hsutter@gotw.ca (Herb Sutter) wrote in message news:<70hflvs4o6rcalghbkjq18raguguu5gofv@4ax.com>...
> On Thu, 4 Sep 2003 21:19:20 +0000 (UTC), maisonave@axter.com (foo) wrote:
> >I've heard that the C++ commitee is considering adding hash_map to the
> >C++ standard.
>
> Pretty much -- more precisely, to the library technical report, which is
> non-normative but its contents are viewed as likely to be eventually
> considered for the standard itself someday. See also my "The New C++"
> articles, notably the just-published-today article:
>
>   "Trip Report: April 2003" [The New C++ #9]
>   C/C++ Users Journal, 21(10), October 2003
>   http://www.cuj.com/documents/s=8890/cujexp0310sutter/
>
> (I'm just about to put a link to that on my website and announce it to my
> mailing list.)
>
> >All three of the following freee STL libraries have hash_map
> >implementation:
> >
> >STLport
> >http://www.stlport.org/
> >
> >SGI
> >http://www.sgi.com/tech/stl/
> >
> >Dinkumware
> >http://www.dinkumware.com/
> >
> >The hash_map interface is not exactly the same.
>
> Correct.
>
> >Does any one know which of the above hash_map implementaions comes
> >closes to what the commitee is considering to add to the standard?
>
> See the latest document at the Library Technical Report page:
>
>   http://std.dkuug.dk/jtc1/sc22/wg21/docs/library_technical_report.html
>
> It's the last item, N1456. Enjoy!
>
> Herb
>
> ---
> Herb Sutter (www.gotw.ca)
>
> Convener, ISO WG21 (C++ standards committee)     (www.gotw.ca/iso)
> Contributing editor, C/C++ Users Journal         (www.gotw.ca/cuj)
> Visual C++ program manager, Microsoft      (www.gotw.ca/microsoft)
>
> ---
> [ 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.jamesd.demon.co.uk/csc/faq.html                       ]

Thank you for the great links.

I have Excpetional C++ and More Exceptional C++, and these two books
are two of my top 5 favorite C++ books.
Good Job!!

While I have your ear, in reading the links, I notice that they want
to change the hash_map name to unordered_map when they add it to the
standard.
Part of the justicfication given was that some vendors where putting
hash_map in the std namespace.

I thought one of the main points of having an std namespace, is so
that the space is reserved for the C++ standard objects, and that all
other objects should be placed out side of the std namespace.

IMHO, this is a poor justification.  I think vendors who are
incorrectly using this namespace for none C++ standard objects, should
have there code break when the standard adds objects of the same name.
This would teach them to keep these objects out of the std namespace.

If the commitee bends over backwards for these vendors, then the
vendors are going to continue to clutter the std namespace with none
C++ standard objects.

IMHO, this course of action is just promoting bad behavior.

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

maisonave@axter.com (foo) writes:
[snip]
> I thought one of the main points of having an std namespace, is so
> that the space is reserved for the C++ standard objects, and that all
> other objects should be placed out side of the std namespace.
>
> IMHO, this is a poor justification.  I think vendors who are
> incorrectly using this namespace for none C++ standard objects, should
> have there code break when the standard adds objects of the same name.
> This would teach them to keep these objects out of the std namespace.
>
> If the commitee bends over backwards for these vendors, then the
> vendors are going to continue to clutter the std namespace with none
> C++ standard objects.
>
> IMHO, this course of action is just promoting bad behavior.
[snip]

Naming the new unordered associative containers hash_* , along the
    lines of existing hashed containers, would break far more user
    code than implementor code, since the new containers have
    different semantics. The commitee's decision IMO aids
    you and I more than it aids the implementors.

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

On Thu, 4 Sep 2003 21:19:20 +0000 (UTC), maisonave@axter.com (foo) wrote:
>I've heard that the C++ commitee is considering adding hash_map to the
>C++ standard.

Pretty much -- more precisely, to the library technical report, which is
non-normative but its contents are viewed as likely to be eventually
considered for the standard itself someday. See also my "The New C++"
articles, notably the just-published-today article:

  "Trip Report: April 2003" [The New C++ #9]
  C/C++ Users Journal, 21(10), October 2003
  http://www.cuj.com/documents/s=8890/cujexp0310sutter/

(I'm just about to put a link to that on my website and announce it to my
mailing list.)

>All three of the following freee STL libraries have hash_map
>implementation:
>
>STLport
>http://www.stlport.org/
>
>SGI
>http://www.sgi.com/tech/stl/
>
>Dinkumware
>http://www.dinkumware.com/
>
>The hash_map interface is not exactly the same.

Correct.

>Does any one know which of the above hash_map implementaions comes
>closes to what the commitee is considering to add to the standard?

See the latest document at the Library Technical Report page:

  http://std.dkuug.dk/jtc1/sc22/wg21/docs/library_technical_report.html

It's the last item, N1456. Enjoy!

