Apoorv Durga's Blog

I have moved to a more permanent location at http://apoorv.info.

Apologies for the inconvenience.

The consolidation continues. Hummingbird, with its strength in Document Management has announced the acquisition of European RedDot solutions. RedDot has its strength in Web Content Management. This gives Hummingbird and RedDot an opportunity to address the ECM space and cross sell to each other's customers.

However, true benefits will only accrue if the two vendors actually integrate the products tightly. This means having a common user repository, common content repository and some kind of workflow interoperability. Otherwise, the two would just be like desparate products, with companies having to duplicate data and users, which defeats any advantages that could be got by a single ECM product.

Okay that might sound frivolous but there are reasons. Google has entered the personalized portal space. Check out http://www.google.com/ig. Although, it is a basic and rudimentory offering but it still is a good beginning. Everyone knows about Google's different search models - google appliance, desktop search, web search and thousand other different types of searches. Besides, they also have a "Search History" and some other cool search related features. In the collaboration space, they have Hello (http://www.picasa.com/picasa/redirect.php?tid=Y2NpZD0zODcx) which could be their IM offering and Google groups (http://groups-beta.google.com/grphp?hl=en&tab=wg). They also do news syndication, maps, froogle and catalogs.
As far as content management goes, they are a leader in blogging which essentially is a poor man's CMS.

They already have API's to access many of these features. With a couple of more acqusitions and some consolidation in different services they have, they might be a player to watch for in the PCM space.

Focus on Usability is going to be a big trend in Portals and Content Management implementations. This follows from two main reasons:

More and more organizations are rationalizing and consolidating different sites and portals that had proliferated in the enterprise. Each of these have a different look & feel, navigation and a different user experience. Hence, it will be important to focus on usability in order to provide a consistent user experience.

Secondly, most of these products force you to follow a "design template" provided by the product. For example, almost all portal products let you choose a single column, 2-column or a similar template. It is certainly possible to have your own design but it is really difficult to actually implement it. Hence, the design is actually driven by what a product can (and can not) do rather than the other way round.

Some analysts think that a "Stand Alone" Portal Server is dead. This follows from the fact that many portal features have been commoditized and made part of infrastructure/application servers. The fact that portal servers are expected to do much more than they initially did means that they have to interact more closely with services provided by application server. This gave app server vendors an opportunity to differentiate by providing portal features. Hence more and more companies are going in for infrastructure vendors like IBM, Microsoft and Oracle. Smaller vendors will either be gobbled up or will need to provide some niche capabilities if they want to survive in this market.

IMO, in the future, more and more clients will buy or use portal features provided by the application server vendor. Laundry requirements like search, content aggregation and integration, basic content management, personalization etc will be provided by these. There will be niche offerings such as eLearning, collaboration etc for which these organizations will buy products that will need to integrate with the application (or portal) server.

We run our group's Portal on a leading commercial Portal Server. Since it is an internal portal, most of the development is done by people when they are not on projects (On Bench that is). This leads to too many issues. If someone has to work on one of these commercial portals (BEA, IBM etc), the client requirements are humongous. For example, BEA Workshop needs 1 GB of RAM on client desktops!! IBM's WSAD also has similar requirements. I know one can use a low end IDE (Eclipse etc) but most of the tutorials and sample examples expect that the required client software is set up. I also feel that because of the fact that these products provide sophisticated features, it is even more difficult to create *simple* applications because the overhead of learning and getting up & running is too much. These products are just too complex for novice users!

I'm thinking of migrating our portal to an open source (OS) offering. One of the uses of this portal is that it is like a brand for our group (I work with Portals and Content Management group of a leading Indian SI) and we use this portal to give demos to clients. Hence i'm spending a lot of time to decide whether this is a right move. However, with OS offerings maturing and more & more clients themselves considering OS, this might be a right move. Plus, if we use an OS product that is standards based (like JSR 168), we can always migrate the portlets to another portal server if there is a specific client requirement.

Because of the features that we require and some other factors, I've short listed Jetspeed and Liferay. However, Jetspeed-1 is not standards (read JSR168) compliant and Jetspeed-2 is not available for download yet. That leaves me with Liferay. Its a good product but I'm not yet convinced that it can compete with the more sophisticated commercial products.

PS: Because of political reasons, we are limited to java based products.

Looks like content is indeed the king these days. Today, Sun announced a proposed take over of StorageTek. This gives them a foot hold in a fast growing Information Life Cycle Management (ILM) space. Next step would be to buy a Content Management vendor. Once that is thru, they will have presence in Server market, Storage and tools to manage that enormous data. This would also give some competition to EMC which became a big player in this area after they bought over Documentum in 2003.

Fatwire Content Server could have been a probable candidate. Its a brilliant product based on J2EE technologies had shared close relationship with Sun. Vignette could have been another choice but recently they've started focusing *again* on .Net. OpenText and FileNet could also be in the race too.

Microsoft already have a good offering in this space after buying out NCompass. IBM, BEA and Oracle too have small time offerings in CM space but they can not match the sophisticated feature set of Vignette, Interwoven et. al.

I think its not too far when all the biggies - IBM, BEA, HP, Oracle and Sun will have a content management offering, most likely thru an acquisition. Smaller players and Open Source will still be there in the market. However, they will need to be niche and should differentiate themselves from these big guys. After all, once Oracle or Sun has a CM offering, no one can stop them from commoditizing it and offering it as part of bundle free of cost.

An interesting thread about CMS modules has started "again" on the cms-forum list. As usual, there's a debate going on on what features should a Content Management System have. Many people seem to think that features like calendaring, discussion forums, polls and surveys, etc. should be a part of CMS. I disagree. These features are certainly very cool and form a part of any web initiative that requires collaboration, including employee intranets and customer facing websites. However, we must separate them from core content management systems. The problems that a CMS tries to solve and those addressed by this applications are very different.
One can always say that the data generated by, say a voting application is content and hence by definition, any application that handles that is a content management system. If we were to follow that line of thinking, then pretty much everything will be a content management system. An ERP, a CRM system, a search engine (Google??) - you name it and it'll be a CMS.

In my opinion, a CMS should provide core capabilities like:

• Content Creation - Web based forms for content entry, integration with content creation software like MS-Office
• Content Management - Repository Management, Workflows, version management etc
• Content Delivery - Some kind of templating that separates content from presentation and a mechanism to publish content

It would be good to have interfaces to search engines and other applications for dynamic delivery of content. This includes applications for security, personalization and so on. If some of these are built into the CMS, that should be considered additional.