Hacker News Reader

For me, aiming for the shortest possible methods and classes forces me to more carefully consider good design principles like Single Responsibility. I don't automatically assume that long methods are wrong, but they certainly smell.

View on HN

Yes, but how long a method does it take to make you claim it is a smell? One of Steve McConnell's books lists a couple of studies that suggests optimal method length between 100 to 250 lines of code.

Smelliness generally increases proportional to length, simply by virtue of the fact that it takes longer for my human brain to parse. A one line function has no inherent smell because usually I can tell immediately if there's anything else smelly about it. A 250 line function smells a bit more since there are more places for more insidious smells to hide. Very short methods tend to be very easy to reason about.

What properties of 100-250-line methods make them optimal?

Number of bugs per method.

How long ago were the studies made? One thing that's changed since then is the disappearance of continous form printing. Decades ago, you could print a 250 line function on four connected pages and lay them down on a table and see the entire function. Nowadays, reading printouts like that is unheard of.

Reply on news.ycombinator.com