Can you explain why it probably doesn’t work with 1..0?

Actually, it works without warnings for 1..0 or any other unequal pair of numbers (actually, it doesn't work at all per this, but at least it not-works consistently | actually, it isn't that consistent (update: see this)):

c:\@Work\Perl\monks>perl -wMstrict -le
"sub get_max_index {
   my $imax = 0;
   ;;
   foreach (@_) {
     $imax = (1 .. 0) if $_ > $_[$imax];
     }
   return $imax;
   }
 ;;
 $. = 1;
 ;;
 my @arr = (1,2,13,4,5);
 my $ans = get_max_index(@arr);
 print qq{i max == $ans;  \@arr[$ans] == $arr[$ans]};
"
i max == 2;  @arr[2] == 13
[download]

(Update: Try this with  my @arr = (1,2,13,4,5, 99); as the test input to see it not-work.)

Note the  $. = 1; statement. Per Range Operators in perlop:

If either operand of scalar ".." is a constant expression, that operand is considered true if it is equal ("==") to the current input line number (the $. variable).

So the following flip-flop values and  $. initializations "work":

• $imax = (  1 ..   0) if $_ > $_[$imax]; for  $. = 1
• $imax = (  0 ..   1) if $_ > $_[$imax]; for  $. = 0
• $imax = ( 42 .. 137) if $_ > $_[$imax]; for  $. = 42
• $imax = (137 ..  42) if $_ > $_[$imax]; for  $. = 137

And yes, this sort of thing should definitely appear in some sort of index of heretical teachings, somewhere.