Imaginary Jabber Conversation
To help you better understand what transpires during a Jabber-based conversation, this section presents a conversation between two friends, Jim and John, and two “assistant” applications at their respective places of work.
Jim and John work at two different companies. Both of them invest in the stock market, and they use a Jabber-based service to check and monitor prices, to buy and sell shares, and to manage their portfolios. John’s company also has a workflow assistant that monitors incoming email and coordinates work items between colleagues.
Jim is notified that John is available:
<presence from='john@company-b.com/Desk' to='jim@company-a.com/home'> <status>Online</status> <priority>2</priority> </presence>
Jim sends a quick chat message to John:
<message type='chat' from='jim@company-a.com/home'
to='john@company-b.com'>
<thread>01</thread>
<body>Hey John, have you seen the latest story on Megacorp earnings?
</body>
</message>
John responds:
<message type='chat' to='jim@company-a.com/home'
from='john@company-b.com/Desk'>
<thread>01</thread>
<body>No, where is it?
</body>
</message>
Jim sends John the URL:
<message type='chat' from='jim@company-a.com/home' to='john@company-b.com/Desk'> <thread>01</thread> <body>Here's the link
</body> <x xmlns='jabber:x:oob'> <url>http://www.megacorp.co.uk/earnings3q.html
</url> <desc>Third Quarter Earnings for Megacorp
</desc> </x> </message>
John receives an alert about the price of ACME Holdings (ACMH) falling below a threshold he previously set:
<message to='john@company-b.com' from='alert@stocks.company-b.com'> <subject>ACMH Fallen below 250p
</subject> <body>ACME Holdings price 248p as at 10:20am today
</body> </message>
He checks the price of Megacorp stock (MEGC) by sending an empty message to the shares assistant application:
<message type='chat' to='MEGC@stocks.company-b.com' from='john@company-b.com/Desk'> <thread>T20</thread> </message>
In reply, the shares assistant application sends the required information:
<message type='chat' from='MEGC@stocks.company-b.com' to='john@company-b.com/Desk'> <thread>T20</thread> <subject>MEGC Current Price
</subject> <body>Megacorp price 1287p at 10:25am today
</body> </message>
John likes the price and decides to buy 100 more shares:
<message type='chat' to='MEGC@stocks.company-b.com'
from='john@company-b.com/Desk'>
<thread>T20</thread>
<body>Buy 100 at 1287 now
</body>
</message>
He sends a message to Jim telling him of his new investment. Jim checks his own portfolio:
<iq type='get' to='portfolio.company-a.com' id='port_01' from='jim@company-a.com/home'> <query xmlns='jabber:iq:browse'/> </iq>
He sees that he already holds 1200 Megacorp shares and decides against buying any more:
<iq type='result' from='portfolio.company-a.com' id='port_01' to='jim@company-a.com/home'> <portfolio xmlns='jabber:iq:browse' type='personal' jid='jim@portfolio.company-a.com'> <stock type='standard' name='Megacorp' jid='MEGC@portfolio.company-a.com' holding='1200'/> <stock type='standard' name='ACME Holdings' jid='ACMH@portfolio.company-a.com' holding='500'/> </portfolio> </iq>
The workflow assistant application, workassist, sends John a notice that he’s received an important email:
<message from='workassist@company-b.com' to='john@company-b.com/Desk'> <subject>New mail from: Alastair B
</subject> <body>You have a new mail waiting; details are as follows: Subject: Incident last week From: Alastair B
</body> </message>
He also receives an invite to a meeting:
<message from='joanne@company-b.com/laptop'
to='john@company-b.com/Desk'>
<body>
Hey John, you're supposed to be helping us decide where to hold
this year's Christmas party!
</body>
<x xmlns='jabber:x:conference' jid='room2@meeting.company-b.com'/>
</message>
John decides to leave the mail until later and join the rest of his colleagues in the meeting room.
There’s a great deal we see in this imaginary, but not unlikely, conversation:
There are human and application participants involved in this conversation; Jabber makes no distinction nor holds any prejudice against either participant type.
XML is used effectively to segregate the conversational chunks (Jim’s opening gambit, John’s response, the price alert, and so on) and to structure and identify data within those chunks (the URL for the earnings story, the portfolio information, the conference room invitation).
Conversational strands are kept in context by use of a
<thread/>
tag, so that Jim and John’s clients have a chance to present what’s said in an appropriate way.Each conversational chunk is a
<presence/>
,<message/>
, or<iq/>
tag—the three Jabber building blocks on which everything is based.The conversation takes place across two Jabber servers; one server is located at
company-a.com
and the other atcompany-b.com
.The conversation flow is asynchronous, in the sense that out of nowhere, a message alerting John to a stock price fall below threshold appears seemingly in the middle of John’s conversation with Jim, as does the message from the workassist agent.
All Jabber entities, human or otherwise, are identified by Jabber IDs such as John’s (
john@company-a.com
) or the stock alert mechanism (alert@stocks.company-b.com
).company-b.com
’s stock system is multifaceted. As well as being able to send (and receive settings for) threshold alerts through thealert@stocks.company-b.com
address, it can also interact using different identities to reflect the context of the stock being discussed. For example,MEGC@stocks.company-b.com
represents the Megacorp (MEGC) stock.In fact, this stock system is a Jabber component, a single entity that takes on each of the stock guises in the conversation.
The most common misconception about Jabber is that it’s solely for instant messaging—Jabber is more than a bridge to other IM systems or a set of protocols. Instant messaging is just one of countless ways that an XML-based messaging technology such as Jabber’s can be applied. Jabber is an extensible client/server architecture that routes XML data between clients and services, which plug into Jabber servers as components. The original and core set of components provide the IM features (and supporting services) that were briefly described in the Preface.
The XML structures that make up the Jabber protocol fall into three categories, each represented by a uniquely named top-level tag (also referred to as an element). These tags don’t necessarily carry human-generated IM message content; as long as the resulting XML is well-formed, anything goes. Furthermore, the Jabber protocol design makes use of an XML feature that allows total flexibility of extension: namespaces.
Note
To prevent flooding of the server with large amounts of data, there are mechanisms in place to “throttle” heavy connections. These mechanisms can be configured in the server configuration, described in Section 4.13.
Bearing this in mind, it’s clear that Jabber can be deployed to provide solutions far beyond the IM space as well as within it.
Get Programming Jabber now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.