The last decade saw a war between Portal and CMS frameworks, in which one had to choose between application integration and collaboration tools vs. information lifecycle management (from authoring to versioning) and Web publishing frameworks. With the advent of Web 2.0 technologies such as discussion forums, wikis, blogs, document spaces, microblogging and the increased use of mobile devices, boundaries among content, composites, contexts, and contributors are rapidly vanishing. This convergence where summarized in the previous blog post by the 4Cs Manifesto.

This trend is now pushing organizations to rapidly create all types of new portals, web sites and rich internet applications. Simple, static brochureware sites have become a thing of the past as each business lines need now to deal with a plethora of different public or private web initiatives such as document workspaces, social communities, interactive knowledge bases, customers and suppliers extranets or personalized information dashboards.

On their side, portal technologies are moving to what the Gartner calls User Experience Platforms. But portlet or mashup environments are still suffering from one big inconvenient: they are by default still not enough bound to any content middleware framework. Most portlet or mashup suites are simple presentation facades, which are clearly insufficient to develop real useful composite content applications.

The convergence of content stores, portal frameworks, combined with powerful context-aware publishing systems and social interactions, is pushing traditional ECM, WCM, Portal, and E2.0 vendors to rely upon a new generation of integrated “Composite Content Platforms” (also called “Content–enabled Enterprise Portals” or “Content Application Servers”).

The Content Application Services Decalogue

Rather than trying to structure such a platform according to a classical IT stack, going from persistence to presentation, I privileged a generic, service-oriented vision to demonstrate how all these pre-integrated services could finally benefit to any of your portals, sites and composite applications initiatives.

The 10 key axes of content and application services for the next generation of content-rich applications are illustrated in the diagram and described below.

1. Library Services
As the CMS industry has matured, library services have become a basic core requirement. They regroup all the services required to correctly manage the lifecycle of any content assets, from persistence to content structure, versioning, file plans, observations and relationships. Thanks to the ongoing standardization efforts and open source commoditization, (e.g. CMIS, JCR with some freely available open source Reference Implementations) such library services will rapidly spread to related industries, such as social software, knowledge management, or learning management. Distributing these services in a vendor-neutral manner in the cloud will be one of the next key challenges.

Standards: JCR, CMIS, WebDAV, NoSQL,
New challenges: Content library services as a standardized, scalable and omnipresent utility; Portable data; Distributed persistence in the cloud; Shared ontologies; Content ownership.

2. Identity Services
In an era of transparency, flatness and openness, privacy concerns risk to quickly come back to the fore. Widespread adoption of a single, simple authorization and authentication mechanism is still a dream. The situation is becoming worse with the spread of distributed content environments and aside-storage systems (search indexes, triple stores)which are often duplicating data without solving security issues. Multiplication of similar user properties across systems will rapidly become an obstacle for end-users. At a content object level, the industry still sticks to standard ACL and DRM-kind of feature is still in its infancy, mainly limited to the media and musical industry.

Standards: LDAP, Kerberos, Shibboleth, SAML, XACML, OpenID, OAuth, and many others
New Challenges: Truly transparent, federated and widely used identity management standards; Digital identity as a utility; Distributed user profiles and properties; Improved DRM on any content asset.

3. Process and Orchestration Services
From BPML to BPEL/BPMN, workflow and orchestration services were often considered as a separate piece of the enterprise IT puzzle. Misunderstandings among orchestration services and human workflows, often combined with overly complex business processes modelers and tougher than initially estimated web services integration issues, limited the widespread usage of such technologies. As standards and implementations matured along with simpler Web 2.0 interfaces, we should see the comeback of business process management services natively and heavily integrated into the next generation of composite content servers.

Standards: BPEL, BPMN, BPEL4People, etc.
New challenges: Native integration in next generation of composite content servers; Focus on simplicity; new generation of Visual BPM Modeler; Better integration with mashup services.

4. Content Interoperability Services
Improved content interoperability is bringing the content silo era to a close. As core content library services are becoming a commodity, and as customers are pressuring vendors to open up their silos, content interoperability services and protocols are gaining in importance. Associating and linking content objects existed from the beginning of the web, but semantic web gurus were perhaps too focused on modeling domains to reach the next level of computer intelligence, rather than integrating the core library services required for a common playground. In the coming years we should see a merger of semantic web concepts with other content interoperability services.

Standards: CMIS, RDF
New challenges: Migrate from a silo’ed to distributed content universe; Migration from proprietary binary files to open content types; Evolution from linked data to linked content objects; Widespread usage of common ontologies among content stores; Secured P2P-kind of services on content objects.

