Document number: N2142=07-0002

Alisdair Meredith

State of C++ Evolution (between Portland and Oxford 2007 Meetings)

This paper presents a quick summary of all papers submitted to WG21 for consideration by the Evolution Working Group, with an indication of their progress towards inclusion in the next C++ Standard.

There are no library issue here, unless highlighted by the Registration Document submitted following the Portland meeting. Otherwise this list focuses entirely on changes affecting clauses 1 -> 16.

A note on the timetable

The whole committee expressed a strong desire to deliver the next C++ Standard in 2009. In order to meet ISO timetables that will mean voting out a Final Candidate Document at the end of 2007. That leaves two more meetings to finish the EWG work, and finalise all wording. Therefore, next version of this paper will not have a Stalled section - anything stalled at the end of the next meeting is too late. Likewise, pressure will build very quickly on CWG as papers receive final wording and require review. We have seen very few papers go through review without some revision, often several revisions. The more we can do to supply papers with complete wording in advance of a meeting the better. I would hope the Blessed By Evolution - Wording Required section could be cleared between meetings. If an offline review of proposed wording with a member of Core can be arranged in advance of the meeting, that should help identify first-order problems so there is a much better chance of the paper making its final revision at the meeting, in order for Core to move forward into the Working Paper.

Key to the table

Papers have been grouped according to theme, and related papers joined into the same row, even when submitted by different authors. The intent is to capture the feature, rather than the workflow. Papers are ordered by theme, and themes ordered by document number for the first submission on that theme.

Note that while the deadline for new proposals for C++09 is long past, it is not unusual to split a specific feature out of an existing proposal, to ease their progress. Likewise, existing Core Working Group issues may grow large enough to merit a paper that would not have previously shown on this list.

Proposals in a green font have advanced a category since the last meeting, those in red have moved back, while those in yellow are new or have been updated since the last meeting.

Issues with a green background have been highlighted as the features that will be delivered in the next version of the standard in the Registration Document.

