5 ms·

It’s not like the idea is completely crazy; pointer misuse is a big source of all kinds of problems, Rust shows how much can be done without constantly using ra

It’s not like the idea is completely crazy; pointer misuse is a big source of all kinds of problems, Rust shows how much can be done without constantly using raw pointers, and what are the C++ Core Guidelines if not an effort in exactly this direction? I finally “got” that this is a joke when they discussed the future alternatives: > For copyable and assignable references you can use std::reference_wrapper. (http://en.cppreference.com/w/cpp/utility/functional/reference_wrapper http://en.cppreference.com/w/cpp/utility/functional/referenc...) and thought, ha ha, an “assignable reference” instead of a “pointer”, that’s obviously completely different. :) What a subtle joke! Only now I see that reference_wrapper is apparently completely real and exists since C++11. Why does this even exist? My head hurts from banging it against the desk.

jopsen 9y ago
It probably exists facilitate type parametrization of pointer implementations. Just a guess. That still crazy though :)
_o_ 9y ago
1. April joke :) The point of c/c++ in comparision to other languages is that you can do anything and having a huge toolbox for that. Pointers are excelent sledgehammer. You can be perfectly fine with not using them but c++ need to have them. The pointers are not source of the problem, the developer is. That's why languages like java are prospering, it prevent incompetent people making stupid mistakes and that is fine. But some people need and want freedom that c++ offers.
- lucidguppy 9y ago
  Raw pointers are like "transporter scalpel" used by star trek doctors. When used with care it does the job and does it well. You need plenty of tests and valgrind. Its dangerous for most programmers to be cavalier with owning raw pointers. If we aren't we'll get another Java.
- laumars 9y ago
  Your point may be technically accurate but the elitism of the tone completely undermines it's point because even the best of us can and do write code with errors in it. I mean just look at all the mature code out there written by experienced developers which fails fuzz testing. I've been writing software for close on 30 years and I know damn well how easy it is to make mistakes. Even on days when I think I'm in the zone, there will be occasions when I'll go back to it and think "WTF was I thinking?" when it bombs out of the first round of automated testing.
- pjmlp 9y ago
  You can have languages that prevent people making stupid mistakes and still offer full control over the hardware, that is what languages like Modula-2 and Ada offer.
  - evincarofautumn 9y ago
    Yeah, also Rust and ATS. A good systems language just needs to allow low-level and unsafe operations; nowadays the industry is warming up to the idea that such operations shouldn’t be available by default. Having to opt in with a machine-checked “unsafe” annotation means you have an explicit indication in the code of everywhere things could go wrong with type/memory safety, which means when something does go wrong, you know it must be caused by one of those unsafe regions.
- FrankHB 9y ago
  The problem is that pointers almost totally violate the single responsibility principle. Technically, the only dependencies on pointers are types of the allocation/deallocation functions which are mandated in ISO C++ even for freestanding implementations (not the case in ISO C), and their derivations (new-expressions, etc). All other sane uses of pointers should be replaced by some better alternatives when appropriate (shared_ptr, unique_ptr, observer_ptr, reference, reference_wrapper without type completeness requirement, uintptr_t, iterator, etc). In that sense every pointer is a badly designed variant type; basically each occurrence of a pointer would likely be mostly incorrect (or not correct enough) in program semantics. I always treat that vagueness as a can of worms, so even merely replacing pointer by observer_ptr somewhere makes a big win to me. There do exist real costs like verbosity and bloating of binary size, but they have to be paid for sin of the father. The ship has sailed too far in a regretful direction; but this is not the excuse to fix it - which is one source of the problem from many developers.
evincarofautumn 9y ago
In all seriousness, reference_wrapper lets you: 1. Have a container of non-null references. As the example in your cppreference link shows, that’s convenient for efficiently providing multiple “views” of the same values. 2. Pass a value with reference semantics to a function that takes an argument by value. Mainly useful to avoid copying or reuse a function object. 3. Explicitly say “this is a reference to something I don’t own”, complementing unique_ptr’s “reference to something I own exclusively”, shared_ptr’s “reference to something I share ownership of”, and raw pointers’ “reference to something I may or may not own”.
netheril96 9y ago
> Why does this even exist? When you use `std::bind` or `std::thread`, the arguments you pass are captured and stored by value. What if you want it to be captured as reference? You use a `std::reference_warpper` instead. Note you cannot use a pointer here because the type won't match, while `reference_wrapper` implicitly converts.
andrepd 9y ago
cppreference explains it pretty well: >std::reference_wrapper is a class template that wraps a reference in a copyable, assignable object. It is frequently used as a mechanism to store references inside standard containers (like std::vector) which cannot normally hold references. >Specifically, std::reference_wrapper is a CopyConstructible and CopyAssignable wrapper around a reference [...] of type T. Instances of std::reference_wrapper are objects (they can be copied or stored in containers) but they are implicitly convertible to T&, so that they can be used as arguments with the functions that take the underlying type by reference. Sounds pretty reasonable and useful to me. No need for over-the-top head banging :)
- mannykannot 9y ago
  It is also an example of a sort of 'combinatorial pump' that inflates the language's complexity, as each new feature needs to work with every existing one. It is my guess that it is C++'s multi-paradigm flexibility that makes this both possible and necessary.