5. Semantic Services
Most of the time “Semantic” only refers to the “Linked Data” terms coined by Sir Tim Barnes Lee, which is mainly about better managing “content relationships” (cf. point 4 above). For the purpose of this post, I clearly distinguish “data and content interoperability services” from “semantic services” oriented toward the extraction and understanding of the meaning of content. The latter integrates all kind of techniques required to enhance and enrich any content assets, from simple text documents to images or videos. Such services rely on various algorithms, such as lexical and syntax analysis, natural language processing, object recognition.
Most existing content objects are still unstructured, with a poor set of available annotations and metadata. We are recently assisting to the rise of new semantic enrichment services – including OpenCalais API, Evri API, or Zemanta API – that dig into your content to automatically extract meaning, provide additional metadata and associate it to existing ontologies. Next-generation of content-enabled applications will rapidly embed such enrichment services to offer extra values to end-users.

Standards/Field of Computer Science: OWL, (P)LSA, NLP, etc.
New Challenges: Semantic lifting; Commoditization of semantic enrichment techniques; Standardization of semantic enhancement interoperability services; Adoption of very simple ontologies.

6. Information Access and Retrieval Services
Most content frameworks still rely on classical keyword searches. The Apache Lucene/Solr project is doing a fantastic job to commoditize access to core search services. But information access should not be limited to enter keywords in a search input form. The growth of content available in various stores, the lack of proper information filters, and the need to access to related information directly from your current content consumption context will foster the creation of new search-driven applications.

Standards: OpenSearch API; Need for a new search API standard?
New Challenges: Standardization of federated or unified search; Social search; Distributed LinkedData search; New information retrieval applications; Preemptive recommendations and suggestions; Genius-oriented kind of features;

7. Analytical Services
Web analytics services, leveraging audit trails, logs, and hits, have existed for years. They are now largely commoditized, mainly due to the apparition of the free Google Analytics service. Once considered simply as “nice to have” options, they now have to be at the heart of any application. These services are leading to the next level of “content intelligence”. Mixing public or private-facing audit trails, weblogs or social graphs will foster the development of more intelligent applications. However, finding the right mix among all possible factors, while leveraging such a mass-volume of data to turn them into relevant metrics and useful personalization services, remains a key challenge.

Standards and Techniques: OLAP, FB Open Gaph, ETL, etc.
New Challenges: Content mining; Interest graph; Usage patterns; Automatic audience segmentation;

8. Social and Collaboration Services
One of the recent goals of any composite content platforms is to let developers rapidly socialize their applications. Most so-called E2.0 software employ a top-down approach focused on the added value of adopting an enterprise social network that mimic Facebook rather than promoting the integration of social as a service.
Last year, socializing an application was mainly about enabling user generated content, such as comments, ratings, likes, and tags. We are assisting this year to the multiplication of all kind of activity and action streams. We should see tomorrow the rise of more standardized and distributed social profiles and collaboration protocols which will be natively integrated into the next generation of content application server.

Standards: ActivityStreams, OpenSocial, XMPP, Wave Federation Protocol, Salmon, SIOC/FOAF
New Challenges: Distributed user profiles; Unification of activity streams, Social archive management; Native integration of standardized collaboration protocols;

9. Mashup Services
Portal technologies have been decoupled from CMS for years, but as the web evolves, there is a growing need to develop content rich applications. The all-purpose MyYahoo-type of dashboards is reaching its limits. Users now want access to their back-end processes and information from within the context of their sites. Native support for portlets, gadgets, widgets, and other types of shareable micro-applications within a composite content platform is becoming a necessity and implies that your platform should natively support a good level of mashability.
Moreover, advanced end-users now want to create all the data, content, or process-driven mashups on their own. All these user-generated mashups will populate a new generation of enterprise app store. Composite Content Platform will also become in charge of offering all the discovery and governance services needed to efficiently manage such a repository.

Standards: JSR168, JSR286, OpenSocial, EMML, etc.
New Challenges: Mixing public-facing data with private-facing information; LinkedData composition; Comprehensive visual mashup builders; Governance of enterprise mashup stores;

10. Rendering Services
Users and developers should be able to rely on strong rendering and delivery services. The multiplication of sites (per audiences, language, topic, and so on), devices (including web, mobile, and social networks), personalization (such as user preferences, implicit personalization, or geolocalized settings), standards (including cross-browser compatibility, or new HTML 5 standard) and multi-scripting support (including JSP, velocity, groovy, or php) make rendering a complex beast. Combined with all the front-end delivery caching issues (CDN or similar natively embedded front-end caches), and the integration of highly dynamic UI components to create a rich user experience, content rendering and delivery has become a critical service for any content application server.

