At 19.34 -0500 1999.01.27, Vicki Brown wrote: >You cheated. You tested with /a/ (aka m/a/). >We have been discussion the return value of the substitution command s/a/. They are the same. >Anyway, the camel sez that s/... > returns the number of substitutions made (on success) > returns false (0) otherwise Where does it say 0? In all my tests, m/a/ and s/a// returned the exact same thing for false. And using -Dctls as Ronald did seems to support that, if i am reading it right. >Considering that false is false but "" is not precisely equivalent to 0 >we can't use the result from a match (m/.../) to experiment with how >substitution works (s/.../). I think you can in this case. Substitute and see for yourself, though. Regardless, we are still left with the fact that m/a/ or s/a// returns some null value when false, that Ronald identified as 0 in numeric context and '' in string context, but nevertheless will use 0 when context is ambiguous, supposedly, though I am not sure yet. -- Chris Nandor mailto:pudge@pobox.com http://pudge.net/ %PGPKey = ('B76E72AD', [1024, '0824090B CE73CA10 1FF77F13 8180B6B6'])