Friday, August 17, 2007

RSS feeds begin to bleed into enterprise applications

You've probably just gotten used to the idea of "mashups" for quickly bringing web services into applications and portals. Well, now get ready for making novel and powerful use of content via RSS feeds in a similar way.

I don't call them mashups, though, I call them "Feed Bleeds." That's because syndicated feeds can be easily bled into one another to form aggregated streams of content. Not only that, the users and/or developers can increasingly control how much of one type of content should be bled in with another.

The use of RSS feeds as conduits for distributing and managing content, data, and media acts as a complement to more programmatic displays of appropriate relational data in applications via such means as ODBC, JDBC and SQL. Whereas these data access protocols target structured content, the RSS and/or Atom feeds open up the spigot to much more information.

What's newly powerful is that nearly any kind of content can be driven through these feeds -- from documents, spreadsheets, and data to video, blogs, podcasts and online HTML instruction manuals. Feed Bleeds allow for human knowledge in natural language to mingle and complement IT-based assets such as data, application services and automated event-driven processes. Think of it as broad integration on the cheap -- and fast.

Unlike programmatic approaches, the developer can hand off to the end users the subscription to and fine-tuning of the content feeds. Users can adjust how much or how little on a subject they want. Businesses can control what feeds are allowed into the network. Get general information on one business subject and highly specific content on another. The work process need determines the right mix of procured feeds.

Indeed, the users can begin to use search in-house or and online directories to find the syndicated content they wish to add to their applications and process views. We're now seeing a lot more custom enterprise applications that contain and exploit RSS-based content. We're seeing enterprises identify more in-house content that they should expose as feeds.

I think the new Feed Bleed benefits are too powerful to ignore. By quickly finding information on almost any topic that's delivered through lightweight syndication, subscriber/aggregators can shape that information flow to help them in their work, then adjust the content qualitatively and quantitatively as needed to best contour the information to the tasks at hand.

Developers can give the users the tools to make content appear in context to business processes. RSS-enabled Windows Vista and many freely available, standalone news readers help, but a cadre of back-end servers with associated APIs also now allow the productive exploitation of feeds, content and services within applications. We're even seeing a conceptual page borrowed from service oriented architecture (SOA) in the form of RSS feed "buses." This really is a case of Web 2.0 leading to Enterprise 2.0, leading to mainstream enterprise IT.

On-premises servers provide the management and integration of feeds. There are also on-demand feed tools. Onsite feed bleed providers include Apatar, Inc., JackBe, Kapow Technologies, RSSBus, and Strikeiron, Inc. These suppliers allow all kinds of content (HTML, XML, PDFs, spreadsheets, CMS, RDBs, SOAP, REST, as well as RSS/Atom) to be bled together, organized, managed and presented. Online mashup tools come from Dapper, OpenKapow, Teqlo, and Yahoo Pipes, among others. Apatar also has a hosted online offering in the works.

What's more, software as a service (SaaS) business applications providers like Salesforce.com (maps and data merge) and Workday are providing more mashups and feeds-based and enhanced services. If it's good for a SaaS provider, it should be good for an enterprise (as it acts as the service provider to its internal and partner users).

The open source world is also a fan of feed bleeds. An increasingly effective lightweight database aggregation approach involves creating specific feeds of data from, say, MySQL data and SugarCRM applications, that are then aggregated into common feeds that provide a single view of a customer, or an order, or a business process. This allows for whole new kinds of workflows, applications, and processes -- but on an agile time-frame. Interfaces are quickly evolving to allow for drag and drop means to create and adapt feeds. Even a business manager could do it!

IBM's vice president of emerging technologies, Rod Smith, is a fan of giving users the ability to finer-tune the content they need for their jobs. IBM itself has produced what it calls a "situational application" tool, a mashup enabler built on Zend Framework called QEDWiki (Quick and Easily Done). Smith recently told me he likes the idea of bringing together the Web 2.0 and enterprise IT communities so they can begin to work together, even if they don't necessarily speak the same language.

As Web 2.0 empowers younger workers to manage content online in new ways, they will want to use similar approaches on the job. Should this de done via an end-run around IT? Or should IT embrace and extend mashups and feed bleeds?

I think it's clear that this one is too good to ignore.