When the new C++ standard will be released.

[Ivan]

The new C++ standard will be release round the conner, but I am not
sure the time. Do anyone know this?

The new C++ standard was designed as C++09 and will be released at
2009. The exciting time is coming.
this standard will at least includes tow major core features, rvalue
references and type concepts.

I know the website of c++ standard committee http://www.open-std.org/jtc1/sc22/wg21/,
and so many issue proposed and approved lists in the page. But I
didn't what I should start to concert with the dynamic information.
Who has this experience, tell me please.

I want to discuss c++ standard with you, including the release time,
how to keep the trace to the development direction and related issue.
Come on!

 

[Ian Collins]

I want to discuss c++ standard with you, including the release time,
how to keep the trace to the development direction and related issue.
Come on!

Sounds like you want comp.std.c++

 

[Ivan]

Sounds like you want comp.std.c++

Yes, it is my fault.

 

[Bo Persson]

The new C++ standard will be release round the conner, but I am not
sure the time. Do anyone know this?

The new C++ standard was designed as C++09 and will be released at
2009. The exciting time is coming.
this standard will at least includes tow major core features, rvalue
references and type concepts.

I know the website of c++ standard committee
http://www.open-std.org/jtc1/sc22/wg21/, and so many issue proposed
and approved lists in the page. But I
didn't what I should start to concert with the dynamic information.
Who has this experience, tell me please.

I want to discuss c++ standard with you, including the release time,
how to keep the trace to the development direction and related
issue.

We don't know the release date yet.

As you say, the committee works very hard to produce a final standards
proposal this year, but we will not know until about October if they
actually succeed.

Otherwise we might get C++10 instead.


Bo Persson

 

[Christof Donat]

Hi,

Otherwise we might get C++10 instead.

I thought it would be a hexadecimal number. So we will get C++0a ;-)

Christof

 

[Howard Hinnant]

Hi,


I thought it would be a hexadecimal number. So we will get C++0a ;-)

0x0a or even 0x0b. 09 is no longer possible. N2800 represents CD1
(first committee draft). This should be viewed as "feature complete",
but definitely not bug free.

-Howard

 

[Ivan]

0x0a or even 0x0b.  09 is no longer possible.  N2800 represents CD1
(first committee draft).  This should be viewed as "feature complete",
but definitely not bug free.

-Howard

I expect the new standard will be release as soon as possible. I have
read an article before at http://www.devsource.com/c/a/Languages/C09-A-Glimpse-into-the-Future/
and it say that:
----------------------------------------------------------------------------------
Following the Portland meeting in October 2006, the C++
standardization committee set 2009 as the target date for C++09, the
next C++ standard. C++09 will include at least two major core
features: rvalue references and type concepts. Other core features are
also included, such as automatic deduction of types from initializers,
delegating constructors, nullptr, and even a solution to the right
angle bracket nuisance. Join me for an overview of core C++09.
----------------------------------------------------------------------------------
So the thing keep in my mind is C++09. I want to confirm the time and
the standard's name, but I can get nothing from the website of C++
standard committees.

Do any one know the true?

 

[Bo Persson]

I expect the new standard will be release as soon as possible. I
have
read an article before at
http://www.devsource.com/c/a/Languages/C09-A-Glimpse-into-the-Future/
and it say that:
----------------------------------------------------------------------------------
Following the Portland meeting in October 2006, the C++
standardization committee set 2009 as the target date for C++09, the
next C++ standard. C++09 will include at least two major core
features: rvalue references and type concepts. Other core features
are
also included, such as automatic deduction of types from
initializers,
delegating constructors, nullptr, and even a solution to the right
angle bracket nuisance. Join me for an overview of core C++09.
----------------------------------------------------------------------------------
So the thing keep in my mind is C++09. I want to confirm the time
and
the standard's name, but I can get nothing from the website of C++
standard committees.

Do any one know the true?

Howard Hinnant is member of the committee, so if he says that 09 is no
longer possible, it would require a small miracle to prove him wrong.
Not that he would mind, I'm sure!

Like he says, the current draft is now "feature complete",
unfortunately meaning that a lot of features were added at the last
minute, attempting to reach the C++09 goal. However, some of the
additions seem to contradict each other and will require some "bug
fixes" before the new standard is complete.

Can this be done during this year? I don't know. Obviously, Howard
doesn't belive so.


Bo Persson

 

[coal]

0x0a or even 0x0b.  09 is no longer possible.  N2800 represents CD1
(first committee draft).  This should be viewed as "feature complete",
but definitely not bug free.

I'm sorry to hear that. I would like to suggest the development of
an on line compiler/code generator able to transform C++0x code into
code that can be handled by existing C++ compilers. If possible
that will make it easier to push C++0x applications to a wide
variety of platforms. I recall writing a while ago that I didn't
think there would be more than 4 or 5 compilers able to adequately
handle C++0x code by 2012 or 2013. That was before I heard this
and before the economy tanked. I thought my prediction was
realistic at the time, but now it seems too optimistic. I think
gcc would be one of those first compilers with C++0x support and yet,
that would mean continuing with the C status quo of gcc, thereby
turning gcc into a bigger black hole than it already is. I guess
that may happen but I pity the poor folks who would maintain it or,
G-d forbid, attempt to extend it to support new functionality.
Others have used the word crisis to describe the state of affairs
and I agree with that.


Brian Wood
Ebenezer Enterprises
www.webEbenezer.net

I recommend the articles on social topics by this physics professor --
http://wuphys.wustl.edu/~katz/

 

[Jerry Coffin]

Howard Hinnant is member of the committee, so if he says that 09 is no
longer possible, it would require a small miracle to prove him wrong.
Not that he would mind, I'm sure!

