[Nel] TCP vs. UDP

Thu, 5 Jul 2001 10:08:09 -0700

----- Original Message -----
+From: "Vincent Archer" <archer@frmug.org>
+Sent: Thursday, July 05, 2001 8:50 AM
+
+
+> To be efficient, any UDP-based protocol needs to be highly aware of the
+> data semantics (not just syntax - i.e. format - but semantics - i.e.
+meaning)
+> of what it carries. Which makes it very complex and usage-specific; you
+> cannot plug in a "random UDP protocol" in your app (in this case, the
+> client-server part of NeL architecture). If you design the protocol in
+> complete isolation of the application (the MMOG) inside, better go
+directly
+> to TCP; it won't be worse than anything you'll design. And it's already
+> made and tested.
+
+I don't believe that this is entirely true.  Let's look at what we need out
+of a protocol for NeL:
+
+1)  Know what data HAS to be delivered and what data doesn't really.
+2)  Know what data can be overwritten by new data if we recieve a large
+chunk of data
+at once or out of order. (If the data is spread out, not read in one call to
+the network object, then we don't need to know if it can be overwritten, and
+in order  the game engine will overwrite it if it's the same class of data)
+
+Problem one is very easy to solve, we simply use a priority argument when
+sending data with our network object.  In this case high priority data could
+be sent via TCP and low priority data could be sent via UDP.
+
+Problem two really isn't all that hard to solve either.  Perhaps we have a
+'class' argument whend sending data via our network object.  When a packet
+is sent it includes it's data class and a serial number, both client and
+server keep track of both.  When we recieve data we check the class and
+serial number.  If the serial number is out of sequence and the data is
+considered old for that class then it's discarded, otherwise it's accepted.
+If the data has no class (class is set to 0) then all data is accepted and
+the serial number isn't checked.
+
+The only problem I see is mainting the large list of data classes.  Really I
+don't think it's that much of a problem and wether or not you're using this
+model if you do discarding of data you'll be doing class checking in some
+form anyway.
+
+I'm sure the rest of you will point out the many other flaws I don't see.
+:P
+
+-E.J. Wilburn
+zane@supernova.org
+
+
+
+