ChangeLogTag:Tue Mar 28 14:40:54 2000 Carlos O'Ryan <coryan@uci.edu>

1 files changed, 13 insertions, 6 deletions

diff --git a/TAO/tao/GIOP_Message_Lite.i b/TAO/tao/GIOP_Message_Lite.i
index 251906c72e1..0815bf79756 100644
--- a/TAO/tao/GIOP_Message_Lite.i
+++ b/TAO/tao/GIOP_Message_Lite.i
@@ -1,19 +1,24 @@
 //$Id$
 
 
-
 ACE_INLINE int
 TAO_GIOP_Message_Lite::parse_header (TAO_GIOP_Message_State *state)
 {
   // Get the read pointer
   char *buf = state->cdr.rd_ptr ();
-  
+
+  // @@ Bala: i added the following comment, does it make sense?
+  // In GIOPLite the version, byte order info, etc. are hardcoded, and
+  // not transmitted over the wire.
   state->byte_order = TAO_ENCAP_BYTE_ORDER;
   state->giop_version.major = TAO_DEF_GIOP_MAJOR;
   state->giop_version.minor = TAO_DEF_GIOP_MINOR;
+
+  // Get the message type.
   state->message_type = buf[TAO_GIOP_LITE_MESSAGE_TYPE_OFFSET];
-  
+
   state->cdr.reset_byte_order (state->byte_order);
+  // The first bytes are the length of the message.
   state->cdr.read_ulong (state->message_size);
 
   return 0;
@@ -25,19 +30,21 @@ TAO_GIOP_Message_Lite::
               TAO_OutputCDR &output)
 {
   // Write the GIOP header first
+  // @@ Bala: i still believe it is a bad idea to have those generic
+  // enums for all the message types....
   this->write_protocol_header (TAO_PLUGGABLE_MESSAGE_REPLY,
                                output);
-  
+
   // create and write a dummy context
   // We dont really need this. But for some reason the reply is not
   // parsable on the receiving side?
-  // @@Bala, look in to this
+  // @@ Bala, look in to this
   IOP::ServiceContextList resp_ctx;
   resp_ctx.length (0);
   output << resp_ctx;
 
   // Write the request ID
   output.write_ulong (request_id);
-  
+
   return 0;
 }