JAX India 2007: Web2.0: What you should do?

Craig McClanahan talked on the Web2.0 at JAX India 2007.You can find more detilas about what he covered in Shishank's post. Here’s a bit of elaboration of those 10 points what Craig suggested to make the web right :-
10 - Expose Data/Logic as services
* Content is more important than presentation.
* Use REST based Services when you “Can”
* Use SOAP based Services when you “must”
9 - Incorporate External Content
* No single application or database can contain everything
* Combine available content from multiple sources
8 - Seek QOS (Quality of Service) deals from Sources
* Authentication guarantee
* Performance provisioning
* API and format compatibility
7 - Give QOS Deals to users
* External consumer will also become dependent upon your content
6 - Adopt Agile Processes
* Continuous iterative improvement model
* Incremental release every 7-14 days
5 - Test Driven Development
* Reduce release testing
* Do aggressive unit /functional testing
4 - Architect for Scalability
* Separate view, logic and persistence
* Break into layers that can be independently scaled
* Add resource as needed for bottlenecks
3 - Embrace Heterogeneity
* Data and logic as “service” insulates layers both internally and externally
* Agile Technology benefits in for fast UI fixes
2 - Reach out to Mobile Clients
* UI for Mobile devices can share existing service
1 - Enable User Provided Content
* Participating in “mashups” counts
* User like to participate not just “view”

The suggestion that Craig has provided, gives a clear perspective on how Web2.0 should be incorporated gradually within your organizations portal, CMS or website.If you read the above points again I feel that some of them directly or indirectly take you to the SOA space where we re-architect our implementations so that our applications can be “loosely coupled” and “interoperable” when used as “services”.

Portals: Intruding the ECM space?

If Portals are the underlying technology for the presentation, aggregation, integration and SOA implementation (as every vendor talks about it ), then the Content Management System (CMS) is what feeds the portals.

Portals have been designed to integrate with CMS products as a part of their offerings. At the same time they have come up with their own version of built-in CMS.

Besides Partnering with vendors from Web Content Management (WCM) or Enterprise Content Management (ECM) arena, Portal vendors are also working successfully towards their own built-in CMS. These built-in CMS provide you the best a CMS product can offer. Journal Content Management, Document Management, Integration with MS office or Open Office, Drag and Drop of your desktop files, Workflow management, Integrated Publishing and Search.

WebSphere Portal 6.0 Integrates externally with CMS products like Interwoven Teamsite ,Documentum etc , also includes IBM Workplace Web Content Management Version 6.0 which itself carries a full fledged WCM capabilities. BEA provides the integration with Stellent, documentum and Vignette CMS products but has its own Content Management system and virtual content repository. Open source player Liferay portal 4.2 has come up with portlets for Alfresco (another open source Leader in ECM),but again it has its own "Liferay Journal" CMS which covers most of the WCM functionalities.

Sun and Bea have partnered with FatWire to provide Portal Server customers with unlimited-use licenses of FatWire Spark Portal Content Management (pCM) software at no cost.

There is a definite and clear separation of a built-in CMS and a third party CMS integration on to the portals. The choice is up to you, it all depends on what satisfies your business requirements. As for small and mid-size customers, these built-in CMS are doing the job, that too with a lot of ease and bringing in cost benefits.

CMS market has always been more mature, streamlined and more professional. On the other hand, the Portal market are not making the impact that was expected of them and was so fiercly predicted by the experts, a few years ago. The reasons could be cost, need, ease of implementation or lack of expertise with the service providers. In such a scenario are Portal vendors adding on CMS to their product feature list to stay alive?

If yes, what next? Would they continue to break ground and venture into ECM territory as they have now ventured into WCM territory?

"Portals" and "Portal-like"

Few days’ back I got to know about the something called “portal-like” . wow...a new word and a new concept after Collaboration, Web2.0 and SOA in the portal space. Drilled down a little deeper I came to know that this concept make sense when it comes to mid-size enterprise ,here it goes....

