In article <1994Oct5.050320.15666@nlm.nih.gov> bkline%occs.nlm.nih.gov (Bob Kline Phoenix Contract) writes:
>Steve Clamage (clamage@Eng.Sun.COM) wrote:
>: >
>: Lead times in book publishing are long, and I would imagine that P. J.
>: was using his crystal ball to predict the state of things when the book
>: was to appear in print. If he reads this discussion, I hope he will
>: comment.
>
>Thanks for the clarification.  The "crystal ball" theory is a little
>weak, though, in light of reference to "the publication, in February
>1994, of the draft C++ Standard for informal public review."  Publishing
>lead times are indeed long, but not _that_ long.  Perhaps the various
>members of the committees differ in their opinion of what the terms
>should mean?

 The WP WAS "published" in Feb 1994 for informal
review. But NOT for public review. There is NO public review
in the ISO processes at all. The Feb 1994 publication was
for ISO member National Bodies to peruse.

 Similarly, the proposed Committee Draft has been
frozen and will be "published" by ISO any time now,
and distributed to National Bodies through the usual
ISO channels. Again: no public review. There is NO public review.

 The public review involved is the ANSI review.
ANSI is just one National Body, and is free to conduct
a public review if it wants. The C++ committee as a whole
will adapt to the ANSI processes "informally": WG21 members
of the committee are interested in the ANSI public review.

 BTW: Plauger's book, as usual, is worth having.

--
        JOHN (MAX) SKALLER,         INTERNET:maxtal@suphys.physics.su.oz.au
 Maxtal Pty Ltd,
        81A Glebe Point Rd, GLEBE   Mem: SA IT/9/22,SC22/WG21
        NSW 2037, AUSTRALIA     Phone: 61-2-566-2189

Steve Clamage (clamage@Eng.Sun.COM) wrote:
: >
: >: There isn't a "draft standard" yet, only the Committee Working Paper.
: >
[snip]
: >
: Few people in the world understand the details of the standardization
: process as well as P. J. Plauger, and he certainly does better than I.

: Nonetheless, the C++ standardization process has not yet reached the
: "[Dd]raft [Ss]tandard" stage. The term is a technical one in the
: ANSI standardization sequence (capitalization aside). It means that
: certain registration and voting procedures have taken place.

: To further complicate matters, the C++ Committee is a joint ANSI and ISO
: committee. ANSI and ISO have different rules and terminology. The C++
: committee is now an ANSI "Type I" committee, a concept introduced by ANSI
: to allow coordination with ISO. ISO terminology and sequences will be
: followed by the committee, in so far as I understand the situation.

: Lead times in book publishing are long, and I would imagine that P. J.
: was using his crystal ball to predict the state of things when the book
: was to appear in print. If he reads this discussion, I hope he will
: comment.

Thanks for the clarification.  The "crystal ball" theory is a little
weak, though, in light of reference to "the publication, in February
1994, of the draft C++ Standard for informal public review."  Publishing
lead times are indeed long, but not _that_ long.  Perhaps the various
members of the committees differ in their opinion of what the terms
should mean?

Bob Kline

Steve Clamage (clamage@Eng.Sun.COM) wrote:

: There isn't a "draft standard" yet, only the Committee Working Paper.

[snip]

: You can read more about IOstreams in the books "C++ IOstreams Handbook" by
: Steve Teale, Addison-Wesley, and "The Standard C++ Library", P. J. Plauger,
: Prentice-Hall.

The first sentence of the second book recommended above (whose title
actually was "The Draft Standard C++ Library" when it finally appeared)
reads: "This book shows you how to use the library classes and functions
mandated by the draft ANSI/ISO Standard for the programming language C++."

So who's right, Clamage or Plauger?  Is there or is there not a draft
standard?  The next paragraph of Plauger's book begins: "The watershed
event was the publication, in February 1994, of the draft C++ Standard
for informal public review."  Are the committees splitting hairs so
finely that "draft standard" and "draft Standard" (differing only in
capitalization) have two different meanings?

