CHANGELOG.md in activesupport-4.2.0.beta3 vs CHANGELOG.md in activesupport-4.2.0.beta4

@@ -1,5 +1,14 @@
+*   TimeWithZone#strftime now delegates every directive to Time#strftime except for '%Z',
+    it also now correctly handles escaped '%' characters placed just before time zone related directives.
+
+    *Pablo Herrero*
+
+*   Corrected Inflector#underscore handling of multiple successive acroynms.
+
+    *James Le Cuirot*
+
 *   Delegation now works with ruby reserved words passed to `:to` option.
 
     Fixes #16956.
 
     *Agis Anastasopoulos*
@@ -17,12 +26,12 @@
         1.minute.eql?(60.seconds)  # => true
         1.minute.eql?(60)          # => false
 
     *Joost Lubach*
 
-*   Time#change can now change nanoseconds (:nsec) as a higher-precision
-    alternative to microseconds (:usec).
+*   `Time#change` can now change nanoseconds (`:nsec`) as a higher-precision
+    alternative to microseconds (`:usec`).
 
     *Agis Anastasooulos*
 
 *   `MessageVerifier.new` raises an appropriate exception if the secret is `nil`.
     This prevents `MessageVerifier#generate` from raising a cryptic error later on.
@@ -33,18 +42,18 @@
     specifying the order in which test cases are executed. This option currently defaults
     to `:sorted` but will be changed to `:random` in Rails 5.0.
 
     *Akira Matsuda*, *Godfrey Chan*
 
-*   Fixed a bug in Inflector#underscore where acroynms in nested constant names
+*   Fixed a bug in `Inflector#underscore` where acroynms in nested constant names
     are incorrectly parsed as camelCase.
 
     Fixes #8015.
 
     *Fred Wu*, *Matthew Draper*
 
-*   Make Time#change throw an exception if the :usec option is out of range and
+*   Make `Time#change` throw an exception if the `:usec` option is out of range and
     the time has an offset other than UTC or local.
 
     *Agis Anastasopoulos*
 
 *   `Method` objects now report themselves as not `duplicable?`. This allows
@@ -57,32 +66,32 @@
 
     Fixes #9933.
 
     *Guo Xiang Tan*
 
-*   Added instance_eval version to Object#try, so you can do this:
+*   Added instance_eval version to Object#try and Object#try!, so you can do this:
 
-      person.try { name.first }
+        person.try { name.first }
 
     instead of:
 
-      person.try { |person| person.name.first }
+        person.try { |person| person.name.first }
 
-    *DHH*
+    *DHH*, *Ari Pollak*
 
 *   Fix the `ActiveSupport::Duration#instance_of?` method to return the right
     value with the class itself since it was previously delegated to the
     internal value.
 
     *Robin Dupret*
 
-*   Fix rounding errors with #travel_to by resetting the usec on any passed time to zero, so we only travel
+*   Fix rounding errors with `#travel_to` by resetting the usec on any passed time to zero, so we only travel
     with per-second precision, not anything deeper than that.
 
     *DHH*
 
-*   Fix DateTime comparison with DateTime::Infinity object.
+*   Fix DateTime comparison with `DateTime::Infinity` object.
 
     *Rafael Mendonça França*
 
 *   Added Object#itself which returns the object itself. Useful when dealing with a chaining scenario, like Active Record scopes:
 
@@ -316,11 +325,11 @@
     deprecated.
 
     *Pavel Pravosud*
 
 *   `HashWithIndifferentAccess` better respects `#to_hash` on objects it
-    recieves. In particular, `.new`, `#update`, `#merge`, `#replace` all accept
-    objects which respond to `#to_hash`, even if those objects are not Hashes
+    receives. In particular, `.new`, `#update`, `#merge`, and `#replace` accept
+    objects which respond to `#to_hash`, even if those objects are not hashes
     directly.
 
     *Peter Jaros*
 
 *   Deprecate `Class#superclass_delegating_accessor`, use `Class#class_attribute` instead.