Revert emplace_back to the non-C++11 push_back

[henningjp]

For speed increases, the push_back() functions were replaced with emplace_back() in #17. However, one of the goals of the CoolProp project is to maintain MSVC 2008 compatibility. Since emplace_back() is a C++11 language standard and not supported in MSVC 2008, this functionality has been reverted, while the other optimization items in #17 have been retained. Users desiring the emplace speed increases can refer to the previous commit (90cfa1e), however, CoolProp/CoolProp will reference this commit or later.

[jowr]

Just a stupid question: Why don't you use a combination of resize (instead of reserve) and then access the elements directly by their index instead of using push_back? The direct access is named element assignment in http://coliru.stacked-crooked.com/a/83d23c2d0dcee2ff that was mentioned in #17

[ibell]

It's a good question. Reserve expands the block of memory so that we can expand it to be at least as many elements as we need via push_back and no reallocations will be required (which slows things down), but the entries are not allocated. With resize, they are allocated, and are filled with default values. With C++11, emplace_back yields moves rather than copies, which, for large objects, can matter a lot. This is, of course, as I understand it, which doesn't mean it is correct.

…

On Mon, May 1, 2017 at 2:24 PM, Jorrit Wronski ***@***.***> wrote: Just a stupid question: Why don't you use a combination of resize (instead of reserve) and then access the elements directly by their index instead of using push_back? The direct access is named *element assignment* in http://coliru.stacked-crooked.com/a/83d23c2d0dcee2ff that was mentioned in #17 <#17> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#22 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABxha2PLyrvzyQLRfmJB14lFcIRP_ngZks5r1j-agaJpZM4NLvZ7> .