Bob Kline

In article <369vfk$bbh@engnews2.Eng.Sun.COM>, clamage@Eng.Sun.COM wrote:
> The IOstreams section has been significantly different in each WP issue
> (three per year). If you want to keep current, you could join the Committee,
> possibly as an observer, and get the mailings.

How does one go about joining the Committee?

> You should be aware that because of the rapid and significant changes in the
> IOstreams section, no vendor has attempted to keep up with it.

Actually, I'm more interested in implementation.

_____________________________________________________________________________
Shahid M. Alam                                         shahid@mail.utexas.edu

bkline%occs.nlm.nih.gov (Bob Kline Phoenix Contract) writes

 > Steve Clamage (clamage@Eng.Sun.COM) wrote:
 >
 > : There isn't a "draft standard" yet, only the Committee Working Paper.
 >
 > [snip]
 >
 > : You can read more about IOstreams in the books "C++ IOstreams Handbook" by
 > : Steve Teale, Addison-Wesley, and "The Standard C++ Library", P. J. Plauger,
 > : Prentice-Hall.
 >
 > The first sentence of the second book recommended above (whose title
 > actually was "The Draft Standard C++ Library" when it finally appeared)
 > reads: "This book shows you how to use the library classes and functions
 > mandated by the draft ANSI/ISO Standard for the programming language C++."
 > So who's right, Clamage or Plauger?  Is there or is there not a draft
 > standard?  The next paragraph of Plauger's book begins: "The watershed
 > event was the publication, in February 1994, of the draft C++ Standard
 > for informal public review."  Are the committees splitting hairs so
 > finely that "draft standard" and "draft Standard" (differing only in
 > capitalization) have two different meanings?

Certainly standards committees split hair so fine that can confuse anyone
(not excluding me). Worse, the powers who control standards activities are
even better at that, and (like all real standards bodies) they change the
rules in mid stream. Thus, the rules for creating an ISO standard with the
cooperation of national standards bodies such as ANSI, AFNOR, and BSI are
not easy to fathom and in flux.

However, I'm pretty sure that no ``Draft C++ Standard'' exist, and that
none will exist until sometime after a meeting in Austin TX March 1995.

At the meeting in March 1994 in San Diego, we for the first time approved a
working paper containing all the parts of a proper standard and on the proper
format for a proper (the standards bureacraties are VERY keen on such aspects
of standards). I suspect that must be what PJP is referring to.

At the meeting in July 1994 in Waterloo (Canada) we voted to submit our
working paper (note: note yet draft standard) for CD registration (CD is
an aconym for Comitteee Draft). This means that the ISO folks get to verify
that the working paper is on the right form and an assertion by the standards
committee that all major components of the language is in place, though we
do not promise that details won't change. Note the scope for disagreement
of how to define `major' and `detail' in that sentense. To confuse matters
further, the various standards bureacraties have already requested and
received informal copies of our working paper and we have received some
informal feedback.

Assuming that all goes well with the CD registration and that the comittee's
technical work progress as we hoped it will, we will vote on submitting a
CD draft at the Austin TX March 1995 meeting. If that vote passes, the working
paper will be revised with the decisions voted in in Austin and the result will
be passed on to the national standards bodies for the formal vote. Yes, the
people who meet, do the technical work, and vote, don't REALLY vote, they
simply forward recommendations to the official national standards bodies
who do the real voting.

The San Diego WP was pretty good, but - as expected - far from being acceptable
as a standard. The current WP (post-Waterloo, dated September 1994) is much
better and the first WP that can be relied on to give a good idea of what the
language and library will look like. I wouldn't trust it on details, though.
Many details like the exact set of library classes or the exact set of member
functions for a given library class will change. Personally, I am quite conservative
in what aspects of iostreams, etc., I'm currently willing to base real code on.
Minimal use of features is the key to safety just now.

