Re: It's v7-> mainline merge time!

[Gabriel Dos Reis <gdr at integrable-solutions dot net>]

Paolo Carlini <pcarlini@suse.de> writes:

| Hi,
| 
| >| -2- Simulated move semantics, basically as-is in the branch. The
| >| entire merging work can be split in the infrastructural bits (e.g.,
| >| moveable.h, slt_construct.h, stl_uninitialized.h, testsuite),
| >| algorithms, trivial containers, vector and deque, testcases  as we
| >| go. I'm pretty sure the work is safe from the binary-compatibility
| >| point of view (or can be made so with minor tweaks): in the most
| >| complex case, vector and deque, in a nutshell, functions like, e.g.,
| >| _M_insert_aux, _M_fill_insert, _M_range_insert, will use __move and
| >| become faster when moveable types are detected.
| >
| >I'm deeply reserved about that experiment going to mainline.
| >  
| >
| Humm, Chris did a lot of work and certainly it was meant to be included
| at some point.

I seem to remember that in the past, when I expressed my concerns, I
was told it was just experimental and, btw it was just on v7 branch.
(And many consider the v7 branch where all sorts of risks are
encouraged to be taken.  I see nothing wrong with that.)

We do have a proposal for move semantics -- that obvisouly affects the
standard library; the committee is working on the synergies.  It is my
opinion that when we have set the details and we've a merge into the
WP, then we may justify possible alteraions of user codes that we're
anticipating the next version of C++0x which already has those
specifications. 

I'm for all sorts of experiments on branches;  however, unless the
experiments solve a major problem (like the visiblity issue), I'm of
the opinion that for production release we shall exercise caution
inversely proportional to what we may do in sanbox.

-- Gaby