I am having a strange problem after upgrading to jruby 1.7.2. Trying to
get the specs to pass, but they’re failing on some utf-8 file
reading/writing. The data is a valid_encoding as a string, but when I
write it to a file and then read it back it is no longer a
valid_encoding. The same code works fine on a mac, and it worked fine on
windows before on jruby 126.96.36.199.
I tried to reproduce with a trivial case but was not able to. Any ideas
what might be going on here?