I was not aware of Tiny::Config::Ordered until today. However, the Hash::Ordered module v0.009 is quite nice IMO.

The following diff is all that's needed to preserve order transparently in Tiny::Config.

$ diff -u Tiny.pm.orig Tiny.pm
--- Tiny.pm.orig    2015-07-06 12:51:54.000000000 -0400
+++ Tiny.pm    2015-07-06 13:02:22.000000000 -0400
@@ -3,6 +3,7 @@
 # If you thought Config::Simple was small...

 use strict;
+use Hash::Ordered;

 # Warning: There is another version line, in t/02.main.t.

@@ -39,7 +40,8 @@
 # Create an object from a string
 sub read_string {
     my $class = ref $_[0] ? ref shift : shift;
-    my $self  = bless {}, $class;
+    tie my %obj, 'Hash::Ordered';
+    my $self  = bless \%obj, $class;
     return undef unless defined $_[0];

     # Parse the file
@@ -59,7 +61,10 @@
             # Create the sub-hash if it doesn't exist.
             # Without this sections without keys will not
             # appear at all in the completed struct.
-            $self->{$ns = $1} ||= {};
+            $self->{$ns = $1} ||= do{
+                tie my %obj, 'Hash::Ordered';
+                \%obj;
+            };
             next;
         }
[download]

The output order matches the ini file (sections and elements).

$ perl test.pl
[insert]
SQL1 = insert into table <table1>
SQL2 = insert into table <table2>
SQL3 = insert into table <table3>
SQL4 = insert into table <table4>
[Truncate]
SQL1 = truncate table <tablename1>
SQL2 = truncate table <tablename2>
[update]
SQL1 = update into <table1>
SQL2 = update into <table2>
[download]

I sent the link to the first post to both authors.

Kind regards, Mario