Yesterday I participated in IBM Softwareday 2006. The theme was "What makes you special?". The background of the theme was that many executives focus more and more on innovation and that innovation gives the real business benefits. After the opening by IBMs Sales Executive there was an inspiring speech by Mads Oevring, who told about the relevance of innovation and innovation in Novo. The Novo case was primary focused on importance of that their employees learn new things before their competitors in the medical industry and that more than 60% of knowledge comes from outside their organization.

After the initial speaking there were six trails, you could follow the rest of the day and I followed the SOA-trail, which was the most popular. The relation between innovation and SOA is of course that innovation requires readiness for change and SOA can make it easy to make changes.

The SOA-trail was introduced by a manager from IBM with focus on Entry points for SOA: How the organizations can start on SOA. The other speaks were among others from Danske Bank and The Danish Medicine Agency (Laegemiddelstyrelsen).

The IBM-manager presented their SOA Reference Architecture, which is much business oriented. The reference architecture was mapped with five categories of entry points, which were:

- People: Interaction Services that enables collaboration between people, process and information
- Process: Process Services that orchestrate and automate business processes
- Information: Information Services that manages diverse data and content in a unified manner
- Connectivity: Facilitates communication between services with ESB
- Reuse: Partner Services that connect with trading partners.

There was also focus on SOA-governance and the challenges. The rest of the speech was about IBM´s services and products.

The CIO from Danske Bank connected their "One Bank - One System"-vision with SOA and one platform for integrating processes and it. This one platform and the whole setup is a condition for their expansion and taking over new banks (fast integration of new banks national and international) - When you buy up a bank, you have to offer them something and one the main offers are Danske Banks architecture and infrastructure. The background of all this is of course efficiency by reuse in the long term, which gives Danske Bank differentiation and opportunities of innovation. This SOA-consolidation strategy results in better and better ROI when taking over new banks. Half of all their functionality is serviceenabled and whole their Netbanking-system is based on services.

The SOA-approach of Danish Medical Agency was rather technical and Bottomup-focused. Their SOA-approach was titled "From spaghetti to Lasagna" (and not SOA-sauce because of lack of Governance and other large challenges in SOA...). Their background for SOA was large non-documented solutions, redundancy in data and functionality, different systems for security, need for integrating silooriented systems and non-standard technology and tools. The first initiative was implementation of a new integrationplatform and a pilotproject. The triggers of the integration platform was need of standards, easy integration, documentation, reuse, less dependence of key persons and more competition between the vendors. The pilot project was a Price-Extranet-solution, where the medical firms now can report their prices online, which was done manual and by faxes before.

The slides from the SOA-trail can be downloaded here.

The National IT and Telecom Agency (IT- og telestyrelsen) is working on common SOA-principles for the public sector. The draft can be downloaded here The agency has formulated 11 SOA-principles in six categories:

- Business supporting
- Documented and communicated
- Flexible
- Based on open standards and contracts
- Managed
- Secure

The background of the initiative is that many in the public sector feel SOA as intangible and it is difficult to get started. One of the solutions on these problems is common SOA-principles for the public sector. The aim is to make a publication after submitting to the OIO-Architecture committee and eventually public hearing. In this context the Agency held a workshop, which I participated in. The purpose of the workshop was to discus business goals and principles of SOA. It was a very interesting workshop with architects from public and private organizations, much good discussion and group work. The outcome of the SOA goals was that it of course depends on the perspective and that goals have different abstraction levels. The generic goal of SOA in e-government was concluded to be efficiency, better service, agility and increased competition among vendors, especially on large tight coupled systems.

The outcome of the discussion of the principles was that the principles were very good, but there were not any principles regarding governance and EA. More over is security a universally concern and challenge in all architectures and must maybe expelled from the principles.

This work is also interesting regarding my thesis. The principles are almost the same as the SOA- characteristics I have analyzed.

One of our EA-projects is migration of 15 systems to. NET. The 15 systems are based on VANG hardware and operative system and the applications are based on Focus. Yes, I had never heard about VANG before and try to google it:-) VANG is a expired technology and according to my boss, they were the first with Word processing and e-mail and some of the guys from VANG made WordPerfect.

Focus is still supported. The vendor is http://informationbuilders.com/ and we use WebFocus and Focus (Host-based) and the products are marketed as "The Standard for Business Intelligence". As mentioned the products are still supported and they are rather innovative and based on standards. So what is the problem? The problem is the Focus-competences in Denmark. In Demark there are about 6-7 Focus-developers and the 4 of them are employed at us! – And according to our policies, we must outsource most development activities as possible and not be depended on rare competences.

So there for, must the 15 systems be migrated and we are working on specifying the requirements.

