***@***.***> wrote: @zarthross <https://github.com/zarthross> is there a snapshot artifact of this branch which I could pull into my build? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#285 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABD2KJeTEziXi_UE76cpIPiutNX63rEsks5vPZZngaJpZM4aJwSU> .