From: Matt S Trout <mst@shadowcat.co.uk>
Date: Sat, 21 Jul 2012 14:42:32 +0000 (+0000)
Subject: document Moo versus Any::Moose in brief with article link
X-Git-Tag: v1.000001~5
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=c687d0168a34cd04fa5e4dbda4c0f390bb8c4abc;p=gitmo%2FMoo.git

document Moo versus Any::Moose in brief with article link
---

diff --git a/Changes b/Changes
index 87d9283..57e5ebd 100644
--- a/Changes
+++ b/Changes
@@ -1,3 +1,4 @@
+  - document Moo versus Any::Moose in brief with article link
   - remove quote_sub from SYNOPSIS and has docs, expand Sub::Quote section
   - localize @_ when inlining quote_sub'ed isa checks (fixes lazy+isa+default)
   - ensure constructor gets regenerated if forced early by metaclass inflation
diff --git a/lib/Moo.pm b/lib/Moo.pm
index 9ccab7f..2fd9419 100644
--- a/lib/Moo.pm
+++ b/lib/Moo.pm
@@ -298,6 +298,25 @@ to your code before Moose is loaded, but bear in mind that this switch is
 currently global and turns the mechanism off entirely so don't put this
 in library code.
 
+=head1 MOO VERSUS ANY::MOOSE
+
+L<Any::Moose> will load L<Mouse> normally, and L<Moose> in a program using
+L<Moose> - which theoretically allows you to get the startup time of L<Mouse>
+without disadvantaging L<Moose> users.
+
+Sadly, this doesn't entirely work, since the selection is load order dependent
+- L<Moo>'s metaclass inflation system explained above in L</MOO AND MOOSE> is
+significantly more reliable.
+
+So if you want to write a CPAN module that loads fast or has only pure perl
+dependencies but is also fully usable by L<Moose> users, you should be using
+L<Moo>.
+
+For a full explanation, see the article
+L<http://shadow.cat/blog/matt-s-trout/moo-versus-any-moose> which explains
+the differing strategies in more detail and provides a direct example of
+where L<Moo> succeeds and L<Any::Moose> fails.
+
 =head1 IMPORTED METHODS
 
 =head2 new