Like he says, the current draft is now "feature complete",
unfortunately meaning that a lot of features were added at the last
minute, attempting to reach the C++09 goal. However, some of the
additions seem to contradict each other and will require some "bug
fixes" before the new standard is complete.

Can this be done during this year? I don't know. Obviously, Howard
doesn't belive so.

I'm quite certain Howard's right. Even if the bugs are fixed this year
(it wouldn't surprise me if they are, at least mostly) the standard
can't be made official this year.

The simple fact is that there are a number of things (votes, mostly)
that require publishing the standard in its current state, and then
waiting a specific amount of time (IIRC, usually 90 days) to give people
a chance to read and respond before the next step is taken.

I added these up once, and IIRC, they work out to an absolute minimum of
a little over a year from the time the _last_ change is made until it
becomes an official standard. Since the last change has not been made
yet, there's no way it can happen this year.

 

[Juha Nieminen]

I'm quite certain Howard's right. Even if the bugs are fixed this year
(it wouldn't surprise me if they are, at least mostly)

I know nothing about the standardization process, but I can't
understand why fixing some "bugs" in a standard draft would take over a
year of time.

If fixing bugs in a program would take that long it would be rather
ridiculous.

 

[Ian Collins]

I know nothing about the standardization process, but I can't
understand why fixing some "bugs" in a standard draft would take over a
year of time.

If fixing bugs in a program would take that long it would be rather
ridiculous.

If the code were written by a large number of geographically distributed
part-timers who had to vote on changes in person three or four times a
year, it probably would !take that long!

 

[James Kanze]

Howard Hinnant is member of the committee, so if he says that
09 is no longer possible, it would require a small miracle to
prove him wrong. Not that he would mind, I'm sure!

Like he says, the current draft is now "feature complete",
unfortunately meaning that a lot of features were added at the
last minute, attempting to reach the C++09 goal. However, some
of the additions seem to contradict each other and will
require some "bug fixes" before the new standard is complete.

Can this be done during this year? I don't know. Obviously,
Howard doesn't belive so.

It's not really a question of what Howard believes (but what he
and every member of the committee knows): ISO has a very
definite procedure for adopting a standard. The current status
is that there is a committee draft, which is being voted on by
the national bodies. If all of the national bodies vote yes,
we'll take it exactly as it is, with no modifications or
corrections, it would still take nine or ten months of paperwork
in ISO before it was formally adopted. And realistically
speaking, I don't think any of the national bodies will go that
far---most (but I'm just guessing here) will probably vote yes,
with comments. The comments must be addressed before the draft
can be sent to ISO, and the nine or ten month delay starts
ticking.

My guess (but Howard is better placed than I am to judge this)
is that the best case would be that all of the comments are
handled at the meeting this July in Frankfurt, and the final
draft adopted then, with the standard being formalized by ISO
next spring. But it wouldn't surprise me if there were enough
serious issues that they couldn't all be handled in one meeting,
either.

 

[James Kanze]

I know nothing about the standardization process, but I can't
understand why fixing some "bugs" in a standard draft would
take over a year of time.

If fixing bugs in a program would take that long it would be
rather ridiculous.

The standard is, in some ways, a "law": it must be voted on, at
different levels. And the votes are by "national bodies", which
in turn have to meet, to decide how to vote. This is why the
delays, at least in part. The national body must receive the
document, then have time for its experts to study it and comment
on it, and then vote.

 

[Ivan]

Thanks for all your comments. Sorry for my lack knowledge about the
standard and process of C++ committee, because I am a new guy in this
group and also new in the official C++ "rule". I know more about the
standard and committee than before. Thanks again.

 

[Guest]

  I know nothing about the standardization process, but I can't
understand why fixing some "bugs" in a standard draft would take over a
year of time.

  If fixing bugs in a program would take that long it would be rather
ridiculous.

the problem is they don't write standards in a formal notation
(don't mention Algol-68, I think I did once, but I think I got
away with it!). If the C++ Standard were written in, say, C++
I'm sure they would be able to get it out of the door much quicker.
You'd be able to run gdb on it for a start!



--
Nick Keighley


"The use of the Chomsky formalism is also responsible for the term
"programming language", because programming languages seemed to
exhibit a strucure similar to spoken languages. We believe that
this term is rather unfortunate on the whole, because a programming
language is not spoken, and therefore is not a language in the true
sense of the word. Formalism or formal notation would have been
more appropriate terms."
Niklaus Wirth

 

[Juha Nieminen]

If the code were written by a large number of geographically distributed
part-timers who had to vote on changes in person three or four times a
year, it probably would !take that long!

Isn't that the case, for example, with the linux kernel? Yet bug fixes
come very fast when bugs are found. Usually in a matter of days (or even
faster).

 

[Juha Nieminen]

The standard is, in some ways, a "law": it must be voted on, at
different levels. And the votes are by "national bodies", which
in turn have to meet, to decide how to vote. This is why the
delays, at least in part. The national body must receive the
document, then have time for its experts to study it and comment
on it, and then vote.

Isn't this kind of bureaucracy a bit detrimental to the whole process?
What are the benefits?

 

[Ian Collins]

Isn't that the case, for example, with the linux kernel? Yet bug fixes
come very fast when bugs are found. Usually in a matter of days (or even
faster).

Do the maintainers have to vote on changes in person?

 

[Bo Persson]

Isn't this kind of bureaucracy a bit detrimental to the whole
process? What are the benefits?

For an international standard to be efficient, most nations with an
interest in the standard (represented by their national standard
bodies) must more or less agree that the standard is correct and
useful. Many contries will issue this as a national standard as well,
if they agree. What if they don't?



A bit of bureaucracy can also be useful to assure that some big
companies don't form their own committee to get its products approved
as a standard. Oops...



Bo Persson