[cmroanirgo]

If everyone can forgive my ignorance... I used to call everything in the std:: namespace as just that, the standard template library (STL) (whatever it's iteration). So is this C++03 /C++11 stuff just updates to this library, or is std::string recognised at the compiler level? (Genuinely confused).

(In old man voice) We ended up rolling out own stuff and marking std:: verboten. Why? Stl was too slow, too verbose, and too hard to grok the stack when debugging. We ended up with less memory fragmentation, less dangling ptrs, etc etc. In the rare case there was something in STL that was actually cool (or faster, which was very rare), we'd gut it and reef out the part that was cool to use in our implementation.

I presume these comments aren't popular (sorry about that, but this is during 90s and early 00s when dev cycles were clearly different). Eg. We had string classes in all different flavours, some would interop, some wouldn't. Eg. We had tree and hash classes that, while templateable, had a few core implementations that made compilation fast. We had various ptr management systems (ref counted, stack based, etc).

We made STL between verboten in all APIs because we been burnt so many times using (/trying to use) other ppls APIs that exposed STL in its library. (We were exclusively a windows shop, if that helps understand my confusion... PS. I'm retired these days and have been out of the c++ game >10yrs)

[kllrnohj]

> So is this C++03 /C++11 stuff just updates to this library, or is std::string recognised at the compiler level?

std::string is in the STL. So is std::string_view. There are language changes as well that enable some of the new STL additions but some are just STL updates and you could "backport" them so to speak. Or do the same thing but yourself.

Making std:: verboten these days would likely be a mistake, though. There's so many things that are just not controversial and not worth re-implementing. Like std::unique_ptr. Or std::array.

Most of the containers are still better off ignored, though, many of them unfixable. The least worst of them is std::vector and it's still _ok_ but even there things like absl::InlinedVector are worth a strong consideration instead ( https://github.com/abseil/abseil-cpp/blob/master/absl/contai... ). Or boost's small_vector ( https://www.boost.org/doc/libs/1_60_0/doc/html/boost/contain... )

[vlovich123]

I still think the STL containers are the sane default to reach for & switch to more obscure ones when the problem domain and performance requirements say to use a different one.

[kllrnohj]

You have that backwards. The STL containers are for when you have a hyper-specific niche use case. They are otherwise terrible defaults and everyone should use boost or abseil by default otherwise.

std::map, for example, is only appropriate if you need a red-black tree specifically. Which almost nobody does. std::unordered_map is less awful, but abseil has a literal straight upgrade. With the same API. So... why would you pick the slower thing when you're using C++? std::vector is only really appropriate if you know you never have small vectors, which is again a more obscure situation.

[rumanator]

> You have that backwards. The STL containers are for when you have a hyper-specific niche use case.

That assertion makes no sense at all. The stl contrainers work very well as basic generic containers that can safely be used for pretty much any conceivable use where performance isn't super critical. I'm talking about cases like, say, you need to map keys to values but you don't really care about performance or which specific data structure you're using. That's stl's domain: robust, bullet-proof implementations of basic data structures that are good enough for most (or practically all) cases with the exception of a few very niche applications.

If you happen to be one of the rare cases where you feel you need to know if a container is built around a red-black tree or any other fancy arcane data structure, and if this so critical to you that you feel the need to benchmark the performance to assess whether you either need to use non-defaults or completely replace parts or the whole container with a third-party alternative... Then and only then the stl is not for you.

[kllrnohj]

This makes no sense. The STL is the specialized containers with obscure performance characteristics & behaviors. Boost & abseil provide the generic, reasonable default ones.

You're arguing it's better to use something that's across the board worse for nearly every user, and by a lot, just because... why? It's slightly more convenient?

[rumanator]

> This makes no sense. The STL is the specialized containers

It really isn't. The whole STL is, by design, a template library packed with generic data structures that are designed to have very robust defaults and still be customizable and extensible.

When the defaults are good enough, which is the case in general, the STL will do. If you have a niche requirement (say, games) or feel adventurous, you adopt custom and/or specialized solutions.