WG21 Number(s) Title Authors
Integrated into working paper
These propsals are already incorporated into the latest working paper, N2134, or the project editor has been directed to include them in the next draught.
N1377 N1385 N1690 N1770 N1855 N1952 N2118 A Proposal to Add an Rvalue Reference to the C++ Language H. Hinnant, P. Dimov, D. Abrahams, J. Adamczyk, A. Hommel
N1381 N1604 N1617 N1720 static_assert R. Klarer, J. Maddock, B. Dawes, H. Hinnant
N1448 N1960 N1987 Extern template Matt Marcus, Garbriel Dos Reis, John Spicer
N1520 N1616 N1722 N1791 Extended friend Declarations William M. Miller
N1545 N1566 N1653 Synchronizing the C++ preprocessor with C99 Clark Nelson, Paul Mensonides
N1564 N1627 "Conditionally-Supported Behavior" William M. Miller
N1727 Changing Undefined Behavior into Diagnosable Errors William M. Miller
N1565 N1693 N1735 N1811 Adding the long long type to C++ J. Stephen Adamczyk
N1746 N1988 Adding extended integer types to C++ J. Stephen Adamczyk
N1581 N1618 N1895 N1986 Delegating Constructors H. Sutter, F. Glassborow
N1610 Clarification of Initialization of Class Objects by rvalues D. Abrahams, G. Powell
N1649 N1699 N1757 Right Angle Brackets Daveed Vandevoorde
N1721 N1794 N1894 N1984 Deducing the type of variable from its initializer expression J. Jaarvi, B. Stroustrup, G. Dos Reis
N1737 A Proposal to Restore Multi-declarator auto Declarations Walter E. Brown
Proposed wording under review in Core
These papers are undergoing final scrutiny in Core Working group, and have all been reviewed at least once. Full wording is available, and has been reviewed at least once by the CWG. However, some draughting issues remain to be resolved before moving into the Working Paper.
N1406 N1449 N1451 N1489 N2112 Template aliases for C++ B. Stroustrup, G. Dos Reis, Mat Marcus, Walter E. Brown, Herb Sutter
N1478 N1527 N1607 N1705 N1978 N2115 Decltype (and auto) J. Järvi, B. Stroustrup, D. Gregor, J. Siek, G. Dos Reis
N1521 N1972 N1980 N2116 Generalized Constant Expressions Gabriel Dos Reis
N1628 N1823 N1955 N2018 New Character Types in C++ Lawrence Crowl
N1944 N2052 Sequencing and the concurrency memory model Clark Nelson, Hans Boehm
Blessed by evolution - wording available for initial review by Core
The design of these features is deemed complete and accepted by EWG, and there is a paper with complete wording for Core to review. However, CWG have not yet devoted any committee time to this issue. Typically papers in this state have been accepted by EWG at the previous meeting, or the version with proposed wording was not avaiable by the pre-meeting mailing.
N1488 N1601 A name for the null pointer: nullptr H. Sutter, B. Stroustrup
N1513 N1579 N1719 Strongly Typed Enums H. Sutter, D. Miller
N1534 N1642; N1970 Proposed addition of __func__ predefined identifier from C99 Alex Rosenberg; Alisdair Meredith
N1546 N1877 N1971 N2140 Adding Alignment Support to the C++ Programming Language Attila (Farkas) Fehér
N1583; N1898 N2119 Inheriting Constructors F. Glassborow; M. Michaud, M. Wong, A. Meredith, J. Maurer
N1592 Explicit Conversion Operators Lois Goldthwaite
Blessed by Evolution - Wording required for Core review
The design of these features is deemed complete and accepted by EWG, but there is not yet a paper with complete wording for Core to review. Complete wording will include specific directions for the project editor to apply directly to the Working Paper.
N1483 N1603 N1704 N2080 Variadic Templates D. Gregor, J. Järvi, G. Powell
N1493 N1509 N1890 N1919 N2100 Initializer lists Bjarne Stroustrup, G. Dos Reis, Daniel F. Gutson
N1528 Syntactic Disambiguation Using the Template Keyword John H. Spicer
Active topics in Evolution
These topics are still actively being developed, and there is a clear intention to incorporate each one in the next standard.
N1400 N1418 N1428 N1496 Draft Proposal for Dynamic Libraries in C++ Pete Becker, Matt Austern
N2117 Minimal Dynamic Library Support Lawrence Crowl
N1401; N2016 N2047 Atomic operations with multi-threaded environments Jan Kristoffersen; Nick McLaren, Hans Boehm
N1874 N1966 Thread-Local Storage Lawrence Crowl
N2090; N2094 A minimal threading API for C++0x Peter Dimov; Howard Hinnant
N2061 N2106; N2096; N2107 Propagating exceptions when joining threads Beman Dawes; Peter Dimov; Alisdair Meredith, Jens Maurer
N1408 N1691 Explicit Namespaces David Abrahams
N1792 N1893 N1912 N2103 A Modest Proposal: Fixing ADL Herb Sutter, Daveed Vandevoorde
N1494; N1827 N2108 An Explicit Override Syntax for C++ Daniel F. Gutson; C. Uzdavinis, A. Meredith, Jens Maurer
N1511 N1892 Extensible Literals I. McIntosh, M. Wong, R. Mak, Bjarne Stroustrup
N1796 N1868 N1961; N2049 Proposal for new for-loop Thorsten Ottosen; Doug Gregor
N1833 N1943 N2128 N2129 Transparent Garbage Collection for C++ H. Boehm, M. Spertus
N1958; N1968 Lambda expressions and closures for C++ V Samko; J Willcock, J Järvi, D Gregor, A Lumsdaine
N2042 N2081 Concepts (unified proposal) Doug Gregor, Bjarne Stroustrup
N2098 Scoped Concept maps Jeremy Siek
N2053 Raw String Literals Beman Dawes
N2062 N2102 PODs unstrung Beman Dawes
N2105 Choosing Keywords Beman Dawes
N2123 Adding the prohibited access specifier to C++09 Alisdair Meredith
Stalled - no progress for at least one meeting
These topics have all been considered by Evolution at some point, and have not yet actively rejected. However, there has been no EWG committee time spent on them for at least one meeting, and unless an extremely strong case is presented at the next meeting, they will all automatically drop out of consideration for C++09 at the next meeting.
A strong case should at the very least include proposed wording that might be applied directly to the Working Paper, and ideally a reference implementation. It is also recommended to contact the EWG chair before doing further work, to be sure that the EWG has resources to review/approve the proposal at the next meeting.
