Today I participated in the IDC Conference about Service Oriented Architecture in Copenhagen. It was a conference with good and giving presentations and I can use much of it as input to our SOA Analysis in Økonomistyrelsen.

The main conclusions where not new: SOA is very complex and though there are challenges with the technologies and methods, the main challenges are in governance and people. The issue about complexity was well addressed by Rob Hailstone from IDC: SOA has been oversold according to simplicity. The architecture in SOA is simple, but the dynamics SOA gives, makes SOA complex, which again explain the need for governance. The presentation of Lennart Klamer from Capgemini was titled "To implementing SOA is more about governance and people and less about technology". The mindset must be changed to thinking on others (thinking global), to be service focused, thinking cross organisational and focus on constant learning. The matureness of the stakeholders in the organisations is different and we Enterprise - and IT-Architects must work hard on talking with the business people. This was also confirmed by Rob Hailstone and a IDC analysis have concluded that the business have least awareness about SOA, though the business should be the driver for SOA. – The analysis shows that the top priority driver for SOA is to better respond to changing business challenges. Other example of the human aspects is that, the developers must for example instead of being systemorganized be serviceorganized, which changes the responsibilities according to Niels Feldthus from DSB.

Andre Rogaczewski from Netcompany started his presentation with a questionnaire to the about 450 conference guests: First: How many are investing in SOA? About 50% of the guests responded yes! Secondly: How many have gained business benefits? Out of the 450 guests only one said yes! The main barrier is the lack of budgets and that the business benefits of SOA are hardly tangible and there are not many finished SOA- projects. There for have we IT-people again a main responsibility of telling the business about the business benefits. His presentation was titled: Stop the academically and theorytical thinking and focus on gaining business benefits. For this he presented following three principles:

- One language - one platform. SOA is about standardization and require a infrastructure platform early in the process and work with common definitions of concepts and data.
- System independent integration requires a organisation, where the developer team corporate closely with the architecture team. The architecture team must not been a "junta", but a partner and reviewer and both teams are responsible for the success.
- Focus on a stepwise strategy and start with a pilot project, so you can achieve early business benefits.

There was presentation of practical cases from Energinet.dk, DSB, 3F and Copenhagen Amt. The main point was again the cultural change. They agreed in that it is not a good feeling´and situation of being some of the first-movers in SOA (in Denmark), but most of them havent any other choices. Some of them can feel the benefits of being more agile, less dependent on single vendors etc., but the all agreed in that it was a investment on the long term. They all agreed in taking a practical approach to SOA (inclusive security) and thinking production already at the start of the projects.

Today I participated in a conference arranged by Qualiware about EA, management information, quality management and risk management with focus on their products. There was a lot of interesting presentations about EA from the CEO of Qualiware and practical implementations at ATP and the Norwegian Post.

I got a lot of input and ideas, which I can use in our project about EA Documentation Framework and EA Analysis Tool.

Qualiware focuses on modelling and compare EA with metamodelling (metamodel is a model which describes a model). The task of metamodelling is a good and value giving activity, which for example find gaps, use of duplicate resources and give the organisation a common language. A part of the metamodelling at ATP was to define business benefits and products, business concepts, Use Cases and services and mapping them with each other.

According to the CEO Qualiware do not sell their EA-product as a drawing tool, but that it supports 8 EA-disciplines like metamodelling, assessments, change management etc.

One of their new products is a projectportofolio management system, which is integrated with the EA-product. Their product for risk management is also integrated to the EA-product.

All of the EA-Frameworks was inspired of Zachman and customized according to the organisations specific needs. Some of them worked with up to 5 detail levels and some had extra aspects/perspectives for stakeholders, organisation etc.

Modelling of especially business processes and workflows is not an easy task and is very time consuming. More over must the organisation be mature, ready for business modelling and must think and act cross organisationally and not silo oriented. In our organisation there are no plans of detailed business modelling in the near future and it is off course not a task of the it-department, but must be done by business people from the business departments.

