Posts Tagged ‘User Experience’

1st international SOA Symposium

Yesterday the first international soa symposium started. The location is very interesting: the Ajax soccer stadium. The line up is great: Thomas Erl, David Chappell, and Dirk Krafzig are among the speakers. The content is very good. For the first time in a long time, I had trouble choosing the right track because I would like to follow all of them!
It was also the first time we (Approach) were sponsoring an event. We have a great spot for our booth. It is a nice mix of tool vendors, consultancies and customers at the conference.
The keynote yesterday was by Thomas Erl. He addressed three groups: He appealed to vendors to stick to standards, to the Industry to make standards understandable and to practitioners to make a distinction between different disciplines of SOA.

We are not reorganizing the universe just because we want to do SOA
The first session I attended was by Paul Brown, a well known author from Tibco. The presentation was well organized, and he made relevant observations. The thing that stood out for me, was that he approached the problem of SOA and BPM from the business. He talked about the current silos and stated that these would not go away, just because we are doing soa, or as he eloquently put it “we are not reorganizing the universe just because we want to do SOA”

Another track I was interested in, was the SOA and Web 2.0 track. I attended two sessions. (For some reason, the first one was in the track “SOA Industry”). It was presented by Edwin Sanders and he talked about user interface services. The concept and architecture was interesting, his approach lacked any acknowledgement of common user experience principles. Either the product (corizon) lacks this vision, or the presentation was just unclear. I hope for them, it is the latter…

The presentation by the Burton Group about user experience was very focussed on User Experience. Anne talked about persona’s, Donald Norman etc. Unfortunately the relevance for SOA was explained very clearly. I believe there is a strong link, hence our user exerience service….

In general the sessions and content is very interesting. It would be nice if the program would mention the company the speakers represent. This is very relevant when you select the sessions that you want to attend.

I am sitting in our booth, typing this. I am looking forward to another interesting day. For me this is already a success and I hope that this conference will be continued next year.

Mike in the Approach Booth

User experience can’t be generated

Lucas Jellema posted a very interesting blog on the AMIS site recently about a topic that I love to argue about .

His claim is, and I quote: “My claim in this article is that there are very few ADF projects that will not benefit from using JHeadstart.” I disagree vehemently, I think there are several reasons why you don’t want to use it , both from an IT perspective and from a business perspective.
Let’s start with the IT perspective:

• you become dependent on the JHeadstart (consulting) team for version upgrades, they tend to lag a little behind. This might not be a big issue, but it is something that you should at least be aware of.
• Generating does not encourage separation of concerns: the same code can be generated into different pages and is not refactored by a developer (because it is generated). When you want to change this code, that is repeated everywhere in the application, you need to change it by hand everywhere or regenerate the page and lose post generation edits.
• you can not reuse generic JHeadstart components in other applications than in JHeadstart applications.
• It is designed to generate (relational) data driven applications very fast. In a SOA environment data is not in the form of relational tables and rows, but in the format of hierarchical XML data and objects.
• You edit properties not directly in the screen (direct manipulation) but indirectly in a property editor. It takes some (a lot?) of learning to estimate the effect of an edit.

However, this is not as important as the business reason.
An application that is used to bind customers to your company (marketing sites), saves operational costs by applying self service, or that is used in a mission critical situation (hospital environment, military type) has totally different requirements. The user interface can add real value to your application and organization in this case. It should be designed independent of the database or any other technical implementation detail! See for a good explanation of this concept this book.

In a SOA environment this becomes even more important. Because you have more information from different sources that you can combine (services), it becomes very important to think about the experience you want to give your users. If they don’t understand it, or they get lost, you can lose business.

Finally, software development costs are not the biggest IT costs in most businesses. Losing potential customers to competitors, maintenance of existing applications, license and hardware costs, correcting technical and business errors (for example incorrect orders), staffing and training the helpdesks etc, etc are a much bigger issue. Saving on your development by generating the most visible part can be a an expensive mistake.

There are different types of applications, of course. An application that is build to maintain some reference data in a database can easily be generated by JHeadstart or any other data driven tool. The question is, does this really save you any money?

My claim is: in most cases it will not