This has been the case since the STL's inception. They are the standard, default library. I can't understand how someone is able to miss this fact.

[pjmlp]

Because STL is part of the compiler, guaranteed to work on every platform supported by the compiler, and does not require lots of paperwork for adoption at many shops.

[mehrdadn]

Small vectors break iterator guarantees, for one thing. They also really only make sense for tiny objects (ints, etc.) given you don't want a pickup truck's worth of data on your stack. They're most definitely not general-purpose.

There are lots of subtleties STL containers have to worry about in designing containers, regarding everything from iterator & pointer invalidation to allocation and allocator propagation. All this is because they're designed to be general-purpose and support most conceivable use cases. Their replacements have to trade off requirements in order to get better performance or otherwise improve on some axes.

[kllrnohj]

> Small vectors break iterator guarantees, for one thing.

It only breaks swap of the container itself during iteration. Which is a super niche condition.

And that swap also invalidates some of std::vector's iterators as well - specifically the end() iterator.

> They also really only make sense for tiny objects (ints, etc.) given you don't want a pickup truck's worth of data on your stack. They're most definitely not general-purpose.

Of course they are still general-purpose. They can (and do) specialize on the size of the object being contained. The only reason std::vector doesn't also have SSO is because it's an ABI break. Not because it's better in some way or less fragile. Legacy is the only reason.

[mehrdadn]

> And that swap also invalidates some of std::vector's iterators as well - specifically the end() iterator.

And they don't invalidate the iterators that point to actual elements, which was kind of the entire point I was making. Don't let that stop you from trying to make it look like I'm just blurting out nonsense, though.

[leetcrew]

how do small vectors break iterator guarantees?

[mehrdadn]

By invalidating iterators (and pointers and references) when swapping: https://stackoverflow.com/a/8191356

[gpderetta]

Not sure about iterators, but they do break reference guarantees. Moving a vector doesn't invalidate pointers to the contained elements.

[Espressosaurus]

Because that way you don't need to haul in dependencies unless you have a real reason.

std::function is fine for prototyping, but its size hit is extreme, so in embedded code we use other implementations. But where size and speed doesn't matter? Why bother?

[kllrnohj]

> Because that way you don't need to haul in dependencies unless you have a real reason.

These are all largely header libraries. You're already hauling in a dependency, and in every c++ file that uses it at that.

> std::function is fine for prototyping

std::function isn't part of the containers library of the STL (containers being all the stuff here: https://en.cppreference.com/w/cpp/container ). I agree std::function is fine, it even has a pretty reasonable small-size optimization.

[mbeex]

> These are all largely header libraries.

That's not even true for boost, no matter if they always advertise that. The lib is also notorious for bad decomposability (using only a subset without installing the whole monster). Not to speak about idiosyncratic naming and build system, making it sometimes hard to include it in meta builds of other libraries and frameworks.

In sum: Anyone sensible, regarding different kinds of footprint and dependencies will think twice, before pulling in these kind of libraries.

[fctorial]

I'll use java where size and speed doesn't matter.

[pjmlp]

I rather have my phone be fast, without the apps taking ages to download.

[steerablesafe]

abseil's might have a similar API but it's most definitely not the same (function signatures looking the same doesn't make it the same API). Some of the standard containers can't be fast because too strict/specific standard requirements, not because they don't try hard enough.

Having said that I think using abseil's containers is reasonable, even as a default, if you can afford the dependency.

> std::unordered_map

AFAIK unordered_map is the most awful of all standard containers.

[leetcrew]

what's wrong with unordered_map? it's at least more useful than map.

[aw1621107]

From what I understand its API is overspecified to the point that it basically forces a "traditional" array-of-linked-buckets implementation, which can be horribly slow on modern processors due to the need to chase pointers. This means a lot of the potential performance improvements of allowing map elements to be unordered are lost.

[pjmlp]

Because C++ is a nice, mostly safe, general purpose language, being spoiled by the 1% "performance above anything else crowd".

I want my OWL, VCL back, not an hash table able to do lookups in micro-seconds

[TeMPOraL]

I know you're exaggerating here a bit, but come on, a hash table not able to do lookups in micro-seconds is just garbage. I expect better of any language, not just C++.

