My name is Vinnie. I'm lookin' for Lex.

The resurrected Velociraptors are pressing hard on doors and windows trying to break through for a tasty human snack. The only hope is for Lex to get all the systems to power up, and get the lights on and the doors locked. For goodness sake, Lex can do it! She presses a button and graphical display of all the park’s systems appears. She exclaims, “It’s a Unix system! I know this!” And predictably, all the techies in the theater laugh and make dinosaur jokes about UNIX. Seventeen years later, UNIX is still around and shows no sign of becoming extinct any time soon, despite the evolution of the Torvalds Penguin.

Lately I’ve thought about the extinction of big box content management systems: “When ECM Roamed the Earth” sounds like an episode of National Geographic. Is their fate sealed? Can they find new life? Obviously little furry Drupals and larger Alfrescos scurry about eating the Documentus’ lunch, and evolving into much larger beasts while they prey on simpler life forms in the food chain. All the while, growing steadily, quietly beneath their feet, and now threatening to consume them all is… MOSS!

With all due respect, consider that Oracle literally gives Stellent (UCM) away, bundled with other software. Alfresco Community Edition is free. You can find Alfresco packaged with Adobe LiveCycle. What’s on the menu at SourceForge today?

Let’s say you are a CIO having to slash your budget 30% or more this year, in the relentless race to the bottom line below the bottom line. Just using EMC as an example, let’s say that they approach you with a deal for a million dollar starter kit of EMC CMA with all the bells and whistles, bundled with the storage you are considering. They offer consulting with it to sweeten the deal. Then you notice Alfresco, and it is “lookin’ gooood!” plus Scot, your able IT director, assures you he can support it with one half FTE. But then SharePoint seems so easy to deploy and people actually like it. What do you, the CIO, naturally select?

Here’s a true story from my personal consulting journal: One of our customers went through a merger. The parent company was very deep in enterprise content management experience. The purchased company had no significant experience in ECM at all. In fact they received some CM software bundled with another product and had been trying to figure out how to put it to use.

A project was initiated to import documents of record in various life cycle states, from the old company’s systems and boxes to the ECM system of the parent. Finally they would get some real document management power! A collaboration site was configured for the project. An existing document management repository was configured for project artifacts. A repository for the records was available. When the technical lead reviewed the final, signed version of the business requirements document, he noticed major differences between it, the verbal agreements made, and an earlier draft version of the document still on the collaboration site. When he questioned it, he was told that the final version was from a share drive the team from the old company had been using. Well, they were collaborating, and how simple a solution can a share drive be?

Is there any case now left to be made for enterprise content management, in all its voluptuous Jurassic glory? Do the little furry animals on the scene have the life force to quickly evolve and take over? If all the business wants is a primitive one-celled solution, where is natural selection and ECM evolution heading? Even with divine directions from on high mandating ECM, the animals are unruly, and dinosaurs and mice can both be scary!

I don’t know how long this blog and its offspring will last, but I think there’s still a case to be made for full fledged ECM, no matter how expedient a smaller cheaper and apparently simpler solution might seem at first. What do you think? Did I mention SharePoint?

I was fortunate to attend the Microsoft Professional Developer’s Conference in LA as a guest of Microsoft’s last month. Two things hit me right off: (one) pretty much every person in attendance blogged, facebook’d and tweet’d—my apologies in advance to OED—seemingly all at pretty much the same time! (two) Microsoft’s cloud, Windows Azure, is more real than not…as are some others. Better wake up the brain cells.

To me, the truly impressive thing about Azure was, well, not its infrastructure, scale, configurability. Though impressive enough, they’re the annonymous linebackers to the superstar QB—the dude you really pay to watch. And he to me, was ability to develop and deploy using Azure. Seemlessly developing against a “cloudy” SQL server like it’s sitting on your very own laptop (actually better because it won’t kill your laptop in doing so!), writing and compiling your code of choice (give or take…though including Java through Eclipse), and deploying using something other than ANT and a thousand man-hours. Lots of exciting stuff!

This rhetoric though really is not about Azure, nor am I in any shape or form an expert in it. And, besides Microsoft, Amazon, IBM, Dell, EMC and a whole bunch of other folks are in various states of making noise in cloud computing. Watch for it. It’ll impact all of us.

What this rhetoric is about is ECM in the cloud. Hosted ECM or ECM SaaS has, of course, been around for a number of years already. There are a number of good success stories around. Just see SpringCM, WordPress or even Salesforce.com…the lines between managing structured and unstructured content are awfully blurry these days. Are these true examples of cloud ECM? Perhaps. Or maybe the definition is still evolving.

