Moderator:
Angelique Matheny, IBM Rational Software
Speakers:
Lauren Cooney, CTO Office, Information Management Group
Sriram Padmanabhan, Distinguished Engineer, IBM Advanced Tech. Group
Info 2.0 Webcast Notes (I assume that replay information will appear at the site):
The Good
- Need to move way from rigid applications to a more free-flowing approach is correct (synergistic with emergence, user-generated content, architecture of participation and other trends re: Enterprise 2.0)
- Also valuable to note that IT needs to create "safe sandboxes" for end users to create their own situational applications in ways that IT cannot predict or effectively support
- Excellent point on need for a transformation layer when talking about data, feeds, user / business context etc.
- Interesting idea on creation of specific data feeds (perhaps different than normal content-centric feeds) to help with mashup scenarios.
The Bad
- No connection of Info 2.0 message with other messages coming from Lotus. Odd given Lotusphere was just a few weeks ago.
- No product alignment or how this fits into the entire IBM portfolio (e.g., Quickr, Connections, etc).
- Slides point out Google Gadgets but IBM announced at Lotusphere that they were heading down the Eclipse path with iWidget.
- Mashup Hub is not a feed syndication platform (e.g., Attensa, NewsGator or KnowNow). The talking points made it sound like it was a real feed server and it's not really at that level of completeness, positioning what IBM is doing with managing feeds vs. other solutions in the market should have been made more clear.
And The Ugly
- IBM demo'd some feed server technology within its Innovation Lab at Lotusphere that could evolve into a feed syndication platform but it seems like this nugget of gold remains undiscovered at the management level.
- Which means IBM does not seem to strategically understand the importance and role of a feed syndication platform - given the need for enterprise organizations to have a cohesive middleware layer to manage feeds consistently - it remains astounding how major vendors continue to not have any clear direction or solution in this infrastructure area.
For Burton Group clients, we recently published a reference architecture template document on the architectural components of a feed syndication platform. A feed syndication platform:
- Functions as both a communication channel and hub
- Implements a publish-and-subscribe architecture that mediates consumers (e.g., applications that receive feeds) and providers of information delivered through feeds, often based on Extensible Markup Language (XML) (e.g., Really Simple Syndication [RSS] and Atom)
- Contains a collection of functional components that expose services and interfaces, which enable those components to interact with other applications, infrastructure, and environments external to the feed syndication platform
- Delivers feeds to destination points that exist on a variety of clients (e.g., PC, mobile), applications, and server platforms
- Integrates with applications that facilitate delivery and interaction with feeds, such as e-mail, web browser, instant messaging, custom applications, virtual workspaces, and portal sites
- Receives information from a variety of systems (e.g., enterprise content management systems, search engines, messaging platforms, and business applications) that it delivers as feeds
- Honors policy sources (e.g., permissions, rights management, and compliance) as they relate to information sources, feeds, and feed items, and applies such policies across the feed syndication platform (e.g., subscriptions, storage, download, and user management)
- Draws on related infrastructure platforms to complete the end-to-end system (e.g., directory, security, and integration services)

Mike,
thanks for blogging about our webcast and especially the "Good" section above!
I'd to add some additional commentary on a couple of your points.
First, we are definitely working with our Lotus folks. We have quite a lot of synergy between the Info 2.0 mashup server vision and the Lotus vision to enable mashups and widgets in a variety of their collaborative engines. Please do stay tuned for more news to come in this area.
Your other point on feed syndication and its importance is quite interesting. Our current vision for Mashup Hub is make it a powerful feed generation and feed transformation engine. At this time, it may not include all the capabilities of feed syndication. However, as I look at the definitional bullets that you list for feed syndication, there is quite a bit of overlap with Mashup Hub. I am curious on your views on the intersection of feed syndication and feed servers -- where one could fit in vs. the other.
Posted by: sriram | February 15, 2008 at 01:20 AM