Replace sparse usage of nullptr with NULL

[xzyfer]

We've agreed that we don't need nullptr and could easily get by with 0 or NULL. #996 (comment)

This change is mostly a noop but at will increase compatibility with older or exotic compilers.

[mgreter]

We don't have any nullptr left-over? What's the definition of done for this issue?

[xzyfer]

Looks like you're correct. If we're happy to go with 0 over NULL we're done here.

@saper suggested we should prefer NULL. He can speak to this more.

[saper]

Replacing nullptr with 0 is simply incorrect, it should be NULL. I have updated #996 to fix this.

Also please be more verbose in the commit messages, saying Add some code improvements is not good!

[saper]

Ah, #996 got closed. It is now #1660.

[xzyfer]

I didn't realise this root issue was specific to just src/util.cpp. We use 0 in place of NULL a lot.

[saper]

I didn't realise this root issue was specific to just `src/util.cpp`. We use 0 in place of NULL a lot.

This is not good. Right now it is probably hard to identify those places.

[xzyfer]

I believe that is why @mgreter wanted to leave it as is. My understanding of NULL in C++ is that it's functionally equivalent to 0 (at least in out use case)

In C++, the definition of NULL is 0, so there is only an aesthetic difference. I prefer to avoid macros, so I use 0. Another problem with NULL is that people sometimes mistakenly believe that it is different from 0 and/or not an integer. In pre-standard code, NULL was/is sometimes defined to something unsuitable and therefore had/has to be avoided. That's less common these days.

If you have to name the null pointer, call it nullptr; that's what it's called in C++11. Then, nullptr will be a keyword.
Source

[saper]

I believe that is why @mgreter wanted to leave it as is. My understanding of NULL in C++ is that it's functionally equivalent to 0 (at least in out use case)

In C++, the definition of NULL is 0, so there is only an aesthetic difference. I prefer to avoid macros, so I use 0. Another problem with NULL is that people sometimes mistakenly believe that it is different from 0 and/or not an integer. In pre-standard code, NULL was/is sometimes defined to something unsuitable and therefore had/has to be avoided. That's less common these days.

If you have to name the null pointer, call it nullptr; that's what it's called in C++11. Then, nullptr will be a keyword.
Source

Ok, so this is not "wrong" just "I don't like it".

Even in the examples http://www.cplusplus.com/reference/cstdlib/strtoul/ says NULL.

The reason is simple:

stroul(buffer, 0, 16); looks like 0 was just an ordinary number (offset into the buffer, maybe?)

And of course then C++11 needed to create nullptr to solve problems it created .... :)
And later they figured out they need nullptr_t, too!

[xzyfer]

Ok, so this is not "wrong" just "I don't like it".

This distinction is good to know.

stroul(buffer, 0, 16); looks like 0 was just an ordinary number

Cool. So generally speaking I'm with @mgreter. We should stick with 0 for now.
However understand what you're saying in this case, and I think it makes sense to use NULL where 0 is ambiguous.

IMHO we should ship #1660.