[koboll]

The key determinant, imo, is precisely and comprehensively descriptive names

Even if this leads to humourously Java-esque code, it's worth it

An abstraction that has one purpose and is named to make that purpose crystal clear is a good abstraction

[lukeramsden]

> Even if this leads to humourously Java-esque code, it's worth it

Some people seem to get mad about the verbose naming common in Java - but it’s one of the biggest blessings I’ve ever experienced. If you name things after what they do, and that name is stupid, then it’s the quickest indicator of bad design I’ve ever seen. Good design is where every name is patently obvious and encompasses the entire purpose of the class / record / method / whatever.

[TeMPOraL]

I think it's not the verbose names that are the problem in stereotypical Java-esque names, but rather the large amount of "scaffolding" classes that implement design patterns - a wealth of middleware that becomes apparent thanks to the verbose naming. That is, it's not the ProcessedWidget::QueryStructuralIntegrityPercent() that's the issue - it's the associated WidgetFactory, ProcessedWidgetBuilder, ProcessedWidgetProcessingManagerDelegateProxy, etc. that bloat the code and tax cognitive memory, and which exist only because the language isn't (or used to not be) expressive enough to express those patterns without giving them a name.

[peterashford]

There's nothing like ProcessedWidgetProcessingManagerDelegateProxy in Java or the standard library. I feel that Java gets whacked with a cudgel that should be aimed at middleware authors.

[TeMPOraL]

There's nothing in JavaScript forcing you to use 500 dependencies across 100 000 folders either. Opinions on a programming language are often, for better or worse, opinions on the language plus its community/ecosystem. And correct me if I'm wrong, but isn't the company responsible for maintaining Java (Sun, now Oracle) also one of the worst offenders in terms of naming things in its middleware?

[eyelidlessness]

> If you name things after what they do, and that name is stupid, then it’s the quickest indicator of bad design I’ve ever seen.

I agree it’s a smell, but I’d caution that it’s also very situational.

Sometimes the domain has a recognizable category of like-things which don’t have a name for their likeness within the same domain, and choosing even a stupid-sounding name is a good start towards choosing a better name.

Other times, a targeted portion of an existing codebase might have a common theme which doesn’t necessarily line up with the domain, but needs to be named something to untangle what outlived its WET shelf life. You can recognize it’s a stable abstraction, you can recognize it's entirely divorced from any domain concern except by happenstance. You just have to give it a name, or let it be an increasingly unnecessary burden.

It’s easier to take such a hard line on the design quality implications of naming when you have a vocabulary to reference and/or compose, or when you have extant design attitudes relatively aligned with that principle. It’s especially difficult to apply the principle in systems with extant design problems of this nature, because whatever established or emergent abstractions do exist might not align with any principle you’d apply, and you can’t move them in that direction without some intermediate step no matter how flawed. You have to name what is before you name what should be.

[rightbyte]

> If you name things after what they do, and that name is stupid, then it’s the quickest indicator of bad design I’ve ever seen.

As a reader of the code I think 'FooAndMaybeBar_SpecialBazSupportV2 ()' is a blessing.

Aiming for cleaness when the problem is not clean is a worse sin than messy code that could be clean.

[throwuwu]

Doesn’t matter how good your names are if your control flow is incomprehensible or if the data representation is some bloated or mangled monstrosity.

[i_am_a_peasant]

and every time i complain about names during code reviews i get accused of bike shedding