README.md in vendorificator-0.5.git.v0.4.0.63.g8e9d54d vs README.md in vendorificator-0.5.0

@@ -1,8 +1,8 @@
 # Vendorificator
 
-> **THIS PROGRAM IS STILL IN BETA**, use on your own risk!
+> **THIS PROGRAM IS STILL IN BETA**, use at your own risk!
 
 ## About
 
 [Vendor everything](http://errtheblog.com/posts/50-vendor-everything). Keep your own copies of upstream dependencies that your project needs somewhere close to your project, best in the same source control repository. But how to do it, and keep track of all the upstream modules, their origin and license, upgrades, local changes, and so on? It's what Vendorificator helps you with.
 
@@ -252,10 +252,13 @@
 Vendorificator. Takes the same arguments as `vendor`, plus:
 
  * `:command` -- command to run to download files
  * `:specs` -- files specifying what to download; these will be kept
    on the vendor branch together with downloaded dependencies
+ * `:extras` -- files that are needed for the tool to work, but which
+   won't be committed to the vendor branch together with specs and
+   dependencies
 
 Two convenience shortcuts are provided, `rubygems_bundler`, and
 `chef_berkshelf`. They take no arguments. Their definitions are
 examples as well:
 
@@ -270,9 +273,32 @@
 tool 'cookbooks', # <- chef_berkshelf
      :path => 'cookbooks',
      :specs => [ 'Berksfile', 'Berksfile.lock' ],
      :command => 'berks install --path vendor/cookbooks'
 ```
+
+#### overlay
+
+Overlays multiple modules in the same directory, instead of each of them being
+conjured in its own.
+
+```ruby
+overlay '/xyzzy' do
+  vendor 'foo', :version => '0.23' do |v|
+    File.open('README.foo', 'w') { |f| f.puts "Hello, foos!" }
+  end
+  vendor 'bar', :version => '0.42' do |v|
+    File.open('README.bar', 'w') { |f| f.puts "Hello, bars!" }
+  end
+end
+```
+
+### Fake Mode
+
+To develop a solution based on Vendorificator without having
+everything committed to a dozen of branches, tagged, and noted, set
+`vendorificator.stub` Git setting to `true`. Then, Vendorificator will
+work live in the main repo, and ignore the output directory in Git.
 
 ## Contributing
 
 1. Fork it
 2. Create your feature branch (`git checkout -b my-new-feature`)