W3C home > Mailing lists > Public > www-ws@w3.org > January 2002

Re: Asynchronous callbacks + www-ws

From: Aleksander Slominski <aslom@cs.indiana.edu>
Date: Fri, 25 Jan 2002 17:25:51 -0500
Message-ID: <3C51DB6F.2F9CA2C8@cs.indiana.edu>
To: Steven Gollery <sgollery@cadrc.calpoly.edu>
CC: www-ws@w3.org
Steven Gollery wrote:

> I'm probably missing something here: the technique being discussed in the
> message you refer to seems to me to be applicable if one resource (reachable
> through a URL) is subscribing to messages from another resource (also
> reachable through a URL). Is this correct? What if the subscriber is an
> application -- is there some way to make this technique work in that case?


there is actually a solution to this: have event publisher to expose pull interface
(instead of push where subscriber must provide contact endpoint with URL that is
used to push events) and have an application to use an agent library that will be
periodically pulling events and then notify application (using call-backs as was
described before).

that will allow to run application to get events even if run behind firewall and if
event publisher maintains messages in persistent storage application can retrieve
events even if there were temporary failures (network going down, event publisher
restarted, etc.).



ps. we are to soon publish XEVENTS that is implementation of event framework on top
of SOAP that provides pull interfaces, generic and persistent event cannel on top
of SQL database and other goodies ...
Received on Friday, 25 January 2002 17:25:37 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:05:07 UTC