Standards: (X)HTML, WCAG, Javascript, AJAX, Multi-scripting, ESI/CDN
New challenges: Dynamic contextual and personalized content renderers; user experience platform; multi-device support; clear separation of form (context) from substance (content, data, metadata); easy content repurposing.

By leveraging at best such a collection of services, developers should be able to more rapidly assemble, reuse, and mashup existing content assets and composites into a new generation of content-rich applications. How many E2.0 solutions are today reinventing the wheel at a content platform level into what will rapidly transform into a new generation of Social Silo?

The rise of composite content servers signals the convergence among application servers, portal frameworks, social services, and content stores.

But in choosing a robust enterprise-grade composite content server, it is then critical to:

• Ensure that your data and content are clearly separated from your front-end applications from both a vendor-neutral and standardized manner.
• Understand how you can rapidly develop and assemble hundreds of contextualized sites and applications, employing all your content assets, social properties, or composites into such dedicated solutions.
• Make sure that your content application runtime can scale and adapt to your needs both on-premises or in the cloud.

Open Source Consequences

Today’s composite content servers are focused on trying to glue together a generic, unified content store and all the related services into a single consistent, robust, scalable and stable middleware platform.

But the IT industry’s commoditization is ramping. In 2005, we assisted to the commoditization of application servers with the rise of middleware such as JBoss, Apache Tomcat, and similar offerings. We should now look for accommodation at the content application layer, both from free proprietary solution (e.g. SharePoint Foundation 2010) or from open source alternatives (e.g. Apache Jackrabbit+Sling, Nuxeo EP, JBoss GateIn, Alfresco Content Application Server, etc).

However delivering such a complete stack requires skills and ongoing investments in so many different specialties that it will certainly and rapidly lead to some level of consolidation in the industry at least in the open source world. Open Source Software vendors will then focus on delivering and selling finished applications on top of the stack while working together at a platform level.

Where’s this all going?

Monolithic, ever growing, content application platforms, even if composed of hundreds of relatively independent sub-libraries, are dinosaurs on the way to extinction. Their lack of flexibility and their high integration and maintenance overheads will spur the turn to a more agile and distributed approach.

So let’s try to foresee some trends and their possible consequences on the content application server market:

Trend #1: The cloud as your main distributed content and data store
Content assets need do become omnipresent. Open, standardized, interoperable, and distributed content stores should allow any content owner to securely store, access, and archive any kind of content asset from a transparent manner. Customers will buy disk quotas and value added services (such as replication, backup, caching, and secured data center) rather than applications. As mentioned by Lee Dallas: “Application archives need to be ultimately independent of the array of systems generating the data they store.”

Trend #2: Distributed RESTful Services
The ever-growing Internet bandwidth, improved security on cloud computing, together with the generalization of distributed RESTful APIs, should lead more and more software vendors to focus on specific niches and offer specialized services on the cloud which could be easily integrated back into your custom solutions.

Trend #3: Rapid assembly of content-rich applications
To best leverage trends #1 and #2, developers and power users should be able to rapidly compose the applications they need by federating content from several stores and by mashuping it together with all the required, local or distant, services they need.

This will lead to the creation of dynamically assembled composite content platforms relying upon widely distributed services which will connect to, reuse, and repurpose information available in various local or distant stores.

However, this vision of a truly distributed content application server seems unrealistic for the next 5 to 10 years, at least for enterprise-grade, mission-critical applications, even if it was already actively promoted by CORBA ten years ago, then by SOAP/UDDI, and now by OSGi. Too many standards are still missing. Cloud-based document storage (such as GDocs and DropBox) should transition into more granular content centers. Content interoperability and data portability should support more complex use cases. Content owners should be able to manage their distributed content in a transparent and secure manner, no matter what content centers they employ. Distributed SOA/WOA services should have the same level of performance and security as local, internally installed services. Identity and authorization mechanisms need to be standardized at an international level. Then the content industry will have entered a new era.

Meanwhile, we should expect to see an intermediate, bridging phase with the delivery of micro-kernels, pre-populated with different pluggable, hot-deployable services, hosted both locally and in the cloud with content openness and federation in mind.

Next Episodes: the saga continues

We can now see how composite content platforms will serve both as development foundations and as a production runtimes to develop your next generation of content-rich applications be (a web site, a portal or a 2.0 kind of applications).
Next time, I will go into more detail about:

• Development and integration tools needed to foster the rapid assembly of composite content applications.
• Illustrations of content-rich applications which can be developed on top of such content composite platforms.
• Products and Tools that let you manage all your content assets across all your sites and applications.