Well, even though I don't like "obfuscation" schemes that include calls to crypt, unpack or chr, I have to admit I don't know the ASCII table by heart. :-)

So, I noticed all those 44s, decided I could at least change the eval for a print, and see what it actually did (mind you, I didn't guess... but the way to do it is still, er, poor).

If the main point in an obfuscation is based on s///, crypt, or chr to make it hard to understand, you can be sure I won't like it. Try to do the same thing in a completely unnatural and complicated way, and there you will have me try to understand what it does, parsing it by hand, and so on...