Republish Article: Soa and Modeling, a Lot of Fud (fear, Uncertainty and Doubt)

Soa and Modeling, a Lot of Fud (fear, Uncertainty and Doubt)

Author: Max J. Pucher

In the following article I would like to point out the difference between Microsoft's "vision announcement strategy" and the ISIS Papyrus real-world product availability.

In 2003 Bill Gates publicly said that Microsoft saw modeling as a focus for Microsoft developer products. Recently – four years later – Microsoft has renewed its modeling vision as part of an SOA (service-oriented architecture) and Business Process strategy codenamed 'Oslo'.

Microsoft has at best made little progress in modeling as part of its development tools but full modeling will take until Visual Studio 10 and supposedly will include the .Net Framework, BizTalk-, Team Foundation-, and SQL Servers.

According to Steven Martin, director of product management, Microsoft will TRY to unify its model-driven development tools, WCF (Windows Communication Foundation) and WF (Windows Workflow Foundation).

Donald Ferguson, the gentleman who invented IBM's vapor ware concept of renaming everything to WebSphere recently joined Microsoft and announced that model-driven development would play a key role in Microsoft's SAAS (software-as-a-service) offerings in a cloud computing scenario. Given that 'WebSphere' is no more than a cloud, that makes a lot of sense and tells us what we can expect from Microsoft. A lot of smoke and mirrors, renamed products, great visions and strategies and a lot of FUD (fear, uncertainty and doubt). Ferguson admits already now that he sees completely executable business models as not achievable. He believes that only the first version of an application can be done by modeling because he still believes that models will be translated to programming languages. There is no FULL ROUNDTRIP development in his mind so we will not see something radically new from him or Microsoft.

Microsoft at least clearly says that the future is in modeling and not in coding and that it has to be focused around a repository. Which is EXACTLY what the ISIS Papyrus Business Information Platform is delivering today. Not a vision and not a strategy but a tightly integrated model-to-production application management platform.

Microsoft only plans to make SOA more accessible and useful to customers by adding SOA features and capabilities to its server, services, framework, tools and repository products to arrive at the 'Oslo' vision.

At the UNKNOWN AVAILABILITY of Visual Studio 10 Microsoft says that it will enable end-to-end application life-cycle management through NEW TOOLS for model-driven design of distributed applications using unified meta-data repositories. Various upcoming (no dates given!) versions of BizTalk (for BPM and SOA) and System Center (for versioning and deployment) will eventually utilize even a common repository.

Not only I am skeptical. Many analysts hail the Microsoft approach but call it non-standard and proprietary. I would not mind to be called proprietary if at least some analysts that I am trying to woo into looking at a REALLY INNOVATIVE products such as ISIS Papyrus would at least look!
I am not worried about Microsoft's potential future products being proprietary, but I know that all this smoke is just that, lulling everyone into waiting and away from truly new stuff.

You must be aware that all this will mean that it WILL BE TOTALLY INCOMPATIBLE in products and functionality to what you do today. Using the same product names Microsoft – as do most other large software vendors – will try to make you believe otherwise.

Everything that Microsoft announces as a radical new strategy we at ISIS Papyrus Software have been delivering to customers for a number of years now.

And then we run into analysts that claim that all the innovation we deliver fails the market as no one needs it.

Article Source:

About the Author

Max J. Pucher is the founder and current Chief Architect at ISIS Papyrus Software, a globally operating company that specializes in Artificial Intelligence for business process and communication. He has written several books, frequently speaks and writes on IT and holds several patents.

Republish Article: The Role of Enterprise Architects

The Role of Enterprise Architects

Author: Max J. Pucher

While I am not a typical Enterprise Architect as I am responsible for our products at ISIS Papyrus Software, I face the same problems as someone being in charge of Enterprise IT. I am in fact linked to EA architects problems with our solutions. Let me point out a few things that I see as problematic with current EA thinking and analyst stance related to the subjects in your email. These are essential architectural issues that are not enough covered and therefore I will also cover them in my upcoming book.

• Re: Effective planning to optimize shrinking IT budgets
Lets face it, IT projects have neither become simpler nor more successful. Rather the opposite. More than half of IT budgets are maintenance cost. More than half of projects still get canceled. More than half of project budget is manpower. And more than half of software cost is 'cost of sales' at the vendor. That leaves around five percent of IT budgets actually going into software development at vendors. With current technology being sold in current market conditions that is hard to change. No amount of EA planning will improve that situation. There is a huge amount of money going up in smoke. Why is anyone surprised? Things have to become way simpler!

• Re: Agility with SOA and BPM (Arrrrggh!)
Very soon this will make me want to scream. The reason business process management (BPM) implementations and service-oriented architecture (SOA) are white hot is that there is huge amount of money spent on marketing! Neither BPM nor SOA make a business more agile. SOA makes IT just a lot more complex (and thus rigid) and BPM is about long and expensive process analysis that is dated the day it is finished and outdated once implemented. Large scale BPM deployment is a failure where ever you look because no one has a handle on metadata, versioned deployment, and process interdependencies. The huge effort of putting SOA in could only make sense with large scale BPM and therefore these two are doomed together. The reason that everything is being outsourced is the immense cost of complexity. Obviously SOA life-cycle management is important, just like life-cycle management of eveything else. It makes however no sense to manage SOA and not manage BPM as an integrated element of the overall application. In all cases that I have seen putting in SOA and BPm meant substantial Java coding, that again is managed some place else. Forget agility …

• Re: The bottom line depends on new capabilities.
I could not agree more! IT is stuck in mud, soggy with programmer's tears. True innovation is the name of the game. Not product buying and renaming as most vendors do. We at ISiS Papyrus do actually deliver game-changing technology. Do enterprise architects care much? Most of them are in CYA mode and if at all they take some vendor from the 'Magic Quadrant' that is chosen from past market share analysis … not really the new stuff. Please, don't use Microsoft and innovation in the same sentence. They have long forgotten how to be innovative. They only know how to spend marketing billions.

How do we address thise issues at ISIS. What is so innovative about our Papyrus Platform?
Papyrus WebRepository promotes 'Experience Sharing' in the enterprise rather than rigid planning, and combined with the consolidated BPM, CRM and ECM capabilites in its own transaction and OR-DB engine it is a unique, innovative solution. We are only a medium size, privately held company, but have around 200 implementations in major organizations worldwide since its availability in 2001. The largest are 3000+ workflow users.
The ISIS Papyrus WebRepository has full life-cycle (DEV-TEST-PROD) project management with alerts and reports. It (project) manages all aspects (analysis, design and implementation) of the business process as well as the service registry, regardless of its implementation … SOA or not. While we can easily manage UDDI and WSDL files and their parameters and use in projects, but we get chosen in most cases because SOA is not mandatory. I yet have to see a company that has a readily available SOA interface that we can simply latch onto. A change in (SOA) interface metadata is propagated automatically to all processes and process elements, such as rules, user interfaces and created content! The user interface for the BPM implementation is much more powerful and extensive than Adobe Flex and runs in either Flash or the Papyrus Desktop with the same definition. The integrated UML capabilities of the WebRepository give it enough power for enterprise architecture planning. No Java, no .NET, no SOA, just process oriented applications that are simple and manageable!

Article Source:

About the Author

Max J. Pucher is the founder and current Chief Architect of ISIS Papyrus Software, a globally operating company that specializes in Artificial Intelligence for business process and communication. He has written several books, frequently speaks and writes on IT and holds several patents.