Hi, as hippo pointed out, this is not an SSCCE, and you are declaring $response twice. Additionally:

• To be clear, you do *not* need to create a request object manually if you use HTTP::Tiny (nor LWP::UserAgent for that matter)
  
• While we don't know how you have stored a URL in $html, it's likely not a good name for a variable that does not contain HTML.
  
• You probably don't want to print the error as HTML if you are not returning output to a browser
  

You need to stop throwing in snippets from various sources and settle on one tool, and take the time to learn it. That involves spending time with the documentation. The example I originally showed was copied almost verbatim from the HTTP::Tiny documentation. I suggest that you start with the SYNOPSIS of the module that you choose, and once you have that running in your environment, you can make changes to it, one line at a time, so you can see immediately if you made a mistake.

This is an SSCCE that should be copied into a file (eg 'foo.pl') and run with $ perl foo.pl http://example.com

use strict;
use warnings;
use HTTP::Tiny;

my $url = shift;
my $response = HTTP::Tiny->new->get($url);

if ( $response->{success} ) {
     print "OK: $response->{content}\n";
} else {
     print "Failed: $response->{status} $response->{reason}\n";
}

__END__
[download]

... or this ...

use strict;
use warnings;
use LWP::UserAgent;

my $ua = new LWP::UserAgent;
# possibly add advanced user agent options here

my $url = shift;
# possibly construct a request object here

my $response = $ua->get($url); # if not using a request object

if ($response->is_success) {
    print $response->decoded_content, "\n";
} else {
    die $response->status_line;
}

__END__
[download]

Hope this helps! There are no shortcuts to learning, not if you want to understand.