W3C home > Mailing lists > Public > ietf-discuss@w3.org > December 2000

Use of the term "Client" and "User Agent"

From: Jacob Palme <jpalme@dsv.su.se>
Date: Wed, 13 Dec 2000 11:43:21 +0800
Message-Id: <p05001902b65c9e028c8b@[10.10.2.85]>
To: cdn@ops.ietf.org
Cc: discuss@apps.ietf.org
I am concerned with the use of the term "Client" in the Content
Distribution Standards Work. In most other standards, two terms are
used:

User Agent = The ultimate origin of a request and the ultimate
recipient of the final delivery.

Client = A role played by one agent in relation to server agent. A
client is the agent which asks the server for some services, and the
server is the agent which delivers these services.

In the case of the DNS, for example, we can have the following structure:

+------------------+    +--------------------+    +------------------+
| DNS User| Client |-->-| First DNS | Client |-->-| 2nd DNS | Client |--> etc.
| Agent   |        |    | Server    |        |    | Server  |        |
+------------------+    +--------------------+    +------------------+

Thus, what you in your work name "Client" is what most other
standards call "User Agent". And most other standards define "Client"
as a role which many different agents can fulfill in regards to a
server.

Perhaps you should revise your terminology to agree with what is most
common in most other standards work.
-- 
Jacob Palme <jpalme@dsv.su.se> (Stockholm University and KTH)
for more info see URL: http://www.dsv.su.se/jpalme/
Received on Wednesday, 13 December 2000 06:10:58 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:38:00 UTC