Object Oriented Programming: An Evolutionary Approach

This post is not about the book Object Oriented Programming: An Evolutionary Approach by Brad Cox. I decided to use the book’s title because the author nailed the connection between software and evolution. It is a good book, by the way. I recommend it.

Last week a coworker sent me a link to the React.js Conf 2015 Keynote video in which they introduced React Native. Because I work on NativeScript, I was curious to see how Facebook solves similar problems as we do. So, I finally got some time and watched the video. The presentation is short and probably the most important slide is the following one.

ReactNativeSlide

But this blog post is not about React Native. The thing that triggered me to write is something that the speaker said (the transcription is mine).

This is a component. This, we feel, is the proper separation of concerns for applications.

I couldn’t agree more. Components have been around for many years (don’t get me wrong, I am not bringing up one of those everything new is well-forgotten old themes). Yet, components and component-based development are not as widely accepted as I think they should be. It is probably because so many people/companies saw a value in software components and started defining/building them as they think it is the right way. And this process brought all the confusion what a component really is.

Beside the fact that the term component is quite overloaded it is important to note that many had tried to (re)define it in different times and environments/contexts. Nevertheless, some properties of software components were defined in exactly the same manner during the 70’s, 80’s, 90’s and later. Let’s see what these properties are.

  • binary standard/compatibility
  • separation of interface and implementation
  • language agnostic

These are some of the fundamental properties of any software component. More recent component definitions include properties like:

  • versioning
  • transaction support
  • security
  • etc.

But this blog post is not about software components either. It is about software evolution. We can define software evolution as a variation in software over the time. This definition is not complete but it is good enough. It can be applied on different levels, whether it is the software industry as a whole or a small application. It is important to say that software evolution is a result of our understanding about software, including an exact knowledge, culture and beliefs, at any point of time. We can reuse the analogy of terms like mutation, crossover, hybrid and so on to describe processes in software evolution.

Combining different ideas is one of the primary factor for software evolution. And this is where we need software components. There is a common comparison between software components and LEGO blocks. The analogy of software genes might be another alternative. An application DNA is defined by its genes.

Software evolution is not a linear process. Do you remember Twitter’s dance between client-side and server-side rendering? It is a great example of survival of the fittest principle. So, what will be the next thing in software evolution? I don’t think anybody know the answer. So far, the software components seem to be a practical way to go. Seeing big companies like Facebook to emphasize on composability is a good sign.

The best way to predict your future is to create it.
– Abraham Lincoln