For those that heavily use virtual documents, D2 makes creating virtual documents a snap. Virtual documents are easily created via an intuitive drag and drop interface, not the clunky and sluggish menu driven interface. In general, the D2 interface is much more responsive than any of the wdk based interfaces.

This then brings us to some D2 architecture specifics. D2, from what I can tell, is not your typical wdk/web based application. I am not yet sure exactly all of the technologies used, but so far, it appears to be an msxml based application using activex/java components. Unfortunately, this means Windows/IE only for real; no getting by sort-of with another browser. D2 communicates primarily via dfs, which is sort of unfortunate, considering how verbose and complicated dfs is; although, depending on the implementation, using dfs might have made it easier to abstract the data layer, and hence the ability to replace the communications layer in the future, than had it been done with dfc and native dfc objects.

As for the notion that D2 cannot be customized, I don’t really see why this is. I am sure some enterprising developer familiar with the underlying technologies can easily customize this application. It’s not like the current wdk based applications are really customizable anyway; anyone who has tried has surely discovered that the wdk component classes are nearly impossible to extend properly; with all of the private methods and attributes, you end up having to decompile and rewrite those classes most of the time anyway. This same technique could be applied to D2 as well.

With EMC’s stance that D2 is meant to coexist along side webtop and other wdk based applications, it would appear that D2 is not meant to play a prominent role in the Documentum UI lineup. However, everyone knows by now that wdk is eventually going to be retired, and the only ‘new’ UI will be the newest Taskspace, as part of xCP 2.0. Could D2 eventually become the one shoe fits all for the content management side of Documentum? I don’t see why not; as far as I can tell, the only thing it is really missing is official customization ‘support’.

Ian Chu

1. Fully integrated installation of new xPlore Search Engine. xPlore gave us significant performance improvements over FAST search engine while requiring less resources.
2. xCP 1.6  Taskspace 6.7 fixed a lot of bugs in v6.6 as well as re-engineered some of the components to improve performance and usability.
3. Records Client – unified UI for various records applications (RM, RPS, and PRM)
4. Initial release of CMIS

Towards the end of 2011, EMC released 6.7 SP1.  Service Packs are typically focused on fixing bugs, but there was some new features that were introduced with SP1 – namely xPlore 1.2 and Stateless Process Engine.  xPlore 1.2 a lot of new capabilities like support for thesaurus, subscription to queries, and cloud support.  Stateless Process Engine allows you to execute series of automatic activities in a single thread.  The performance improvement can be greater than ten-fold.

Here are some exciting trends I see for 2012:

1. Marketplace for solutions will continue to grow.  I am not just talking about Apple App Store or the Android Market.  With the growth of certified xCP solutions, I am hoping that EMC will open its own market place for solutions.  Obviously, a customer will not be able to download and install xCP solution in their environment, but a market place might be perfect place for customers to test drive solutions and get feedback from other customers.
2. Cloud based adoption will continue grow, but security breaches will force customers to re-think if they want to invest into public cloud or have vendor manage their private cloud.  EMC OnDemand for Captiva, Documentum, and Document Sciences is the first step in taking advantage of the private cloud.  However, EMC will have to make their pricing model more competitive to further the adoption rate of OnDemand.
3. Integration between Big Data and business automation will become tighter.  About a year ago, I wrote on my blog about how Captiva Dispatcher being deployed on the Cloud could significantly enhance its ability to categorize invoices.  The next big step would be to use data collected from the invoice processing to dynamically adjust the processing of the invoices to efficiently process and approve invoices.  Currently, most of our customers have strict business rules on who has to approve invoice based on invoice type, department, and invoice amount.  These rules are configured in the process automation software.  However, the configuration is rather static in nature; there is no analytics feedback on whether the rules make sense.  Big Data is the answer to this and hopefully EMC will come out with an xCP xcelerator between Greenplum and xCP in 2012.

Similarly to my state of union last year, if EMC can address #1 and #2 for 2012, I will be thrilled.  If #3 happens, I will be excited.  The potential of cloud based solutions being integrated with big data capabilities is enormous.

Johnny Gee
CTO & Principal Architect