The days of the multinational running a full-blown, multi-purpose ECM platform on a cloud-ready “traditional” ECM toolset (take your pick: Documentum, Filenet, Oracle, Sharepoint, Alfresco, et al) is not far off. There are some challenges left for sure. A little technology. A lot inertia.

Public or private cloud? Oracle or SQL? Java or .net? Or maybe, who cares? Our cloud challenge may well become how to ask more relevant questions to get more useful things done faster, better.

What just happened? I presented a “quote.” I discredited the quote and myself by admitting that I did not do even minimal research on a commonly used Web source. I misspelled John Lennon’s name, the person credited with saying it, and I did not use spell check to catch the misspelling. Your opinion of me is now “in the tank,” if it wasn’t already. Of course John “Lennin” is underlined in red here as I type it in Word – automatic spell check is so convenient. However, being the obtuse luddite that I am (“luddite” is also underlined in red but a perfectly good word), I consider the warning just a decorative protest. However, our relationship is now fundamentally altered all because I did not use culturally accepted technology.

Technology changes how we live. Well, duh. More than that, each advance in technology changes the social motif in unexpected ways. From each point we branch to develop faster, better, newer technologies to support our new set of expectations and the way we live. We relate to each other differently. Our language changes – “Google” becomes a verb, and the culture changes. Facebook allows me to extend my circle of friends and to keep up with acquaintances I would otherwise lose track of. I have 85 Facebook “friends.”  My daughter has over 200. We follow “Tweets” at 140 characters a pop. “Microblogging” becomes a new gerund and culture changes again. Technology that was once the “big bang,” becomes obsolete. We develop applications around Twitter. The pattern repeats itself at higher and higher frequencies, and technology twitches.

On June 9, Armedia and our partner Alfresco hosted an event in Atlanta. It was a packed room. John Newton, co-founder of both Documentum and Alfresco, made a presentation, “Simple Enterprise Content in Complex Environments.” He is an interesting speaker, and has been on the leading edge of technology at least twice, especially as it relates to content management. He spoke of open source, now fully accepted and commercially viable, and in equilibrium with closed source software models. OSS won’t go away. He looks forward to content management “officially” becoming a platform, its common law status made honest by the CMIS standard (“Content Management Interoperability Services”). He also gave us an interesting observation relative to the current world-wide recession. He said that in each of the previous major recessions, over the past 30 years or so, technology advanced in fundamental ways. Mainframes to client server, mini-computers, WWW, and now the whole “iWhatever” phenomenon and technologies to support it – all these advances changed how we live and relate to the world.

So then is “content management” as we think it will become, already obsolete, along with the tools, language, and hardware we use to support it? Just for fun, consider if “Twitter” avoids being crushed by the dinosaurs to become the dominant species, and ascends to be the normative way we relate online. How do we respond? Do we revert to building software and systems tuned to small message blocks? What becomes of the Web itself?  Are rumors of its death exaggerated? What does “content management” mean if most content is immediate and reactive? Define “version management” for Twitter. Of course the term does not apply in the context I normally think of, and that is my point.  My pondering is probably already dated.   The future is so like old.

My son designs and builds iPod Touch style applications presented in very large screen format. He builds them for point of sale applications and information kiosks. His customers are as diverse as automobile dealers and resort hotels.  Things are changing fast.  Contracts are rolling in and his company is making money.  A few weeks ago he complained that he had found only one company with the expertise to deliver the expected content in the format his company needs. The vendor is outside the U.S. Last week we talked again. He told me that he had to let the old vendor go because he had found another one who could deliver what he needs faster and better.  Such a shark!

We can plan for technology as it is, or as we think it will become. However, “Life is what happens when you are making other plans.” (John Lennon). And you can quote me.

So what does this have to do with content management or Armedia?  I’ll take the easy one first.  Armedia and its employees are helping support the effort with funds and labor.  Also, we are planting seeds for future Armedians. : )  As for content management, the hope is to eventually deploy a CMS solution to management ANLM digital assets.  Okay, this is not exactly what you looking for however every organization produces content.  Hopefully, they are managing it in an effective manner.  What does effective mean?  I’ll save that for another blog entry.

James

Today’s topic will cover some simple advice on how to extract more performance out of your Content Server by simply changing the way you ask it questions in your code. Documentum offers plenty of advice on how to trim & tune your Content Servers, but the simplest advice that goes the longest way is: Use DQL.

