Michiel Salters<Michiel.Salters@cmg.nl> wrote in message news:<YKYX6.2229$yp1.58532@www.newsranger.com>...
> In article <E15C05n-000ET0-00@f6.mail.ru>, Kiril Avdeiv says...
> >
> >It is no surprize to all advanced users of C++ that some chapters of the ISO
> > 14882 standard (14 templates for instance) are still in the draft state.
>
> It would be a surprise. They are in an inperfect state, but draft? And entire
> chapters? Some parts may not end up as intended. That can be a defect.
>
> >Defect reports are counted by hundrends, and some of them have a significant
> >impact on the language. The parts of the language that are most flawed are
> > the whole chapter 14, Koeing lookup 3.4.2 and the like.
>
> As in, the newest parts are the most flawed cq. the least perfect. So?
>
> >Just take a look at:
>
> [ SNIP ]
>
> >http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_defects.html#45,
> > which changes access checking to a way having little in common with
> > the current state of the standard.
>
> >Well, why has the language been ratified in such a raw state, instead of
> > keeping it in draft state until it more or less stabilize?
>
> Because it good enough. Making it perfect will mean making it obsolete,
> because the time it takes for perfection is infinite.

Well, this is merely an opinion of yours. The facts are:
1. There are hunderds of defect reports.
2. It is impossible to implement the language based on the ISO 14882.
Because the latter doesn't even provide a definitive definition of the
language.

If you were to implement a simplistic C++ parser, numerous questions
with no answers to would come over you, as soon as you encountered a
non trivial template-declaration or an explicit-specialization. Just
give it a try and see that it works...

> >Why has the disclaimer so proudly shining on the cover of the draft version
> > as of 2 December 1996 and candidly saying that "This document is known to
> > be... incorect, incomplet, and inConsiStent" has been removed?
>
> Because no one was able to convince even a single country that critical
> parts were missing/incorrect/etc.

That's again an opinion, for you were not there. And I'm very much
saddened by seeing opinions disguised as facts.

Now, my opinion is that the reason was that no vendor was even close
enough to implementing the language even as stated in CD2. So no one
was able to try out in practice whether everything from the 14-th
chapter or 3.4.2 was making sense. And ratifying the standard in that
state was a compromise.

> >Are there any plans to revive the disclaimer for the next revision of
> >the standard.
>
> Sure, all drafts of C++0x will have something similar.

Again merely an opinion...! Why?

> >Hope to hear an answer and not to have this honest posting rejected.
> >Kiril
>
> The value of any standard, C++ included, is in the quality it adds to the
> products made. We should judge the C++ standard by how much change is needed
> to port real-world programs, before and after the standard. Using this
> quality measure, I'd say the standard is complete enough for its purpose.

I didn't notice a drastic improvement in the quality of
implementations, really, even the EDG front-end, I've found some bugs
in it. Not always bugs though, just things that are ambiguous or
inconsistent but have to be interpreted somehow since the grammar
allows for them.

> If it is complete, why are we extending it? Why C++0x? Because the purpose
> of C++98 was to enable compiler writers to make C++ compilers now, not in
> 2008, and the purpose of C++0x is so they can make them in 2008. Different
> purposes, different standards.

My final opinion is that you might want to read our exchange of ideas
with James Kanze on comp.lang.c++.moderated, the thread being called
"C++ standard facets in locals".

Thank you
Kiril

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.research.att.com/~austern/csc/faq.html                ]

In article <E15C05n-000ET0-00@f6.mail.ru>, Kiril Avdeiv says...
>
>It is no surprize to all advanced users of C++ that some chapters of the ISO
> 14882 standard (14 templates for instance) are still in the draft state.

It would be a surprise. They are in an inperfect state, but draft? And entire
chapters? Some parts may not end up as intended. That can be a defect.

>Defect reports are counted by hundrends, and some of them have a significant
>impact on the language. The parts of the language that are most flawed are
> the whole chapter 14, Koeing lookup 3.4.2 and the like.

As in, the newest parts are the most flawed cq. the least perfect. So?

>Just take a look at:

[ SNIP ]

>http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_defects.html#45,
> which changes access checking to a way having little in common with
> the current state of the standard.

Classes can contain classes. Functions can't contain classes. That is a
fundamental difference. If access checking follows this form, has the same
fundamental difference, is that a bad thing?

>Well, why has the language been ratified in such a raw state, instead of
> keeping it in draft state until it more or less stabilize?

Because it good enough. Making it perfect will mean making it obsolete,
because the time it takes for perfection is infinite.

>Why has the disclaimer so proudly shining on the cover of the draft version
> as of 2 December 1996 and candidly saying that "This document is known to
> be... incorect, incomplet, and inConsiStent" has been removed?

Because no one was able to convince even a single country that critical
parts were missing/incorrect/etc.

>Are there any plans to revive the disclaimer for the next revision of
>the standard.

Sure, all drafts of C++0x will have something similar.

>Hope to hear an answer and not to have this honest posting rejected.
>Kiril

The value of any standard, C++ included, is in the quality it adds to the
products made. We should judge the C++ standard by how much change is needed
to port real-world programs, before and after the standard. Using this
quality measure, I'd say the standard is complete enough for its purpose.

If it is complete, why are we extending it? Why C++0x? Because the purpose
of C++98 was to enable compiler writers to make C++ compilers now, not in
2008, and the purpose of C++0x is so they can make them in 2008. Different
purposes, different standards.

Regards,
Michiel Salters

--
Michiel Salters
Consultant Technical Software Engineering
CMG Trade, Transport & Industry
Michiel.Salters@cmg.nl

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.research.att.com/~austern/csc/faq.html                ]

It is no surprize to all advanced users of C++ that some chapters of the ISO 14882 standard (14 templates for instance) are still in the draft state.

Defect reports are counted by hundrends, and some of them have a significant impact on the language. The parts of the language that are most flawed are the whole chapter 14, Koeing lookup 3.4.2 and the like.

Just take a look at:

http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_defects.html#90
http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_defects.html#213

http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_defects.html#45, which changes access checking to a way having little in common with the current state of the standard.

or

http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/cwg_active.html#224, which seems to suggest an absolutely different interpretation of dependent names.

Well, why has the language been ratified in such a raw state, instead of keeping it in draft state until it more or less stabilize?

Why has the disclaimer so proudly shining on the cover of the draft version as of 2 December 1996 and candidly saying that "This document is known to be... incorect, incomplet, and inConsiStent" has been removed?

Are there any plans to revive the disclaimer for the next revision of the standard.

Hope to hear an answer and not to have this honest posting rejected.

Kiril





--
Posted from 172.16.27.62, 172.16.17.153 via proxy [212.72.53.70] by way of mx10.port.ru [194.67.23.89]
via Mailgate.ORG Server - http://www.Mailgate.ORG

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.research.att.com/~austern/csc/faq.html                ]