N1427 N1945 Making Local Templates more Useful Anthony Williams
N1524 Nested Namespace Definition Proposal Jon Jagger
N1526 Proposal to add namespace references to C++ Benjamin Kosnik
N1611 Implicitly-Callable Functions in C++0x Walter E. Brown
N1614 N1625 N1726 N1740 Macro scopes Thomas Plum, Bjarne Stroustrup
N1664; N1703 Toward Improved Optimization Opportunities in C++0x W. Brown,M. Paterno; Francis Glassborow
N1671 Overloading Operator.() & Operator.*() G. Powell, D. Gregor, J. Jarvi
N1676 Non-member overloaded copy assignment operator Bronek Kozicki
N1706 N1891 N2141 Toward Opaque typedefs in C++0X Walter E. Brown; Alisdair Meredith
N1784 N1821 Extending Move Semantics To *this (Revision 2) D. Vandevoorde, B. Kozicki
N1785 Toward a Proposal for Object Templates in C++0x Walter E. Brown
Heading for a separate TR
These topics are deemed too important to wait for another standard after C++09 before being published, but too experimental to be finalised in time for the next Standard. Therefore, these features will be delivered by a technical report at the earliest opportunity.
N1567 N1776 N1839 N1965 N1977 Decimal Types for C++ Robert Klarer, P.J. Plauger
N1736 N1778 N1964 N2006 N2015 N2073 N2074 Modules in C++ Daveed Vandevoorde
Not ready for C++09, but encourage work to continue
Papers in this category have been reviewed in EWG, and seen to solve real problems. While it is hoped that work will continue, they are not ready to be finalised in time for C++09. It is anticipated that this category will grow as time pressure restricts the feature list for C++09.
N1471 Reflective Metaprogramming in C++ Daveed Vandevoorde
N1751; N1775 Reflection in C++ Detlef Vollmann; Walter E. Brown, et al.
No active interest today, but open to resubmit in future
These papers were presented, but failed to spark sufficient interest. There is no reason these topics should not be resubmitted in the future, but there is no intention to revisit these topics either. Any future submission should include a strong motivation and possible futher use cases that were not considered first time around.
N1384; N1600; N1615 PME: Properties, methods and events John Wiegley; Daveed Vandevoorde; Lois Goldthwaite
N1440 Proposed Addition to C++: Class Namespaces Carl Daniel
N1445 N1582 N1702 N1717 Explicit class and default definitions F. Glassborow, L. Goldthwaite
N1461 Security and Standard C Libraries Martyn Lovell
N1462 Safe Exceptions and Compiler Security Checks Brandon Bray
N1463 N1529 Draft proposal for adding Multimethods to C++ Julian Smith
N1466 Expliciting default parameters Daniel Gutson
N1470; N1741 Extending / checking Switch statements Daniel Gutson; Francis Glassborow
N1585 N1716 N1742 Uniform Calling Syntax (Re-opening public interfaces) Francis Glassborow
N1602 Class Scope Using Declarations & private Members Francis Glassborow
N1613 N1669 N1773 N1800 N1866 N1867 N1962 Contract Programming For C++0x L. Crowl, T. Ottosen, D. Abrahams, J. Widman
N1694 A Proposal to Extend the Function Call Operator Peter Dimov
N1695 A Proposal to Make Pointers to Members Callable Peter Dimov
N1696 Language Support for Restricted Templates Peter Dimov
N1825 Addressing Exception Specifications for Next Generation of C++ Alisdair Meredith
No interest, or superseded by other proposals
These papers have either been superseded by different set of proposals, or were identified as a something we do not want to do. While everyone is free to submit papers they feel strongly about, it is unlikely any of these topics will gain favour for a future standard unless something significantly new is presented with the updated paper.
N1394 Some proposed extensions to C++ language David Miller
N1464 Anonymous array members Daniel Gutson
N1465 Constant inheritance Daniel Gutson
N1467 N1824 Extending Aggregate Initialization Alisdair Meredith, Daniel F. Gutson
N1584 N1701 Regularizing Initialization Syntax Francis Glassborow
N1468 Self methods Daniel Gutson
N1469 Inline Constants Daniel Gutson
N1492 Exclusive Inheritance Daniel Gutson
N1510 N1522 N1536 N1782 N1886 Concepts - syntax and composition B. Stroustrup, G. Dos Reis
N1758 N1798 N1848 N1849 Concepts for C++0x J. Siek, D. Gregor et al.
N1605 Extending Template Type Parameters I: Namespace and scope Daniel Gutson
N1612 How we might remove the remaining shortcomings of std::complex<T> Thorsten Ottosen
N1869 Wording for imaginary numbers Thorsten Ottosen
N1875 C++ Threads Lawrence Crowl
N1959 Member Initializers M Spertus
Background papers for reference
The papers do not contain actual proposals, but rather motivation and background to support other papers.
N1395 Aspects of Forwarding (was C++ Support For Delegation) Lois Goldthwaite
N1648 Motivation, Objectives and Design Decisions Attila Fehér
N1680 N1738 N1777 N1876 N1911 N1942 N2010 Memory model for multithreaded C++: Issues A. Alexandrescu, H. Boehm, K. Henney, B. Hutchings, D. Lea, B. Pugh
N1815 ISO C++ Strategic Plan for Multithreading Lawrence Crowl
N1834 A Pleading for Reasonable Parallel Processing Support in C++ Detlef Vollmann
N2012 Thread Subcommittee minutes from Berlin Lawrence Crowl
N2075 Prism: A Principle-Based Sequential Memory Model for Microsoft Native Code Platforms H Sutter
N2138 A Less Formal Explanation of the Proposed C++ Concurrency Memory Model Hans-J. Boehm
N1799 C++ Language Support for Generic Programming J. Siek, D. Gregor, R. Garcia, J. Willcock, J. Järvi, A. Lumsdaine
N1885 A formalism for C++ G. Dos Reis, B. Stroustrup
N1899 Concept proposal comparison Matt Austern
N1956 Design Rationale for C++/CLI H Sutter
N1976 Shared dynamic objects : review and issues Benjamin Kosnik
N2027 A Brief Introduction to Rvalue References Howard Hinnant, Bjarne Stroustrup, Brokek Kozicki
N2087 A brief introduction to variadic templates Doug Gregor