[pjmlp]

Indeed I was.

The point I was trying to make, was that from productivity point of view, there are more relevant stuff to fix in C++ than algorithm complexity of STL implementations.

Like catching up with what Java 5 standard library offers for networking and parallel/concurrent programming.

[gpderetta]

on the contrary, std::map is a good default container with predictable performance. If you need fast O(1) look up std::unordered_map is really not fit for purpose and requires you to come up with an hash function.

[kouteiheika]

> on the contrary, std::map is a good default container

If you care about performance then it's not. (And if you don't care about performance then why are you using C++?) The standard requires that `insert` will not invalidate iterators, which basically forces everyone to implement `std::map` as a red-black tree, and those are pretty bad performance-wise on modern hardware mostly due to cache misses.

> If you need fast O(1) look up std::unordered_map is really not fit for purpose and requires you to come up with an hash function.

Modern hash table implementations (along with a modern hash function) are exactly what you should use if you need fast average-case O(1) lookup, so I'm confused why would you say that it's not fit for purpose? Unless you specifically meant only `std::unordered_map` which, yes, is pretty atrocious performance-wise (again, due to the iterator invalidation requirements).

[gpderetta]

I meant specifically std::unordered_map, because how it is specified in the standard it is very hard to implement it efficiently. If you need performance, yes, use a good hash table implementation. But even in C++ you do not need to be shaving cycles everywhere and there std::map is better than std::unordered_map.

[vlovich123]

Every new standard incorporates language & library changes. A perfect example of this is r-value references. That was a new language feature in C++11 & was adopted by all the standard library containers & algorithms.

Not sure which part of std::string you're referring to but the compiler generally doesn't contain any knowledge of the library itself. It does goes the other way though where the standard library has to know how to implement certain functionality on a given compiler (some type traits functionality IIRC isn't possible to implement without compiler builtins that expose the AST to you). I think Rust has taken a more sustainable approach with their macro system which can modify the AST instead of relying on builtins but even in Rust I suspect they use builtins in certain places of their standard library.

Today's STL implementations are going to be better performing and more robust than anything you'd write yourself so generally a good idea to stick to it as a rule of thumb for the majority of code.

[atq2119]

> even in Rust I suspect they use builtins in certain places of their standard library.

There are places in the rust standard library which just omit the implementation because it's magically filled in by the compiler based on the type and function names. Which, for really low-level and fundamental functionality, seems fair game to me.

[gpderetta]

As non exhaustive list, compilers have have intimate knowledge of:

* various operator new * all the type traits * a lot of the names inherited from the C library

compilers could do much more, but in general they prefer to implement generic optimizations instead of targeting a specific library name (for example removing allocation for stack allocated std::strings was not done until the generic removal of alloc/free was implemented).

[MaxBarraclough]

Many of the type-traits can be implemented without specific compiler support. That's how Boost managed it. Here's their implementation of is_signed if you're curious [0].

I believe some standard-library type-traits cannot be implemented without specific compiler support.

[0] https://github.com/boostorg/type_traits/blob/059ed883/includ...

[kccqzy]

That is indeed a very different time. I believe in the 90s the STL was considered too radical in its use of templates to be practical. People's perception has come a long way. Heck, instead of sane std::list<T> people used to do intrusive linked lists by having T inherit from a linked list base class. What a terrible idea.

[AstralStorm]

Intrusive lists are so good they are likely to land one of these days in standard. Except via templates and traits not silly unnecessary inheritance. (Boost-like.)

[gpderetta]

Yes, boost::intrusive is particularly great.

[taejo]

Do people use std::list? I can't think of a time when I've needed a non-intrusive linked list (either vector is better or I need an intrusive list).

[pansa2]

> We ended up rolling out own stuff and marking std:: verboten.

I think that’s still pretty common, especially in game development and/or when compiling with exceptions disabled. AFAIK std::vector is fine, and probably std::string, but for things like std::[unordered_]map, Abseil and Folly have equivalents that are slightly non-standard and much faster.