SSoT.me – the ‘Single Source of Truth’ Toolbox.
They say, “Fast, Good, Cheap… Pick 2” – but SSoT.me let’s you pick all three.
Creating software based on a Single Source of Truth allows you to do so in a fraction of the time time vs a “Traditional” development model. This speed translates into significantly reducing the overall cost of a project.
As a side affect benefit, this approach also results in dramatically better code along a variety of dimensions.
Us Humans
First and foremost though, Derivative Code can always “explain itself” in Natural Language English. This is the most important layer for all of us Humans!
This “Derivate Documentation” always (and this is important) matches “the tech” which it is describing, because is are is derived. Because is is not “source code”.
Derivative Code
“Derivative Code” is code derived from a Single Source of Truth or SSoT. When compared to code not developed with an authoritative SSoT, Derivative Code tends to be more:
- Complete
- Stable
- Flexible
- Durable
- Understandible
- Internally-consistent
- And Responsive
The reason that the Derived Documentation always matches the tech it’s describing is that neither it, nore the derived-tech it is is describing is the “Source” description. Instead, both are “derivative code” – both from one, central, authoritative single source of truth. When the SSoT changes – both the documentation and the code change, because they are are derived, because they are not “source code”.
“Source Code” or “Hand Code”
“Source code” is expensive. “Source Code” has to be maintained by hand. Human Hands. And Human Hands are expensive.
The goal of ever project should be to reduce the total amount of source code, since 1,000,000 lines of source code will inevitably cost more to create, maintain, upgrade, manage, etc than a 1,000 line app all else being equal.
And one way to dramatically reduce the amount of “source code” which you have to manage is to turn 50, 60, even 90% or more of it into “Derivative Code”. What that means is that when the SSoT changes – 80% of the code in the project automatically updates itself to match the new design.
To be clear, the “source code” would still have to be updated “by hand”. So the remaining 20% of the changes needed would still need to be done by human beings – but hopefully “the interesting bits” in most cases. In other words….
What is the “Low Hanging Fruit?”
The 80% that is most likely to be turned into derived code first, is “the plumbing”. The … well … “derivative” … parts. i.e. The parts of the code that are not too interesting, and are primarily following a well established pattern. This includes any code which, if pointed to, is 1-of-n. Are there 2? Maybe it’s derivative. Are there 3? Or 10? Definitely derivative.
What we’ve found is that most developers are surprised at how much of the code in their project can be isolated and turned into derivative code.
