This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: C++ PATCH: PR 20599 (1/3)


Doug Gregor <dgregor@cs.indiana.edu> writes:

| On Sep 25, 2006, at 2:42 PM, Richard Guenther wrote:
| > This practically means the compiler should just provide _one_
| > -std=c++0x-draft (or whatever name) and _not_ provide preprocessor
| > macros for features included in the particular release.  This avoids
| > giving false promises about stability of these language extensions.
| 
| Given that we're only considering features that have been written
| into the C++0x Working Draft for this -std=c++0x(-draft?) mode, I
| imagine that the chance of features being removed is rather small.

yet non-existent.  I've seen C++98 seriously reworked at the latest
minutes (with treat of "no" vote) -- that gave interesting components
in the library.

| 50
| + members of the committee have said "yes" to a feature at least two
| times, usually with more than a year in between, before it gets into
| the working draft. It would take a significant event for them to
| change their minds.

The way the standard process works is roughly:

  (1) you have public reviews that are seriously considered (in the
      previous round, the WP are twice amended).  The call for public
      review will happen soon (i.e. in a year of so) I hope.

  (2) national bodies, and only those, vote to approve the standard --
      no matter how many committee members were attending the meetings.

Those points should be seriously considered.  The deal is not done,
just because it is in the WP.  You still need the voting national
members to agree.  

-- Gaby


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]