Hey All,<div><br></div><div>When playing with the current hackage versions of Epic and Idris to make them play nice with ghc7.6rc1</div><div><br></div><div><a href="http://hackage.haskell.org/package/idris-0.9.2.1">http://hackage.haskell.org/package/idris-0.9.2.1</a></div>

<div>and </div><div><a href="http://hackage.haskell.org/package/epic-0.9.3">http://hackage.haskell.org/package/epic-0.9.3</a>    (current version on github now builds on ghc 7.6, <a href="https://github.com/edwinb/EpiVM">https://github.com/edwinb/EpiVM</a>)</div>

<div><br></div><div>I ran into some funny type inference problems. Namely, using the idris-0.9.2.1  source and iteratively seeing how ghc complains, </div><div>I repeated found that ghc would infer extraneous class constraints with variables that don&#39;t appear in the function type!</div>

<div><br></div><div>eg <span style="background-color:rgb(255,255,255);font-family:&#39;Bitstream Vera Sans Mono&#39;,Courier,monospace;font-size:12px;line-height:16.78333282470703px;white-space:pre">(Num a, Ord a) =&gt; PArg -&gt; Doc</span>, when the <b>correct</b> type to infer would  be <span style="background-color:rgb(255,255,255);font-family:&#39;Bitstream Vera Sans Mono&#39;,Courier,monospace;font-size:12px;line-height:16.78333282470703px;white-space:pre">PArg -&gt; Doc.</span></div>

heres some gists with links to more info<div><span style="background-color:rgb(255,255,255);font-size:12px;line-height:16.78333282470703px;white-space:pre"><font face="Bitstream Vera Sans Mono, Courier, monospace"><a href="https://gist.github.com/3365312">https://gist.github.com/3365312</a></font></span></div>

<div><span style="background-color:rgb(255,255,255);font-size:12px;line-height:16.78333282470703px;white-space:pre"><font face="Bitstream Vera Sans Mono, Courier, monospace"><a href="https://gist.github.com/3365073">https://gist.github.com/3365073</a></font></span></div>

<div><span style="background-color:rgb(255,255,255);font-size:12px;line-height:16.78333282470703px;white-space:pre"><font face="Bitstream Vera Sans Mono, Courier, monospace"><a href="https://gist.github.com/3364775">https://gist.github.com/3364775</a></font></span></div>

<div><span style="background-color:rgb(255,255,255);font-size:12px;line-height:16.78333282470703px;white-space:pre"><font face="Bitstream Vera Sans Mono, Courier, monospace"><br></font></span></div>Anyways, I&#39;m not sure what to make of this, is this a reasonable artifact of  type inference getting confused on functions with a large number of case analyses when various typeclass extensions are enabled? Or  Is this a bug in terms of what inference should be able to handle?<div>

<br></div><div>Just to be clear, when I add the infererred type ascriptions without the type class constraint, everything type checks in those modules. So my confusion is why the inference adding those unused class constraint variables! </div>

<div><br></div><div>thanks all,</div><div>-Carter</div>