Maintainability
Why ?
Because software is an always-running evolving solution to still-unknown and changing problems
What ?
Maintainability is inversely proportional to the amount of time it takes to make a change and the risk that change will break something
How ?
-
design
-
documentation
-
tests
-
process
-
architecture
Maintainable architecture
Limit brain RAM needed
Contracts
-
explicit responsibility
-
stable interface
-
deliberate evolution
3 principles
-
functional design
-
coupling as hierarchy
-
domain abstraction layer
1. Functional design
The problem with object-oriented languages is they’ve got all this implicit environment that they carry around with them. You wanted a banana but what you got was a gorilla holding the banana and the entire jungle.
Joe Armstrong
Show me your code and conceal your data structures, and I shall continue to be mystified. Show me your data structures, and I won't usually need your code; it'll be obvious.
Fred Brooks
Values & Functions
-
no hidden context
-
controllable side effects
-
clear encapsulation
-
easy composition
2. Coupling as hierarchy
-
parent
-
children
Only talk to your
-
module interface
-
re-exports
-
wrapped delegation
How ?
class Engine::Docs
def self.default_list(project)
if Engine.early?(project)
Engine::Docs::EarlyList.build(project)
else
Engine::Docs::List.build(project)
end
end
endclass Engine
def self.docs(project)
Docs.default_list(project).map(&:slug)
end
def self.early?(project)
project.maturity.in? Engine::Project::EARLY_MATURITIES
end
endImmediately identify impacts of change
3. Domain abstraction layer
Less code is better
int countSubstring(const char *str, const char *sub)
{
int length = strlen(sub);
if (length == 0) return 0;
int count = 0;
for (str = strstr(str, sub); str; str = strstr(str + length, sub))
++count;
return count;
}def countSubstrings(str, sub)
str.scan(sub).length
endGlobal coupling
Stability
The domain is the most stable candidate in the face of changing requirements
Questions ?
Maintainability