README.md in fluent-plugin-elapsed-time-0.0.5 vs README.md in fluent-plugin-elapsed-time-0.0.6

@@ -8,17 +8,47 @@
 
 Use RubyGems:
 
     gem install fluent-plugin-elapsed-time
 
+## Illustration
+
+This illustration draws how fluent-plugin-elapsed-time measures elapsed times.
+
+```
+     +–––––––––––––+    +––––––––––––––+   +––––––––––––––+   +–––––––––––––––+
+     |   Input     |    |  ElapsedTime |   |   Output     |   |   Output      |
+     +––––––+––––––+    +––––––+–––––––+   +––––––+–––––––+   +–––––––+–––––––+
+#on_message |                  |                  |                   |
+            +––––––––––––––––––>                  |                   |
+            |      #emit       | start = Time.now |                   |
+            |                  +––––––––––––––––––>                   |
+            |                  |      #emit       +–––––––––––––––––––>        
+            |                  |                  |     #emit         |        
+            |                  |                  <– – – – –  – – – – +        
+            |                  <– – – – – – – – – +                   |
+            |                  | elapsed = Time.now - start           |
+            <– – – – – - – – – +                  |                   |
+            |                  |                  |                   |
+            +                  +                  +                   +
+```
+
 ## Configuration
 
 Example:
 
-Following example measures the max and average time taken to process [fluent-plugin-grep](https://github.com/sonots/fluent-plugin-grep) => [fluent-plugin-parser](https://github.com/tagomoris/fluent-plugin-parser) => out_stdout chain in messages. Please notice that this plugin measures the total processing time until match chain finishes.
+Following example measures the max and average time taken to process [fluent-plugin-grep](https://github.com/sonots/fluent-plugin-grep) => out_stdout chain in messages. Please notice that this plugin measures the total processing time until match chain finishes.
 
 ```apache
+<match greped.**>
+  type stdout
+</match>
+
+<match elapsed>
+  type stdout
+</match>
+
 <match **>
   type elapsed_time
   tag elapsed
   interval 60
   each message
@@ -26,26 +56,10 @@
     type grep
     exclude foobar
     add_tag_prefix greped
   </store>
 </match>
-
-<match greped.**>
-  type parse
-  format ltsv
-  key_name message
-  remove_prefix greped
-  add_prefix parsed
-</match>
-
-<match parsed.**>
-  type stdout
-</match>
-
-<match elapsed>
-  type stdout
-</match>
 ```
 
 Output will be like
 
 ```
@@ -98,9 +112,17 @@
     Emit 0 on the next interval. This is useful for some software which requires resetting data such as [GrowthForecast](http://kazeburo.github.io/GrowthForecast).
 
         elapsed: {"max":1.013,"avg":0.123,"num"=>0}
         # after @interval later
         elapsed: {"max":0,"avg":0,"num"=>0}
+
+## Relatives
+
+* [fluent-plugin-measure_time](https://github.com/sonots/fluent-plugin-measure_time)
+
+  * This plugin make possible to measure the elapsed time of not only output plugins but also input plugins.
+  * Also, this plugin has a flexibility to measure arbitrary methods of plugins. 
+  * But, this does not have flexibility of output messages such as `add_tag_prefix`, `aggregate` options because of its limitaion of internal mechanism.
 
 ## ChangeLog
 
 See [CHANGELOG.md](CHANGELOG.md) for details.