By InbalL | Mar 10, 2020 12:37 PM | Tags: None
A trip report from the release of C++20, in Prague WG21 meeting.
Trip report: ISO C++ standards in Prague
by Inbal Levi
About the report
This trip report is not exhaustive but limited to my experience, and is just a taste from some of the topics I find most interesting.
By Blog Staff | Feb 21, 2020 11:26 AM | Tags: None
The following are notes from Bjarne Stroustrup, regarding the completion of C++20 on Saturday and summarizing his comments to the committee in the room:
After the 79-0 vote deeming C++20 done on February 15, 2020 in Prague, I made a few comments to emphasize the significance of the event. I got to the lectern just after Herb Sutter had shown photos of the C++ standards committee members at the first meeting and the meetings that voted for the C++11, C++14, C++17, and C++20 standards. I was in all of those, so was Mike Miller from EDG and chair of the Core Working Group. We were the only ones in all photos. This is roughly what I said:
This is a historical event:
- 30 years of C++ standardization.
- 40 years of C++.
- C++20 is the 6th standard, the 3rd major standard; by “major” I mean “changes the way people think.”
- This is something like the 75th meeting; I have been at about 70 of those.
I’d like to add a personal note. For me, C++20 is special because it has essentially all from “The Design and Evolution” (1994). In particular, it has concepts, modules, and coroutines:
- Concepts were not in D&E, but there were three pages of apologies for not having them. Then, neither I nor anyone else know how to design and implement them sufficiently well.
- Modules were just a dream then, but they were specifically mentioned – I dreamed of a day where we could finally eliminate the preprocessor.
- Coroutines were the bread and butter for C++ during the first 10 years; they were missing in most other languages. I was sore having lost them to implementation problems on SPARC architecture.
In addition, we get improved concurrency and a library with ranges, dates, and span.
We (the C++ standards committee members) must be careful and responsible; we serve a huge community:
- Serve the community at large, rather than just experts – “keep simple things simple.”
- Be careful, the world changes; what seems essential or fashionable today may not be good in the longer term.
- Be pragmatic, not doctrinaire; pragmatic and principled.
This has guided the development of C++ so far. No, that’s not easy. We must balance many concerns. I suggest we:
- Pursue the goal of a completely resource-safe and type-safe C++.
- Support a wide variety of hardware well.
- Maintain C++’s record of stability (compatibility) – “stability is a feature” – as much as makes sense.
These aims have served C++ well throughout.
I wonder who’ll be here in 30 years? That’ll be 2050.
Thanks for listening. See you in Varna! (Varna is the location of the next ISO C++ Standards meeting).
By Adrien Hamelin | Feb 20, 2020 01:59 PM | Tags: community
All you want to know.
Freestanding in Prague
by Ben Craig
From the article:
The C++ standards committee met in Prague, Czech Republic between Feb 10 and Feb 15. The standard is wording complete, and the only thing between here and getting it published is ISO process. As is typical for me at these meetings, I spent a lot of time doing freestanding things, Library Incubator (LEWGI) things, and minuting along the way (15-ish sessions/papers!)...
By Adrien Hamelin | Feb 18, 2020 01:07 PM | Tags: community
Houra!
C++20 is Done!
From the article:
C++20, the most impactful revision of C++ in a decade, is done! ������
By Blog Staff | Feb 18, 2020 09:46 AM | Tags: None
Here's a 9-minute taste of what it was like to be at the Prague meeting, with attendees talking about their favorite highlights of C++20...
By Blog Staff | Feb 16, 2020 12:39 AM | Tags: None
A landmark release of C++:
Winter ISO C++ standards meeting (Prague)
by Herb Sutter
From the article:
C++20 is done! ...
By Adrien Hamelin | Feb 12, 2020 02:57 PM | Tags: community
Did you know them?
Five Awesome C++ Papers for the Prague ISO Meeting and C++20 Status
by Bartlomiej Filipek
From the article:
Continuing the tradition for other ISO C++ Meetings, I prepared a blog post where you’ll learn about:
- The current status of C++20
- an overview about the Prague ISO C++ Meeting (10th till 15th February 2020)
- a few interesting papers that are worth reading
Let’s start!
By Corentin Jabot | Feb 7, 2020 11:45 AM | Tags: c++23
SG-11, the study group charged of all things concurrency and parallelism made forward progress and sent the proposal to LEWG - with the hope of landing a future revision in the C++23 draft. This is rather big news given that this work has been brewing for about a decade.
A Universal I/O Abstraction for C++
by Corentin Jabot
From the article:
The year is 2020 and even consummer CPUs feature double digits number of cores, storage offers 10GB/s read speeds and networks have to accommodate ever-growing traffic.
For a long time, the C++ committee seemed to think that either async file I/O didn’t make sense or was fundamentally irreconcilable with networking. This belief would lead to two inter-incompatible APIs in the standard, which would be a nightmare in term of usability (aka ASIO and AFIO).
It seems that there is finally a way to resolve these divides:
io_uring offers very high performance I/O which doesn’t discriminate on device type.
Sender Receiver provides the composable, low-cost, non-allocating abstraction while offering a simple mental model for asynchronous operations lifetime.
Coroutines make asynchronous i/o dead simple for the 99% use case.
Asynchronous Networking is nice.Asynchronous I/O is better.