DQL? But we are in Java land now. Java is object oriented, and object oriented is power. Why would we cast aside our right to exercise Object Oriented Programming and go back to the stone aged days of relational query languages? Speed. DQL queries consistently outperform DFC fetches in most every practical application.

Natively using DFC to operate on objects often incurs a full-fetch or docbase consistency check penalty at the time the object is retrieved. Practically speaking, it fetches extra data that you may NOT want to use or change and you have no concern with. (Do you ALWAYS want to know the last user who saved the document? Or its content type? Or better yet, its filestore location index?)

By using DQL, we are explicitly & clearly instructing Documentum to do two things. 1) Only operate on the attributes we provide in the query, and 2) Translate our statements natively and quickly into SQL, which can be done by a little simple boiler-plate string manipulation by the Content Server.

How much faster? As this issue stems partially from performance fetch and partially from code strategy, it is probably unfair to report any “official” numbers. However, in my experience, I saw an increase from operations that took 150ms to those that could execute in 35ms. That’s it? 115 milliseconds, you say? If you find yourself asking this question, you may not have reached a performance threshold where you are concerned. In this case, you should file this tip away for future reference to say the least. If you have passed the threshold, and you consider how many queries you may be incurring PER operation PER user (say a few fetches to gather data from a few various object types for each concurrent user), you might want to consider implementing this simple optimization trick.

After you begin to use DQL, you may discover that not only is the DQL operation faster, but you can combine some of your multiple-fetches into complex fetches that join faster in the content server and database than programmatically at the point of DFC consumption. Do remember that part of speed is not just choosing faster operations, but choosing smarter strategies.

Thanks for tuning in.

Rahul Raina <rraina@armedia.com>

So why hasn’t there been more progress on this front?  Technology really has little to do with it, technology is actually the easier part of the total effort.  There are several answers and they can vary slightly but it really boils down to the same reasons so many enterprise content management initiatives fail - because there is a lack of analysis, requirements definition, and careful strategic planning up front.  Plus, once systems are deployed, there isn’t sufficient governance and it quickly becomes a support and maintenance headache so the benefits and your ROI disappear.  Something as complex as electronic health records can also be hampered by key stakeholders from different areas failing to agree on a core set of requirements and policies in order to move forward.  If there isn’t a commitment among these interested parties to work together towards a common set of goals and objectives producing a successful solution will be extremely difficult.  This has certainly happened in the past but hopefully with the current momentum this obstacle can be overcome.

One of the key things for organizations planning this type undertaking is to understand and accept that it is not a one time thing.  These aren’t single projects they are programs.  Many projects have failed because the scope becomes way too ambitious and requirements are never properly defined.  With something as complex as electronic health records getting to a refined, specific set of requirements that can be implemented is challenging at best.  Which is why these programs should be broken up into a series of smaller, more tangible projects with scope that can be properly defined and is obtainable during a reasonable period of time.  Taking on more, smaller projects allows you to set more realistic goals, minimize risk, and learn from each iteration so you deploy a more reliable system that meets expectations and yields positive results.  I can’t stress the importance of properly defining requirements down to a specific level, if the requirements are ambiguous and cannot be broken out into use cases you will not wind up with a sound design and the solution will more than likely fail to meet expectations.  I am going to leave this point here since I am starting to cross over into software development methodology and process areas and I have whole separate article planned for that subject.  It may take more than one if I really get on my soapbox!

There certainly have been some good examples, if you haven’t read about it check out the Delaware Health Information Network.  The DIHN (pronounced The Din) as it is lovingly called has been an ongoing program and has been pretty successful.  Will it be a model for other states and possibly more?  I can’t say for sure but having a new VP from the great state of Delaware certainly can’t hurt!

So what can be done to help ensure that an electronic health records system is successful?  For starters realize what I mentioned above, it isn’t a single project, it is a program made up of multiple projects driven towards a common goal.  Of course stakeholder support and resources are a given as with any program but having stakeholders involved that know the different parts of the industry (i.e. Physicians, research, clinical, etc.) and are committed to the common goals is vital.  As with any type of record, health records have compliance regulations that have to be met, the most notable of course is HIPAA.  This is a risky area for electronic health records systems, one that has stumped efforts in the past.  How do you ensure patient privacy when these records become digital, especially with so many potential access points?  This is a big reason key stakeholder input is critical and why there is a great need for governance of the solution once it is deployed.  Roles and permission levels can be established but it is the subject matter experts that need to define them in order to help ensure accuracy.