jopsen 9y ago

It probably exists facilitate type parametrization of pointer implementations. Just a guess. That still crazy though :)

_o_ 9y ago

1. April joke :) The point of c/c++ in comparision to other languages is that you can do anything and having a huge toolbox for that. Pointers are excelent sledgehammer. You can be perfectly fine with not using them but c++ need to have them. The pointers are not source of the problem, the developer is. That's why languages like java are prospering, it prevent incompetent people making stupid mistakes and that is fine. But some people need and want freedom that c++ offers.

lucidguppy 9y ago
Raw pointers are like "transporter scalpel" used by star trek doctors. When used with care it does the job and does it well. You need plenty of tests and valgrind. Its dangerous for most programmers to be cavalier with owning raw pointers. If we aren't we'll get another Java.
laumars 9y ago
Your point may be technically accurate but the elitism of the tone completely undermines it's point because even the best of us can and do write code with errors in it. I mean just look at all the mature code out there written by experienced developers which fails fuzz testing. I've been writing software for close on 30 years and I know damn well how easy it is to make mistakes. Even on days when I think I'm in the zone, there will be occasions when I'll go back to it and think "WTF was I thinking?" when it bombs out of the first round of automated testing.
pjmlp 9y ago
You can have languages that prevent people making stupid mistakes and still offer full control over the hardware, that is what languages like Modula-2 and Ada offer.
- evincarofautumn 9y ago
  Yeah, also Rust and ATS. A good systems language just needs to allow low-level and unsafe operations; nowadays the industry is warming up to the idea that such operations shouldn’t be available by default. Having to opt in with a machine-checked “unsafe” annotation means you have an explicit indication in the code of everywhere things could go wrong with type/memory safety, which means when something does go wrong, you know it must be caused by one of those unsafe regions.
FrankHB 9y ago
The problem is that pointers almost totally violate the single responsibility principle. Technically, the only dependencies on pointers are types of the allocation/deallocation functions which are mandated in ISO C++ even for freestanding implementations (not the case in ISO C), and their derivations (new-expressions, etc). All other sane uses of pointers should be replaced by some better alternatives when appropriate (shared_ptr, unique_ptr, observer_ptr, reference, reference_wrapper without type completeness requirement, uintptr_t, iterator, etc). In that sense every pointer is a badly designed variant type; basically each occurrence of a pointer would likely be mostly incorrect (or not correct enough) in program semantics. I always treat that vagueness as a can of worms, so even merely replacing pointer by observer_ptr somewhere makes a big win to me. There do exist real costs like verbosity and bloating of binary size, but they have to be paid for sin of the father. The ship has sailed too far in a regretful direction; but this is not the excuse to fix it - which is one source of the problem from many developers.

evincarofautumn 9y ago

In all seriousness, reference_wrapper lets you: 1. Have a container of non-null references. As the example in your cppreference link shows, that’s convenient for efficiently providing multiple “views” of the same values. 2. Pass a value with reference semantics to a function that takes an argument by value. Mainly useful to avoid copying or reuse a function object. 3. Explicitly say “this is a reference to something I don’t own”, complementing unique_ptr’s “reference to something I own exclusively”, shared_ptr’s “reference to something I share ownership of”, and raw pointers’ “reference to something I may or may not own”.

netheril96 9y ago

> Why does this even exist? When you use `std::bind` or `std::thread`, the arguments you pass are captured and stored by value. What if you want it to be captured as reference? You use a `std::reference_warpper` instead. Note you cannot use a pointer here because the type won't match, while `reference_wrapper` implicitly converts.

andrepd 9y ago

cppreference explains it pretty well: >std::reference_wrapper is a class template that wraps a reference in a copyable, assignable object. It is frequently used as a mechanism to store references inside standard containers (like std::vector) which cannot normally hold references. >Specifically, std::reference_wrapper is a CopyConstructible and CopyAssignable wrapper around a reference [...] of type T. Instances of std::reference_wrapper are objects (they can be copied or stored in containers) but they are implicitly convertible to T&, so that they can be used as arguments with the functions that take the underlying type by reference. Sounds pretty reasonable and useful to me. No need for over-the-top head banging :)

mannykannot 9y ago
It is also an example of a sort of 'combinatorial pump' that inflates the language's complexity, as each new feature needs to work with every existing one. It is my guess that it is C++'s multi-paradigm flexibility that makes this both possible and necessary.