Herb

---
Herb Sutter (www.gotw.ca)

Convener, ISO WG21 (C++ standards committee)     (www.gotw.ca/iso)
Contributing editor, C/C++ Users Journal         (www.gotw.ca/cuj)
Visual C++ program manager, Microsoft      (www.gotw.ca/microsoft)

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

Good Morning, "foo" (Whatever...)

foo schrieb:

> I've heard that the C++ commitee is considering adding hash_map to the
> C++ standard.
> All three of the following freee STL libraries have hash_map
> implementation:
>
> STLport
> http://www.stlport.org/
>
> SGI
> http://www.sgi.com/tech/stl/
>
> Dinkumware
> http://www.dinkumware.com/
>
> The hash_map interface is not exactly the same.
>
> Does any one know which of the above hash_map implementaions comes
> closes to what the commitee is considering to add to the standard?

Maybe you should have a look at the current state of the proposal:

http://std.dkuug.dk/jtc1/sc22/wg21/docs/papers/2003/n1456.html

Greetings from Bremen,

Daniel Spangenberg






---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

In article <c11783b0.0309031220.6fb5fab0@posting.google.com>, foo wrote:
> I've heard that the C++ commitee is considering adding hash_map to the
> C++ standard.

The pending additions to the library are part of a Technical Report which
does not have standard status; they will probably be added to the next
major version of the standard but that will be considered later.

Here's the last published version of the hash tables proposal:
<http://std.dkuug.dk/jtc1/sc22/wg21/docs/papers/2003/n1456.html>.

The hash table based containers were accepted for the Technical Report,
but with new names substituting "unordered" for "hash" to avoid
conflicts with the pre-standard versions.

> All three of the following freee STL libraries have hash_map
> implementation:
>
> STLport
> http://www.stlport.org/

This implementation is derived from and similar to SGI's, as is that in
GNU libstdc++.

> SGI
> http://www.sgi.com/tech/stl/
>
> Dinkumware
> http://www.dinkumware.com/

Add Metrowerks to that.  Their library documentation is available at
<http://www.metrowerks.com/MW/Support/dev_resources/msl.htm>.

> The hash_map interface is not exactly the same.
>
> Does any one know which of the above hash_map implementaions comes
> closes to what the commitee is considering to add to the standard?

It seems to be closest to the SGI implementation, fairly close to
Metrowerks's and furthest from Dinkumware's.

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

maisonave@axter.com (foo) wrote in message news:<c11783b0.0309031220.6fb5fab0@posting.google.com>...
> I've heard that the C++ commitee is considering adding hash_map to the
> C++ standard.
> All three of the following freee STL libraries have hash_map
> implementation:
>
> STLport
> http://www.stlport.org/
>
> SGI
> http://www.sgi.com/tech/stl/
>
> Dinkumware
> http://www.dinkumware.com/
>
> The hash_map interface is not exactly the same.
>
> Does any one know which of the above hash_map implementaions comes
> closes to what the commitee is considering to add to the standard?

See http://std.dkuug.dk/jtc1/sc22/wg21/docs/papers/2003/n1456.html

For a list of all proposals accepted so far for the Library Technical
Report, see http://std.dkuug.dk/jtc1/sc22/wg21/docs/library_technical_report.html.

--Beman Dawes

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

"Herb Sutter" <hsutter@gotw.ca> wrote in message
news:70hflvs4o6rcalghbkjq18raguguu5gofv@4ax.com...
>
> See the latest document at the Library Technical Report page:
>
>   http://std.dkuug.dk/jtc1/sc22/wg21/docs/library_technical_report.html
>
> It's the last item, N1456. Enjoy!
>
There isn't much to enjoy there, is there?  There's the stupid bucket stuff
(which seems to run completely contrary to any other container by exposing
implementation details like buckets -- if you're going to do that, how about
you let me access maps as the red-black trees they're almost always
implemented as?  Will we have buckets for deques, as they too are normally
segmented storage?  What about strings?  Strings would be allowed to be
segmented storage (I doubt any are, but it doesn't seem to be forbidden), so
should they have the bucket interface too?).  There's crappy iterator
requirements (if the only thing guaranteed is forward iterators then the only
thing I can use is forward iterators, even if my implementation does the right
thing and provides bidi iterators).


---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]

I've heard that the C++ commitee is considering adding hash_map to the
C++ standard.
All three of the following freee STL libraries have hash_map
implementation:

STLport
http://www.stlport.org/

SGI
http://www.sgi.com/tech/stl/

Dinkumware
http://www.dinkumware.com/

The hash_map interface is not exactly the same.

Does any one know which of the above hash_map implementaions comes
closes to what the commitee is considering to add to the standard?

---
[ 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.jamesd.demon.co.uk/csc/faq.html                       ]