Re: [Boost-users] [Containers Library Review] The review of the Containers library by Ion Gaztanaga starts today

Subject: Re: [Boost-users] [Containers Library Review] The review of the Containers library by Ion Gaztanaga starts today
From: Szymon Gatner (szymon.gatner_at_[hidden])
Date: 2011-08-17 08:11:28

Hi,

It is my first review ever, sorry if something is not right. Here it goes:

>
>   *  What is your evaluation of the design?

As it is designed according to C++0x specs not really much to
debate here. It allows us (users) to get used to both new functionality
like emplacing and move semantics.

>   * What is your evaluation of the implementation?

Only went briefly through vector implementation while debugging
to see how exactly move semantics is implemented. Code is very
clean and easy to follow in fact way easier to grasp than visual
studio's std::vector implementation.

>   * What is your evaluation of the documentation?

Lacks. I was especially interested in new allocators. Rest is pretty
much the same as C++98 equivalents so no problematic to figure out.

>   * What is your evaluation of the potential usefulness of the library?

It is imho very important and frankly long overdue. Move semantic are in
debate for years yet it was never available in boost even tho library-level
emulation was fully possible (like Alexandrescu's mojo) for years.

>   * Did you try to use the library? With what compiler? Did you have any
> problems?

I used vector in VS 2008 SP1. No problems.

>   * How much effort did you put into your evaluation? A glance? A quick
> reading? In-depth study?

Didn't have time to do in-depth study. Made a class movable with boost.move
and played a bit with vector of them. Found something very interesting tho -
check Ion's answer to my question regarding forward declaring movable class.

>   * Are you knowledgeable about the problem domain?

Understand it ;)

>
> And finally, every review should answer this question:
>
>   * Do you think the library should be accepted as a Boost library? Be sure
> to say this explicitly so that your other comments don't obscure your
> overall opinion.
>

3 x yes:
1) It gives the possibility to start think in new way (like return
containers by value) already.
2) I can enjoy it with older (VC2008) compiler
3) without boost.container boost.move is only 1/5 complete

Best regards,
Szymon Gatner