C++ Core Guidelines: Naming and Layout Rules


The C++ core guidelines have about twenty naming and layout rules. A few of them are obvious; a few of them may be controversial. Let's see, what I mean.


 mistake 968334 1280

Naming and Layout Rules

First of all, consistency is more important than these naming and layout rules. With this in mind, here is an overview of the rules.

  • NL.1: Don’t say in comments what can be clearly stated in code
  • NL.2: State intent in comments
  • NL.3: Keep comments crisp
  • NL.4: Maintain a consistent indentation style
  • NL.5: Don’t encode type information in names
  • NL.7: Make the length of a name roughly proportional to the length of its scope
  • NL.8: Use a consistent naming style
  • NL.9: Use ALL_CAPS for macro names only
  • NL.10: Avoid CamelCase
  • NL.11: Make literals readable
  • NL.15: Use spaces sparingly
  • NL.16: Use a conventional class member declaration order
  • NL.17: Use K&R-derived layout
  • NL.18: Use C++-style declarator layout
  • NL.19: Avoid names that are easily misread
  • NL.20: Don’t place two statements on the same line
  • NL.21: Declare one name (only) per declaration
  • NL.25: Don’t use void as an argument type
  • NL.26: Use conventional const notation

I will not write about the naming and layout rules of the C++ core guidelines which have already sufficient explanations. I only write about the rules, which need additional wording or are often discussed in my seminars.

NL.1: Don’t say in comments what can be clearly stated in code

Honestly, I'm not a friend of writing a comment to each piece of code. Commenting each piece of code may be a code smell because you code it too sophisticated. I'm more with the Python rule: explicit is better than implicit. I only write a comment if I have to apply a trick which is not apparent. For example, young professionals tend to remove curly braces from your code such as you would remove redundant round braces from an arithmetic expression. If you don't believe me, visit my seminars. But curly braces are essential to scope RAII objects such as locks or smart pointers. Removing a curly brace from a lock may give you a slower program, or a deadlock. On end, a lot of my customers comment the following usage of curly braces.

std::mutex mut;
{   // necessary to manage the lifetime of the lock
    std::lock_guard<std::mutex> lock(mut);


What is bad about comments is that they become out of date. By definition, this is not possible for the source code. It's always up to date.  As a freshman, my job was quite often to refactor legacy code. Honestly, I had often no clue what it does, and I was, therefore, quite frustrated. To my rescue, I found a few comments. But the comments were completely out-dated. It took me quite a time to recognise this, and you can't image how I felt. Comments have to be maintained as code but are often very sloppy maintained.

NL.5: Don’t encode type information in names

Really? I thought since the last millennium we are done with Hungarian Notation. I mean this good old times in which our variables had no type. Hungarian Notation is redundant with the type-checking of the compiler, contradicts generic programming, and - this is my main arguments - becomes out-dated as comments become out-dated. Can you guess for what the following variable names stand?



If you don't know, here it the solution: Hungarian Notation.

NL.7: Make the length of a name roughly proportional to the length of its scope

This rule sounds strange, but we are already used to it. Giving a variable the name i or j, or giving a variable the name T will make the intention of the code immediately clear: i and j are indices, and T is a type parameter of a template.

template<typename T>    
void print(ostream& os, const vector<T>& v)
    for (int i = 0; i < v.size(); ++i)
        os << v[i] << '\n';


There is a meta-rule behind this rule. A name should be self-explanatory. In a short context, you get with a glance what the variable means. This self-evidence will not automatically hold for longer contexts; therefore, you should use longer names.

NL.16: Use a conventional class member declaration order

Okay, this is an easy but quite helpful rule.

  • When you declare a class, use the following order: constructors, assignments, destructor before functions, and data.
  • The same goes for public, protected, and private: Use the public before protected before private order.
  • Don't use the access specifiers in a class more than once:
class X {   // bad
    void f();
    int g();
    // ...

NL.19: Avoid names that are easily misread

Can you read this example without any hesitation?

if (readable(i1 + l1 + ol + o1 + o0 + ol + o1 + I0 + l0)) surprise();


Honestly, I often have issues when I type-in a password having the number 0, or the big capital O. Two years ago it took me quite a time to log into a server. The automatically generated password had a character O.

NL.20: Don’t place two statements on the same line

Let me give you two examples. Did you spot the two issues?

char* p, p2;
char a = 'a';
p = &a;
p2 = a;                              // (1)

int a = 7, b = 9, c, d = 10, e = 3;  // (2)


p2 is a char  (1), and c is not initialised (2).

With C++17, we got an exception to this rule: structured binding. Structured binding allows me to declare more than one name in one declaration (line 1).

Now, I can write the if statement with an initialiser cleaner and more readable.

std::map<int,std::string> myMap;

if (auto [iter, succeeded] = myMap.insert(value); succedded){  // (1)
    // ...
    // ...
} // iter and succeeded are automatically destroyed            

What's next?

DONE! After more than one hundred posts to the C++ core guidelines, I have two good news for you.

First, I write a book about the C++ core guidelines, in which I try my best to make a readable and concise story out of this highly valuable content. My idea is it, to base my story on C++17 and these rules of the C++ core guidelines, which are essential to write modern C++. Of course, modern C++ means, in particular, to write type-safe, bounds safe, and lifetime safe C++ code. So, stay tuned, I start in the next days and give you an update, if appropriate.

Second, my blog will shift to the hot topic in C++: the upcoming C++20 standard. My posts start in a breadth-first search and will end in a depth-first search. C++20 is presumably as powerful as C++11. You can, therefore, assume, that I have a lot to write. 


Thanks a lot to my Patreon Supporters: Matt Braun, Roman Postanciuc, Tobias Zindl, Marko, G Prvulovic, Reinhold Dröge, Abernitzke, Frank Grimm, Sakib, Broeserl, António Pina, Darshan Mody, Sergey Agafyin, Андрей Бурмистров, Jake, GS, Lawton Shoemake, Animus24, Jozo Leko, John Breland, espkk, Wolfgang Gärtner,  Louis St-Amour, Stephan Roslen, Venkat Nandam, Jose Francisco, Douglas Tinkham, Kuchlong Kuchlong, Avi Kohn, Robert Blanch, Truels Wissneth, Kris Kafka, Mario Luoni, Neil Wang, Friedrich Huber, lennonli, Pramod Tikare Muralidhara, and Peter Ware.


Thanks in particular to Jon Hess, Lakshman, Christian Wittenhorst, Sherhy Pyton, Dendi Suhubdy, and Sudhakar Belagurusamy. 



I'm happy to give online-seminars or face-to-face seminars world-wide. Please call me if you have any questions.

Bookable (Online)



Standard Seminars 

Here is a compilation of my standard seminars. These seminars are only meant to give you a first orientation.


Contact Me

Modernes C++,


My Newest E-Books

Course: Modern C++ Concurrency in Practice

Course: C++ Standard Library including C++14 & C++17

Course: Embedded Programming with Modern C++

Course: Generic Programming (Templates)

Course: C++ Fundamentals for Professionals

Subscribe to the newsletter (+ pdf bundle)

Blog archive

Source Code


Today 7885

Yesterday 7863

Week 33759

Month 147222

All 5444326

Currently are 105 guests and no members online

Kubik-Rubik Joomla! Extensions

Latest comments