And for those of you who ran this, but were too lazy to figure out what it was doing, and those who didn't run it, but are marvelling at its obfuscation anyway, here is an equivalent version:
foreach $_ (split / /, "'s q,s q,q 3g sg 'g {i {) }{ sg 3g 7i i' q,g "
                       . "qq} qq) ii i} qq{ q,q q,, 3g ', q,q qqs q,' "
                       . ord "\n")
{
    $_ =~ s/\)/5/i;
    $_ =~ s/\'/8/i;
    $_ =~ s/q/1/g;
    $_ =~ s/s/4/g;
    $_ =~ s/,/0/g;
    $_ =~ s/{/6/g;
    $_ =~ s/g/2/g;
    $_ =~ s/i/9/g;
    $_ =~ s/}/7/gx;
    print chr($_);
}

The "cleanups" I made:

  • removed the q{string} and qq{string} syntax, and replaced it with more normal "string" syntax.
  • changed the pattern being used to split from ";" to " ", which is more readable.
  • got rid of all implicit arguments
  • changed the substitutions from the forms;;; to s///. (Note that most of the gs and all of the is in the substitutions are not technically necessary. In addition, the gx on the last substitution is extraneous; it could be just g, or even left off entirely.)
  • added newlines and indentation
  • broke the main string into two pieces, so that it wouldn't screw up E2's width

So what the program does is this:

  1. starts with a string that is a list of ASCII values for the string "The *REAL* Obfuscated Perl\n", where each digit of each ASCII value has been replaced by a letter or symbol. (The first symbol, "'s", stands for "84", which is the ASCII value for "T".)
  2. splits this string into the list of encoded ASCII values
  3. decodes each value, by substituting the appropriate digit for each letter/symbol
  4. prints out the character represented by the decoded ASCII value

*sigh*.

I suppose now we will all have to wait for The *REAL* *REAL* Obfuscated Perl... ("E2 source code, what?")