Portals Implementation gives you the solution but with a huge investment of money, infrastructure, resources and support.
Mid sized Customers who wants the portal features (with less investments) now can use the “Portal-like” products to leverage the features what a portal offers. These “Portal-like“ frameworks uses the Jsp, Jsf environment as a core language to built the applications and provide the addition runtime environment (portlet container) to consume/include the portlets on top of the jsp, jsf application. These frameworks support the deployment of standard based portlets (jsr168, wsrp1.0) together with the vendor specific portlets.
End users can also be provided with “virtually” customization features that will be a part of JSF/JSP/AJAX implementation.
These products are well suited where the core application takes more functional space than the utility portlets, which is opposite in the case of actual portal implementation.

There are few pros and cons of these frameworks/tools/products -
Pros-
1) Comparatively less investment than the Portal product
2) No portal expertise required to develop “Portal-like“ application
3) Use of standard based pre-built portlets (blogs,wikis,cms,rss etc )
Cons-
1) Application can not be easily scaled to portlets in the near future :)
2) Cannot be used as Integration point with other Enterprise/legacy applications.
3) Confidence of the customer towards the solution vendors.

The Introduction of such a framework is done by Oracle with its Web Center Suite. That provides the Content Integration, Runtime Customization, Metadata Management and Communication Services (Again a new Word for me “Presence Server”)

Enterprise Portal and Service Oriented Architecture (SOA)

Service-oriented architecture (SOA) is the talk of the enterprise fraternity. Not only the business is getting benefited but the potential of SOA is affecting the speed of application development process.

This buzzword strikes Portal technology as well. These days organizations are in the process of migrating from traditional client/server, monolithic or disparate n-tier architecture to a more loosely coupled and interoperable environment, this process is then better termed as Service-Oriented Architecture (SOA).Most of us are aware of SOA and its benefits, I wont be repeating those here but will discuss that how Portal can be used to leverage SOA.

When we talk about the underlying components of SOA i.e. Interoperability, Scalability, and
Integration, the best suite that does and support all these is an Enterprise Portal Implementation.

A Portal based on standard like JSR168 and WSRP can be the most eligible candidate for SOA. When we look at the Portal’s Infrastructure, we find that it comprises of nuts and bolts of Service Oriented Architecture.

So the next thought that comes to mind is how Portal fits in SOA?

Portlets are reusable web components which forms the core of a Portal Desktop, providing relevant and customized information.SOA’s Interoperability is achieved by deploying JSR 168 Portlet to any JSR 168 compliant Portal server. Also, a portlet can be reused in different Portal pages as well as in different containers which proves reusability aspect of SOA.

Web Services, one of the main building blocks of SOA, fits nicely with Portal as WSRP based portlets. WSRP enabled portlets, which adhere to Web services standard such as SOAP and WSDL, enables portal to easily include services from a third party into a portal page. This eases and reduces the development process and makes use of already available enabled services. WSRP provides much support for the basic pillars of SOA like SSO, Reusability, Access management, service integration with loose coupling (here I mean to say that different processes should be integrated but without the processes being dependent on each other ).

Lastly, Role based content delivery model in Collaborative Portals makes a value addition for role based SOA Implementation.

The above information perfectly fits with the Gartner’s conclusion that

“The portal can be a logical and appropriate first step toward SOA implementation because its fundamental nature lends itself to SOA approaches”

RedHat to acquire JBoss

The JBoss war has finally take a not so expected turn when Red Hat announced to acquire JBoss. Red Hat and JBoss have been recognized as open source leaders in their respective domains. The large and vibrant communities around Linux and JBoss prove that the open source development model creates innovative, quality software, while providing a flexible and low cost model for customers. This acquisition is expected to accelerate enterprise adoption of open source infrastructure, and broaden the entire market opportunity for existing and new Red Hat and JBoss partners who are building
value-added enterprise solutions.

RedHat have already tried its hands on middle ware front (Jonas as application server),and much of the R&D was on portal and CMS with no success. With this acquisition RedHat will get a powerful web stack and the customer will get a product with more performance ,security and the branded and well respected support with a low –price.

As far as my speculation, I don’t think that RedHat will offer any support for Windows Version, rather it may stop JBoss’s windows version, because that will effect its own Linux distribution.

Anyways “All well that ends Well” --- OpenSource to OpenSource” :)