Any suggestions to remove the CR's after the value $one?

Looks like that's been covered, but I have some other suggestions, if you're open to receiving them.

1. Use lexical variables.
2. Stick $! somewhere in your error messages for a failed concat().
3. Iterate over the data lines using natural Perl idiom, instead of C idiom.
4. Don't backwhack things that don't need it.

Now, beyond that, it seems to me that you're doing the same operation in more than one place -- namely, filtering and reformatting valid data lines. This makes it a good candidate for a subroutine.

    sub valid_lines
    {
        map
        { 
            my( $two, $one ) = split /,/;
            $one =~ /@.*\./ ? ( "$two,$one" ) : ()
        }
        @_
    }


    open FILE, "< $datadir/$editlist.txt" or
        error_message( "Can't read data file $datadir/$editlist.txt - 
+$!" );
    my @lines = valid_lines( <FILE> );
    close FILE;
    $ccc = @lines;

    local $, = "\n";
    print <<EOF;
<textarea cols="60" rows="5" name="list">
@lines
</textarea><br/>
EOF

    # Then the edited code is written to the db:

    sub change_list
    {
        my $editlist = $FORM{'editlist'};

        $FORM{'list'} =~ s/\r//g; # kill useless carriage returns.

        open FILE, "> $datadir/$editlist.txt" or
            error_message( "Can't write data file $datadir/$editlist.t
+xt - $!" );

        for ( valid_lines( split /\n/, $FORM{'list'} ) )
        {
            print FILE "$_,x\n";
        }
        close FILE;
[download]

Just an idea.

jdporter
...porque es dificil estar guapo y blanco.