I think the idea is not bad, but the JAPH phrase is a bit long for that, so it gets boring quickly. 
sub e{sub _{-72} er}eval($/='sub J{map{substr$/,ref('.q:)?pop@$_\:$_,$
+$_[0]||1}
@_}sub f{uc join "system",J(@_)}print J 4=>1,[2,13],-1,010-1,[ _ ],42-
+020,hex"e
";print+h,e,$",f(__LINE__*2+2),e,qw,l,,J(-1,-hex 20),q{ack},e,J hex or
+d(0)/4 :)

[download]
A variant that is a bit profane: 
sub k{join shift,@_}sub e{sub _{-60}er}eval($/='sub J{map{substr$/,ref
+()?pop'.q
:@$_\:$_,$$_[0]||1}@_}sub f{uc k"system",J(@_)}print J 4,1,[2,13],-1,0
+10-1,[_],
42 -020,hex e;print+h,e,$",f(__LINE__+6),e,qw,l,,J(-1,-hex 16),q{ack},
+e,J 18 :)

[download]
لսႽ† ᥲᥒ⚪⟊Ⴙᘓᖇ Ꮅᘓᖇ⎱ Ⴙᥲ𝇋ƙᘓᖇ

Replies are listed 'Best First'.
Re: JAPH with a smiley
by choroba (Cardinal) on Sep 04, 2013 at 11:37 UTC
    For those who cannot decipher it themselves:
    The main idea is to build the output string from the source itself. The main function J just returns characters at given positions (it can return more than one). There are several ways how to get the positions, including octal literals, functions returning numbers, or some other tricks. Note that 42 is present in both versions, the place that generates "h" generates it from itself, ack is mentioned, and system does not do anything bad here. I also like hex 16 in the second version, it might be fun to include oct 8 as well.
    لսႽ† ᥲᥒ⚪⟊Ⴙᘓᖇ Ꮅᘓᖇ⎱ Ⴙᥲ𝇋ƙᘓᖇ
[reply]
[d/l]
[select]

Back to Obfuscated Code