diff --git a/examples/negation-character-classes.js b/examples/negation-character-classes.js
new file mode 100644
index 00000000..8d88079a
--- /dev/null
+++ b/examples/negation-character-classes.js
@@ -0,0 +1,21 @@
+var mm = require('..');
+
+console.log(mm(['bar/bar'], ['foo/**', '!foo/baz']));
+console.log(mm(['bar/bar'], ['!foo/baz', 'foo/**']));
+console.log(mm(['bar/bar'], ['!**', '!foo/baz', 'foo/**']));
+console.log(mm(['bar/bar'], ['**', '!foo/baz', 'foo/**']));
+console.log(mm(['bar/bar'], ['!foo/baz', 'foo/**']));
+
+// 1. No error, but pattern has a different behavior and my tests are failing because of this: https://ci.appveyor.com/project/electerious/rosid/build/1.0.72/job/srfjb4x7p6t1l5r5
+//'*/[^_]*.{html,ejs}*'
+//'a/[^_]*.{html,ejs}*'
+// '!_*.{html,ejs}'
+
+// 2. Throws no parsers registered for: "]*.{h".
+// [^_]*.{html,ejs}
+
+// 3. Works on macOS, but fails on windows because of a different behavior: https://ci.appveyor.com/project/electerious/rosid/build/1.0.74/job/17hbg0e822r8pdw3 => I guess this shouldn't be the case. Is [^_] evil?
+// '[^_]*.html + [^_]*.ejs'
+
+// 4. Could be a workaround, but !_*.{html,ejs} seems to be different to what [^_]*.{html,ejs}* does.
+// '!_*.{html,ejs}'