ietf
[Top] [All Lists]

Datagram? Packet? (was : APEX)

2002-09-24 08:46:40
As a trainer, I like to tell that the word DATAGRAM is built on the same basis 
as the word TELEGRAM. It's actually a kind of "telegram-of-data".

I also usually explain that Packet Switching relies on two different modes : 
DATAGRAM vs VIRTUAL CIRCUITS. The datagram mode is mainly connectionless and 
consists of sending data units through the network using all possible routes 
(e.g. IP), while the V.C. mode is mainly connection-oriented and consists of 
establishing one virtual route prior to sending data units, thus trying to 
virtually reproduce the way of working of circuit switching (e.g. X.25, Frame 
Relay).

Last, while I definitely, clearly prefer calling Layer 2 data units "FRAMES", I 
sometimes [over-]simplify the terminology of Layer 3 by making the following 
distinction : "a datagram is the data unit before fragmentation" ; "a packet is 
a piece of a fragmented datagram".

However, as Dave said, there is no fixed, definitive, universal definition of 
the words "datagram" and "packet". In 10 different books on the subject, you 
will find 10 different definitions. I'm just happy with mine. :)

E.T.

-----Original Message-----
From: vinton g. cerf [mailto:vinton(_dot_)g(_dot_)cerf(_at_)wcom(_dot_)com] 

Louis Pouzin at INRIA coined the term datagram for use in his CIGALE/CYCLADES 
network around 1974.

vint

At 07:17 PM 9/23/2002 -0700, Dave Crocker wrote:
At 06:46 PM 9/23/2002 -0700, Fred Baker wrote:
A packet is a unit of data carried in a packet network,

this just moves the question over the definition of a packet network.  
(I was trying to compress things.)


regardless of the layer. For some reason, we generally refer to layer 
two packets as 'frames' (such as Ethernet, Frame relay, or LAPB 
'frames'), X.25 packets as 'packets', and TCP packets as 'segments'.

please note your use of the word "generally".  as I said, use of these 
terms is flexible.


I can point out literature that refers to each of those as 'packets'.

that was my point.


But a datagram is quite clearly defined.

I believe the term datagram predates IP and seem to recall hearing it 
during the late 70's and early 80's in non-IP venues.

The fact that it is well defined for one use does not mean that the 
definition is, well, definitive.


...from the application's perspective, a wad of data is directed to a 
specified destination, and the application can essentially fire and 
forget.

I could imagine the APEX folks using the term "datagram" in that 
latter sense.

indeed locking the term down for Apex is probably the more useful path.

(but it is slightly amusing that your definition is the same as I used 
for packet-switching...)

d/


----------
Dave Crocker <mailto:dave(_at_)tribalwise(_dot_)com>
TribalWise, Inc. <http://www.tribalwise.com>
tel +1.408.246.8253; fax +1.408.850.1850



<Prev in Thread] Current Thread [Next in Thread>