How to Refactor Messy Apps for a New Architecture?

When facing a legacy code base, changing the mess to an orderly architecture can cause confusion: where to start? What to do? An exemplary question:

How do I refactor a Big Ball of Mud into layered architecture?

Refactoring1 the existing code base seems like a logical step; after all, refactorings are designed to improve existing code, and improvement is what you’re up to.

But:

My advice:

Create a new model of the application. Start from scratch, and start with drawings on paper. This is an exercise to understand the problem and solution space in a new way. Then you may decide if your legacy code can be re-used to fit the better architecture or if you have to rewrite parts of it. If you look at the old stuff all the time, you can only force yourself to make incremental improvements, not re-designs.

It helps to factor the new model into modules and submodules in advance to start somewhere. Instead of stopping at 4 layers in your app with a bunch of objects in them, identify submodules and communication paths, too, so you can hunt for a small-ish part of the app that you can improve.

To learn how to apply a particular architecture, it’s better to start with a greenfield app or a toy project so you can get a hold of the concepts first without worrying about breaking a shipping app. With experience, it’ll be easier to identify new seams in old code bases and change things for the better.

  1. This is only a refactoring on the app level, as in “improving its code,” but not in the usual meaning: “Refactoring is a disciplined technique for restructuring an existing body of code, altering its internal structure without changing its external behavior.” (http://www.refactoring.com/) I’ll go with the colloquial use of the word here anyway. 

Browse the blog archive