04 October 2024
Why fiddle with models if you can actually know…?
3 min read
Right after a big conference, I always try to put my mind at rest by putting my thoughts on paper. Coming back from the Gartner IT Xpo Symposium in Orlando, I had quite a few hours on the plane, and with my 1m86, I only have a few hours of sleep. Since I usually spend my sleeping hours like a mummy, the confined shape of the seat doesn’t help either, I guess. So an all-night flight gives my mind enough time to keep tossing and turning and generating questions about what I experienced.
One of the bigger questions that kept on popping up was most probably triggered by the many enterprise architecture booths that were present on the IT Xpo floor. And it is not any different from what many of us are already wondering: Why do architects love modeling so much?
Don’t get me wrong, I like models as much as anyone. Heck, even more I would say, we have based much of our business around them. Models are great. Once you realize that they are a simplification of reality, they become very useful and powerful. You just have to be mindful to keep some rules in mind when working with them, to avoid assigning them truth beyond their meaning.
However, a model does aim to reflect the practical world around us. The whole idea of a model is that it allows us to abstract from some of the complexities around us that make it difficult to focus on the core information and/or insights we’re looking for. So the aim is to make life easier to understand while retaining as much information from reality as possible.
And this is where my surprise comes in.
While many architecture tools use APIs to pull information from the farthest corners of the organization, they all seem to ignore the connection to what really matters: the source code.
Because when it comes to building applications, there is one dimension of architecture that matters more than any other; that of the actual setup and build of the application. The connections between the various components and the calls throughout the application itself.
Yes, the other categories of meta-data and information do matter. Who owns the application? Where does the budget come from? Who works on it? Who are the users? And many tools try to gather this information together and then enable the ‘modeling of the application itself’. Information that architects must input to transform the current state into a new, better and desired state.
But without an actual link to the source code, these are just derived guesses as to what reality looks like. And that’s not necessary. Because the information can be obtained. Instead of using modeled versions of reality, architects should be empowered to work with the actual representation of reality: the architectural view as composed from the source code. That which is actually there.
So I wish a better world for all the architects in our midst. Don’t take for granted the current world in which you have to fiddle around with boxes and flow charts to represent what your architecture looks like. Instead, demand a view of reality. Because only when you know what is actually out there, you can shape the world to what you want it to be.
Let’s keep in touch
We'll keep you posted on the latest news, events, and publications.