W3C home > Mailing lists > Public > xml-dist-app@w3.org > March 2001

Re: WG:RPC

From: Ray Denenberg <rden@loc.gov>
Date: Thu, 08 Mar 2001 14:28:33 -0500
Message-ID: <3AA7DD60.5B911ECC@rs8.loc.gov>
To: Mark Nottingham <mnot@akamai.com>
CC: marwan sabbouh <ftdata@yahoo.com>, rden@loc.gov, ms@mitre.org, xml-dist-app@w3.org
Mark Nottingham wrote:

> * Those that supply generic, 'pluggable' services, like encryption,
>   authentication, etc. These can be used in any message, they're not
>   transport- or application-specific.
>
> * Those that are application toolkits, to standardize common
>   functions, like RPC. These might bring requirements in terms of
>   transport or message correlation requirements, and will only be
>   used for certain applications.
>
> * Those that are applications. Not sure of an example here.

I need  to be sure that an idependent application protocol with its own state
machine (and I have several in mind: Z39.50, Open Archives, Circulation
Interchange, and Interlibrary Loan, for example) fits within this architecture.
One possible view is that these are the "applications" you're talking about.
But what if one of these protocols want to avail itself of the services provided
by these other two modules?  For example, suppose we want to cast the OAI (Open
Archive) protocol in terms of  RPC? Or we want Z39.50 to utilize the pluggable
module, for encryption or authentication?  Then we can't have these three
side-by-side (can we?) So I'm not sure that "application" is a module. It may be
a layer on top, and the collections of modules together is a layer.

--Ray
Received on Thursday, 8 March 2001 14:29:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:58 GMT