Re: Latest draft of C++ 11 Organization: Roundhouse Consulting, Ltd.
On 7 jun, 13:33, Daniel Kr?gler <daniel.krueg...@googlemail.com>
wrote:
I'm unaware of any official proposal of static_if. Maybe this proposal
had a different name, but I couldn't find any, therefore I would say
that this lack is a failure of those who wanted it. Participate in the
process of standardization and you can improve the standard. It is much
easier to criticize a state that has been baked by others.
Maybe I should have proposed it, you are right.
This does not sound like any reasonable form of criticism to me.
"Perfect forwarding" is not defined in the standard, so I don't see how
you can argue against the usage of a non-official term.
I've read an article
http://www.artima.com/cppsource/rvalue.html
which states it as a key motivation for rvalue references, but Scott
Meyers has already collected a few examples were perfect forwarding
fails miserably:
http://tinyurl.com/3rzwsao
I miss to see
any specific arguments for "Initializer lists are so wrong!" and
I've tried to discuss something in this forum, but got no responses:
http://tinyurl.com/6b6c2av
"A good interface for valarrays?", therefore these are buzz words to me.
Josuttis heavily criticises the valarray interface in his "The C++
Standard Library - A Tutorial and Reference". It seems that the issue
has not, so far, been addressed. You probably know that valarrays are
fundamental to turn some optimizations on that are not possible due to
the well known aliasing problem.
If there are problems that you are aware of (not handled yet by existing
issues), why don't you submit a corresponding issue?
I would if it were under the direction Stroustrup or some "benevolent
dictator", but the mere idea of going through a bureaucratic committee
makes me itch. Sorry, I know you won't agree with such a posture.
A standard is never perfect and you really *need* at some point in time
complete (or near to complete) implementations to find more subtle
defects.
Agreed.
An ISO standard has a very specific mechanism to handle defects
and problems, these are issues and technical corrigenda, so I cannot
agree with your conclusion. Based on your suggestion we should better
never publish a standard, because we will have a hard time that this
standard is perfect.
It is not quite the mechanism that bothers me. It is the fact that
decisions seem not to be taken based on the scientific merit of the
proposals, but rather on political influences and the such. Of course
this may be only an impression, but is it?
Forget all the technical issues, and try explaining me the fact that,
as Weimer pointed out:
On the other hand, almost everybody in the C++ community hopes that
the FDIS continues to leak to the Internet and is consulted by
programmers because everything else would have a disastrous impact on
adoption.
So, they want to sell something expecting it to be leaked, because if
not there will be no adoption. Does it make sense?
Thank you for your time and attention,
Elias.
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]