How to reduce the cognitive load of your code

Re:How to reduce the cognitive load of your code by Retric on Wednesday March 30, 2016 @09:57AM (#11389276)

If your stripping boilerplate then your operating on different things. Which seems more readable?

  SortGuiElement(){junk; SortUserNames(); junk;}
  SortUserNames(){junk;}

vs.

  SortNames(){junk; sortNames() junk;}
  sortNames(){junk;}

It might seem obvious and the code might be identical, but I see the second case very frequently in other peoples code.

[ Parent | on HN ]

Re:How to reduce the cognitive load of your code by kazinator on Wednesday March 30, 2016 @11:36AM (#11390057)

Re-using an identifier with just a case difference is criminal.

You want:

  SortNames(){junk; SortNamesImpl(); junk;}
  sortNames(){junk;}

or whatever: SortNamesGuts(), InternalSortNames(), DoSortNames(), LowLevelSortNames(). Anything but just flipping the case of a letter or two in the "SortNames" identifier.

[ Parent | on HN ]

Re:How to reduce the cognitive load of your code by Retric on Wednesday March 30, 2016 @12:13PM (#11390335)

I agree that reusing the same name with different capitalization is bad, but I have seen this from several developers across many different teams and several languages. These same people often have Foo(x){junk; Foo(x,y,z); more junk;} sure if it's a pass though that's fine but if you have 10 foo('s) that all have internal logic then please come up with some actual names.

Amway IMO, SortNames() and Internal/Do/LowLevel/SortNames() are almost as bad because just looking at the names it's not obvious what's going on. Yes, visually they look different which helps and consistency can make things even more readable. But, even just SortNamesHappyPath() gives some idea of what's going on.

[ Parent | on HN ]