By Asim Hanif at asimblogged.com

As mentioned here and here, one of our EA projects is "Common methods and templates for requirement specification".

In our organization we have chosen Use Cases for identifying, structuring and documenting functional requirements. I am convinced of the strength of Use Cases and have used it a lot at my former working place. The biggest strength is the focus on the user, because Use cases encapsulate functional requirements in the Users cases and working situations. After the requirement process, you can use the Use cases in the design-phase, test-phase and making of user manuals and online help, while u have 100% focus on the users and requirements in the rest of the project!

I have evangilize Use cases to our business departments (I am supporter of Event-driven Use cases) and after explaining the concept for the business people, they can easily understand it. In next month we will use Use Cases to specify three systems, which are based on VANG and Focus and must be modernizied to MS .NET.

It is important to realize that working with Use cases is an iterative process, where the Use cases are becoming more general and can there for be reused by more actors. More over is more and more business concepts identified until you almost have a complete business concept model, that in the design phase can me transformed to a datamodel.

As mentioned here, one of our EA-projects is "Common methods and templates for requirement specification". The project was initiated by our GAP-analysis and the main background is to mature our organization in requirements. The initiative is perfect for incorporating our EA-principles, reference architecture, standards etc. in practical projects. I am also working with implementation of a common process for system development and -acquisition based on Unified Process (UP), which is much centered at Use cases and Architecture. Though this kind of architecture is focusing on a lower level than EA, we have made good guidelines and processes for connecting our EA to architecture requirements and non-functional requirements in requirement specifications.

In context of requirement specification I participated in a network group meeting arranged by Dansk IT. It was a very giving meeting based on a case and with a lot of good discussion. Dr. Soren Lauesen also participated in the meeting and initiated a lot of good discussions.

Some of the new points for me, are:

- The purpose of the specified business unit (and system) is very important. The purpose can be defined as "A short and clear description of how the project will change our business situation" or where the project will move the specified business unit. In this context all the functional requirements must be mapped against the purpose. You should ask, if and how each functional requirement (encapsulated in Use cases) is supporting the purpose or not. If not, maybe is the requirement not relevant. So is this process very important in scoping projects and managing the Cost/benefit for the clusters of functional requirements.
- Background for the system is also very important and can be defined as a "Short and clear description of why it is important to start the new project and which consequences an unchanged as-situation will have". With focus on this and with good governance many projects will still be ideas and only real business improvement projects will bee implemented.
- The main elements of a requirement specification are the business functionality aka functional requirements for a system. This is according to UP-best practices and in most successful practical cases done by Use cases and Business Concepts (-domains – Danish: Forretningsbegreber). This is done by iterative working with Use cases and Business Concepts, by keeping Use cases against the Business Concepts and asking, which Business Concepts do the Use cases influence and more detailed which Business Concepts to the Use cases create, read, update and delete? This can be done with a CRUD-matrix by keeping all the Use cases against the Business Concepts. This will result in a complete Business Concept model, which can be mapped to a datamodel or ER-diagram in the design-phase. In this way you can be sure to have specified all the requirements.
- The typical Use cases in requirement specifications are system-focused. For getting more innovation and be more business focused, there is need for something before the systemoriented Use cases. This can be done by either Business Use cases or Business Process Modeling (BPM). I think this aspect is very interesting and I will further work with this and clarify the transition from generic holistic across organizational Business Processes to more Systemoriented Usecases.
- Sketches of GUI are very good in communication Use cases with end users and can be used in the requirement process, but the sketches may not be distributed to the vendors, because it will reduce their innovation. We have experienced, that some vendors are seeing sketches as design and they implement the sketches in design by 1:1:- )

I have now blogged in a half year. Much is happen since I started. I hope my English is better now. My hit statistic is constantly increasing. In august I had totally 1245 visits and 16.157 hits. Searching "Asim Hanif" OR "asimblogged.com" on Google gives 14.100 hits. Unfortunenally is my name not 100% unique, so the 14.1000 hits are also referring to others, but it still shows how famous I and my blog is in cyber.

The subjects I have blogged about have been ranging between EA, SOA , WS, requirements and system development. Unfortunenally I have not been able to write much about SOA in context of my thesis. There for I will try to have more focus on this.

I am happy to see that I have a lot of international visitors. The statistics shows with decreasing order: US, Europe, Denmark, Netherland, Argentina, China, Panama, UK, Spain etc.

I see the main challenge in blogging to get feedback and comments from the visitors. I also suffer with this challenge, but I dont think u can do so much other than personal, serious and relevant blogging and marketing your blog.

On this day I also will thank Dr. John Gotze for getting me into blogsphere and inspiring me.

Regards and thank you

Asim Hanif:-)