fixed some funny characters in comments
diff --git a/qa/apps/qa-manual/WEB-INF/src/qa/utils/MessageGenerator.java b/qa/apps/qa-manual/WEB-INF/src/qa/utils/MessageGenerator.java
index ff12ba5..bcd0c45 100755
--- a/qa/apps/qa-manual/WEB-INF/src/qa/utils/MessageGenerator.java
+++ b/qa/apps/qa-manual/WEB-INF/src/qa/utils/MessageGenerator.java
@@ -21,9 +21,9 @@
* for different channels/endpoints being used by the test applications
*
* This Generator class is instantiated by passing in how many generate calls should happen per second (targetRatePerSecond),
- * and how many CPU time slices (runs) per second youd like it to shoot for (suggestedGeneratePassesPerSecond). It sleeps
- * between batches/runs so you dont want so many batches that its giving up its CPU time
- * too quickly and isnt able to get any work done within a second, but not so few that it
+ * and how many CPU time slices (runs) per second you'd like it to shoot for (suggestedGeneratePassesPerSecond). It sleeps
+ * between batches/runs so you don't want so many batches that it's giving up its CPU time
+ * too quickly and isn't able to get any work done within a second, but not so few that it
* just generates a big burst of messages once a second up front rather than a stream.
* Each generate call (either the doDataMessageGenerate(flex.messaging.io.amf.ASObject) or the doAsyncMessageGenerate() method)
* would create and send a message in an actual test. All in all, gives us the adaptive behavior
