Neither problem is a problem. On Windows your version of Perl should read by default as a text file and it will be fine. In Unix binary and text works fine. The only issue is if you read a DOS file in Unix, in which case I would trust no script not specifically designed to do that. (My production ones are though.) As for "for" and "for" working, the spec said find anagrams. Not find unique anagrams. The program works according to what my co-worker specified.

As for the input, output issue, mea culpa.