The map, in numbers
Thirty-six engineering patterns, distilled from seven repositories. This page reads the shape of that graph so you can decide what to study first, what to reuse, and where to dig next. Five patterns carry most of the structure; eight domains hold only one.
The backbone
Five patterns hold the whole map together
Connection count is influence: a pattern wired to many others shapes how the rest fit. These are the hubs. Read them first and the surrounding patterns fall into place faster.
The foundations
The highest-leverage code to reuse
Counting only "depends-on" arrows pointing in shows what other patterns are built on top of. These are the load-bearing abstractions. Build or borrow these first and everything downstream gets easier.
Coverage
Deep in a few domains, thin in many
How many patterns sit in each domain. The single-pattern domains are not weaknesses, they are the obvious places to distill more.
Sources
One repository supplied a third of everything
Where the patterns came from. Repos near the bottom are barely mined, more value is likely still in them.
How it connects
The graph is mostly "x builds on y"
Every connection has a type. The mix tells you what kind of knowledge this is: a web of dependencies and shared lineage more than a list of alternatives.
Compare and decide
Where different projects solved the same problem
These pairs are independent answers to the same question, linked as "similar pattern" or "alternative to". When you hit that problem yourself, these are the approaches to weigh against each other.
Where to start
Using this to make a call
Orient with the backbone
Start from the five hub patterns above. Understanding them gives you the most context per pattern read.
Reuse the foundations
If you are building, lift the most depended-upon abstractions first. They unlock the patterns stacked on them.
Compare before you commit
Facing a problem with two answers in the convergent list? Open both in the map and read their build specs side by side.
Expand the thin domains
The single-pattern domains are the clearest gaps. Point distill-it at a new repo in one of them to deepen coverage.