Fix an ambiguity on julia 0.6

[timholy]

Sometimes, it's the simplest things that fail. Detected in a test failure for ImageFiltering, https://travis-ci.org/JuliaImages/ImageFiltering.jl/jobs/200619941.

[codecov-io]

Codecov Report

Merging #74 into master will increase coverage by 0.45%.

@@            Coverage Diff             @@
##           master      #74      +/-   ##
==========================================
+ Coverage   82.96%   83.42%   +0.45%     
==========================================
  Files           4        4              
  Lines         182      181       -1     
==========================================
  Hits          151      151              
+ Misses         31       30       -1

Impacted Files	Coverage Δ
src/normed.jl	90.8% <100%> (+1.03%)	✅

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9a1b40a...3bf4701. Read the comment docs.

[vchuravy]

We have the ambiguity tests currently turned off for v0.6 we should probably turn them on again, (there might have been a Base issue though)

[timholy]

Agreed. Currently the problem is that Base reports a lot of ambiguities right now. Some of them are fictitious (JuliaLang/julia#20246, JuliaLang/julia#20567), others may be real.

I've sometimes used setdiff to "subtract off" ambiguities in Base from those in other modules. But when you're extending a Base method, that won't help.

[timholy]

Should have mentioned, the second of those bug reports was motivated by trying to turn on the ambiguity tests here. 😄

[timholy]

We'll want to tag again soon, but ideally we'd incorporate #73.