Once a system is deployed the Governance Committee should be the authoritative body that provides change management guidance and helps ensure the common goals of the solution are being met, including compliance.  As with any good program a road-map should be drawn so the team knows where they are and where they need to go.  The road-map should define key milestones for the program and basically layout the projects to be undertaken along with their high-level scope.  Earlier milestones will probably be more specific and it is OK if future milestones are more nebulous, this should be a living artifact that is revisited, and potentially revised, after each individual initiative at least.  The goal is to lay a solid foundation and build upon it with additional functionality while also factoring in changes in regulation and in the environment.

Although not as important as the aforementioned functional and human factors in ensuring success,  technology certainly has an important role to say the least.  Using content management platforms effectively with intelligent classification, metadata standards, and search applications to manage, relate, and access these records will be vital.  Usability is usually an after thought which is why new system adoption and sustainability can be difficult so it’s role in an initiative like this will be huge.  If the system is difficult to use and frustrating for the end user they simply won’t adopt.  You can build the best solution with all the functionality one could ever need but if nobody uses it the effort will still be an utter failure.  Using proven usability techniques and best practices will help facilitate positive results.  With any major industry shift towards technology, especially with as big as this movement is likely to be, the establishment of certain architectural standards will go a long way to help prevent data from being siloed and to make sure the information can be shared and can cross systems.  There has been success on this front as well that can be used as examples, the Global Justice XML Data Model is certainly one.  Another standard which is in a related area is Regulated Product Submission which is a standard that would allow regulated organizations such as those in the Medical Device, Pharmaceutical, and Veterinary Medicine industries for example to submit regulatory information to the applicable agencies using a single standard message format.

The potential benefits to electronic health records systems are there but what is debatable is whether or not it is realistic to think we can get there.  Eventhough there is strong support there are also those that feel that it is too grand a goal and that the benefit is simply not there.  I was just reading an editorial in the newspaper today where the authors basically said that the money for electronic health records technology would be better spent somewhere else, in fact these authors cited statistics that indicated that this type of technology may actually detrimental to health care.  Sorry, I just can’t buy that.  I just don’t believe we have been doing this long enough and effectively enough to gather accurate data on the subject.  Are there other areas of the health care industry that need funding, of course, this isn’t a silver bullet and to think so could possibly set unrealistic expectations and doom the effort before it ever really got going.

I for one believe that we should, can, and will get there.  Make no mistake, it will be a long complicated journey but one that is long overdue and it looks like it will have a great deal of Federal funding and support which will definitely help.  We really can’t afford not to take this step.  Many experts agree the current state of medical records contributes to rising health care costs.  Furthermore, there is a great deal of knowledge that is sitting there but isn’t being utilized simply because it can’t be found and linked together in context.  I mean, when you go into your Doctor’s office or into a lab to have a test performed do they pull up all your information digitally in a unified context?  No, I am sure they don’t, they walk back to those rolling file bins and pull your folder with hard copies of all your health records, once they find it that is, and only for what you have had done with that Doctor or by that clinic.  Now, think about it again, what if your Doctor is able to have all your records including x-rays, lab results, etc. from anywhere at her finger tips and can be referenced easily?  Or what if you move or are out of town and need to visit a Doctor or hospital and they can have all your records pulled up within seconds?   Once it starts to be put in perspective it becomes a very compelling argument.

What I think has the potential to be a tremendous benefit and improvement in overall health care but isn’t being talked about too much are the eDiscovery possibilities that begin to open up.  Think about it, if this information is digital, well attributed, and classified (content management professionals reading this are chuckling at this notion of ECM Zen but stay with me) this starts to not just be a system that manages discrete electronic health records it starts to become a knowledge warehouse.  For example, with all all of your health records stored this way your Doctors can cross reference past lab results and your history to make a better, more accurate diagnosis.  Intelligence can be built in to flag items if certain tolerances have been met and send alerts.  Reporting across geographical areas can be generated to identify potential outbreaks and if you have ever had the flu you certainly want Doctors and organizations like the CDC to have all the information they can at their disposal.  I genuinely believe that this will not only improve our economy and our health care system but that it will lead to more cures and better health in general.  Health records hold a great deal of valuable information and but they suffer from the same plight that a great deal of other content does, not that it doesn’t exist but that it can’t be found.