I participated in a network-meeting at Økonomistyrelsen about OIOXML (The Danish XML-project) with presentations from Jan Braun from the National IT and Telecom Agency (IT- og telestyrelsen) and Rene Løhde from Microsoft. It was an interesting meeting and I will mention some of the main points.

New data standardising-initiative in the sectors

The National IT and Telecom Agency have until now most focus on semantic and technical interoperability and they will in the future have more focus on process and organisational interoperability. The new data standardising-initiative in the sectors is based on the 4 focus areas, which are described here.

Domain language and OIOXML

Rene Løhde talked about semantic as a domain language in a sector. The domain language is often understand by a couple of persons in the domain-area or/and from the vendors. It makes the organisations vulnerable and he introduced the "Bus-factor", which is the dependence of the knowledge and experiences from few people. There for must the organizations ask them self: "What if the person is driven down by a bus tomorrow?". Semantic knowledge can not bee bought from outside the organisation as the case in f.eks. technical knowledge. One solution for reducing the Bus-factor in the area of semantic/domain language is to find the most important concepts in the business-area and define them by making a conceptual model based on OIOXML-schemas. One of the EA-projects in at Økonomistyrelsen is the Common basic datamodel (Faelles stamdatamodel) across the organisation. The 11 most important and common concepts are defined and the background (and the BusinessCase) for the project is that the same data are updated by the users in different systems. The further process is the implement the common concepts in the systems after defining with OIOXML, so the same data don´t have to be updated in different systems.

UML and DSL

UML is a standard for modelling of f.eks. concepts and is recommended by the National IT and Telecom Agencys Interoperability Framework (OIO-kataloget). Microsoft agrees in UML, but never used it them self, because the long distance from the modelling by UML and the programming of code. There for Microsoft have introduced the DSL (Domain Specification Language) and with the DSL-tools you can create your own designer for a visual domain-specific language. The tools help you define the domain-specific language and generate source code and other files. However DSL-Tools requires Microsoft Visual Studio 2005 Professional Edition or higher and you can try out a virtual lab session with the DSL Tools in a hosted environment at MSDN.

Since SOA is one of my greatest interests and SOA is a concept with ongoing development, It is interesting to follow the discussions of the relationships between SOA and Web 2.0.

The burst of the dot-com bubble in 2001 marked a turning point for the web and many people concluded that the web was overhyped and dead. More positive people like OReilly Media and MediaLive International said that bubble-bursting was a common feature of all technical revolutions and it marks the point at which an ascendant technology is ready to take its place at center stage. In this center stage, the real success stories will show their strength, and there begins to be an understanding of what separates one from the other. From these viewpoints the concept of Web 2.0 began with conference brainstormings session between the two organizations and the concept is still being discussed at the Web 2.0 Conferences.

Web 2.0 can be visualized as a set of principles and practices that are based on the Strategy of "The Web As Platform" and the core competencies of Services, Architecture of Participation, Cost Effectively Scalability etc. The whole setup is mapped in the mindmap, which u can see here.

According to Dion Hinchcliffe are some of the similarities between the two concepts:

- Both are slippery and nebulous concepts with many different definitions
- Both encourages the liberation of the underlying functionality of software systems as services by providing open access to everyone that need it
- Both provide the building blocks for creating people-centric processes

The main difference is that most SOAs are conceptually inside an organization's firewall or VPN while Web 2.0 envisions the global Web as scope for building functionality. There for must the organizations SOAs get out of the organizations boundaries so Web 2.0 can be conceptualized as a global SOA. More over can the organizations working with SOA as their architectural model connect it to their Web/Web 2.0.

Conclusion: There are synergies between the two concepts. Other synergies is that, SOA has much central control, management, and governance while Web 2.0 is autonomious, decentralized, grassroots, and with absolutely no command and control structure. Web 2.0 also talks about presentation or GUIs, while SOA is largely silent on this issue of presentation. So SOA tends to be generic and faceless where Web 2.0 have more human/service focus.

This was about web 2.0. Some IT-Architects and SOA-evangilists talk about 3. generation internet and semantic web and the connection to SOA. - Do anyone know anything about these concepts?