The post-Austin WP (March 1995) will be the first to be officially called a
draft and the first to be submitted to an official public review. In the meantime,
comments on the working paper - especially constructive ones - are as usual
wellcome. Please note, though, that the committee has voted to avoid major
changes (additions, modifications, deletions) for a while to work on the WP.
Between now and Austin, we might change all the members of class string (that
would be making changes of details), but we will not add persistence to the
language (that WOULD be a major event).

 - Bjarne

Take all this with a caveat: ``this is true and accurate to the best of my
knowledge, but the rules of standardization are subtle, capritious, and
subject to immediate change - though reputedly not malicious.''

In article 12962@nlm.nih.gov, bkline%occs.nlm.nih.gov (Bob Kline Phoenix Contract) writes:
>Steve Clamage (clamage@Eng.Sun.COM) wrote:
>
>: There isn't a "draft standard" yet, only the Committee Working Paper.
>
>[snip]
>
>: You can read more about IOstreams in the books "C++ IOstreams Handbook" by
>: Steve Teale, Addison-Wesley, and "The Standard C++ Library", P. J. Plauger,
>: Prentice-Hall.
>
>The first sentence of the second book recommended above (whose title
>actually was "The Draft Standard C++ Library" when it finally appeared)
>reads: "This book shows you how to use the library classes and functions
>mandated by the draft ANSI/ISO Standard for the programming language C++."
>
>So who's right, Clamage or Plauger?  Is there or is there not a draft
>standard?  The next paragraph of Plauger's book begins: "The watershed
>event was the publication, in February 1994, of the draft C++ Standard
>for informal public review."  Are the committees splitting hairs so
>finely that "draft standard" and "draft Standard" (differing only in
>capitalization) have two different meanings?

Few people in the world understand the details of the standardization
process as well as P. J. Plauger, and he certainly does better than I.

Nonetheless, the C++ standardization process has not yet reached the
"[Dd]raft [Ss]tandard" stage. The term is a technical one in the
ANSI standardization sequence (capitalization aside). It means that
certain registration and voting procedures have taken place.

To further complicate matters, the C++ Committee is a joint ANSI and ISO
committee. ANSI and ISO have different rules and terminology. The C++
committee is now an ANSI "Type I" committee, a concept introduced by ANSI
to allow coordination with ISO. ISO terminology and sequences will be
followed by the committee, in so far as I understand the situation.

Lead times in book publishing are long, and I would imagine that P. J.
was using his crystal ball to predict the state of things when the book
was to appear in print. If he reads this discussion, I hope he will
comment.

---
Steve Clamage, stephen.clamage@eng.sun.com

Greetings,

I'm looking for some documentation on IOStreams as currently defined in
the draft standard.  Any help will be greatly appreciated.

_____________________________________________________________________________
Shahid M. Alam                                         shahid@mail.utexas.edu

In article 2709941343260001@catalpa.lib.utexas.edu, shahid@mail.utexas.edu (Shahid Alam) writes:
>Greetings,
>
>I'm looking for some documentation on IOStreams as currently defined in
>the draft standard.  Any help will be greatly appreciated.

There isn't a "draft standard" yet, only the Committee Working Paper.

The IOstreams section has been significantly different in each WP issue
(three per year). If you want to keep current, you could join the Committee,
possibly as an observer, and get the mailings.

You should be aware that because of the rapid and significant changes in the
IOstreams section, no vendor has attempted to keep up with it. If you write
code to some specific version of the IOstream section of the WP, it is
unlikely to compile or run correctly for that reason, unless you stick with
the subset conforming to the de-factor standard of the original AT&T release
of IOstreams.

You can read more about IOstreams in the books "C++ IOstreams Handbook" by
Steve Teale, Addison-Wesley, and "The Standard C++ Library", P. J. Plauger,
Prentice-Hall.

---
Steve Clamage, stephen.clamage@eng.sun.com