AWAIT ALL THE THINGS!
By Adrien Hamelin | Jan 20, 2020 11:29 AM | Tags: community
Did you know about the proposal?
Why were abbrev. lambdas rejected?
by Barry Revzin
From the article:
In November, 2017, I presented my proposal for abbreviated lambdas (P0573R2) to Evolution in Albuquerque. It was rejected (6-17), though the group was willing to consider future proposals for shorter lambdas which present new technical information (18-2).
Since then, there’s been a lot of confusion about what actually happened and why it was rejected, and part of this was my fault for not doing a good job communicating this to interested parties. Actually, that’s generous - I didn’t really communicate anything. So here is my making up for lost time by actually conveying this information...
By Blog Staff | Jan 18, 2020 03:51 PM | Tags: None
The 2020-01 mailing of new standards papers is now available.
| WG21 Number | Title | Author | Document Date | Mailing Date | Previous Version | Subgroup | Disposition |
|---|---|---|---|---|---|---|---|
| P1970R1 | Consistency for size() functions | Hannes Hauswedell | 2020-01-13 | 2020-01 | P1970R0 | Library | |
| P1976R1 | Fixed-size 'span' construction from dynamic-size range | Tomasz Kamiński | 2020-01-03 | 2020-01 | P1976R0 | Library | |
| P1985R0 | Universal template parameters | Gašper Ažman, Mateusz Pusz | 2020-01-13 | 2020-01 | EWGI | ||
| P1988R0 | Allow Templates in Local Classes | Steve Downey | 2020-01-13 | 2020-01 | EWGI, Evolution | ||
| P1990R0 | Add operator[] to std::initializer_list | Daniil Goncharov, Antony Polukhin | 2019-12-16 | 2020-01 | LEWGI, Library Evolution, Library | ||
| P1993R1 | Restore shared state to bulk_execute | Jared Hoberock | 2020-01-12 | 2020-01 | P1993R0 | SG1, Library Evolution | |
| P1997R1 | Relaxing Restrictions on Arrays | Krystian Stasiowski, Theodoric Stier | 2020-01-13 | 2020-01 | P1997R0 | Evolution | |
| P1998R1 | Simple Facility for Lossless Integer Conversion | Ryan McDougall | 2020-01-08 | 2020-01 | P1998R0 | SG6, LEWGI | |
| P2000R0 | Direction for ISO C++ | Michael Wong | 2020-01-13 | 2020-01 | WG21 | ||
| P2002R0 | Defaulted comparison specification cleanups | Richard Smith | 2019-12-17 | 2020-01 | Core | ||
| P2003R0 | Fixing Internal and External Linkage Entities in Header Units | Michael Spencer | 2020-01-13 | 2020-01 | SG2, Evolution | ||
| P2004R0 | Numbers and their Scopes | Antony Polukhin | 2019-12-17 | 2020-01 | SG6, LEWGI | ||
| P2005R0 | A Brief 2D Graphics Review | James Berrow | 2019-12-21 | 2020-01 | SG13 | ||
| P2006R0 | Eliminating heap-allocations in sender/receiver with connect()/start() as basis operations | Lewis Baker, Eric Niebler, Kirk Shoop, Lee Howes | 2020-01-13 | 2020-01 | SG1, Library Evolution | ||
| P2007R0 | `std::from_chars` should work with `std::string_view` | Mateusz Pusz | 2020-01-10 | 2020-01 | LEWGI, Library Evolution | ||
| P2008R0 | Enable variable template template parameters | Mateusz Pusz | 2020-01-10 | 2020-01 | EWGI | ||
| P2009R0 | SG16: Unicode meeting summaries 2019-10-09 through 2019-12-11 | Tom Honermann | 2019-12-28 | 2020-01 | SG16 | ||
| P2010R0 | Remove iostream operators from P1889 | Daniil Goncharov | 2019-12-27 | 2020-01 | SG6, LEWGI | ||
| P2011R0 | A pipeline-rewrite operator | Barry Revzin, Colby Pike | 2020-01-07 | 2020-01 | Evolution | ||
| P2013R0 | Freestanding Language: Optional ::operator new | Ben Craig | 2020-01-10 | 2020-01 | SG14, Evolution | ||
| P2014R0 | Proposed resolution for US061/US062 - aligned allocation of coroutine frames | Lewis Baker, Gor Nishanov | 2020-01-13 | 2020-01 | Evolution | ||
| P2016R0 | A step parameter for iota | Corentin Jabot | 2020-01-14 | 2020-01 | Library Evolution | ||
| P2017R0 | Conditionally safe ranges | Barry Revzin | 2020-01-07 | 2020-01 | Library | ||
| P2020R0 | Locales, Encodings and Unicode | Corentin Jabot | 2020-01-13 | 2020-01 | SG16 | ||
| P2021R0 | Negative zero strikes again | Victor Zverovich | 2020-01-09 | 2020-01 | Library Evolution | ||
| P2024R0 | Bloomberg Analysis of Unified Executors | David Sankel, Frank Birbacher, Marina Efimova, Dietmar Kuhl, Vern Riedlin | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2025R0 | Guaranteed copy elision for named return objects | Anton Zhilin | 2020-01-08 | 2020-01 | EWGI | ||
| P2026R0 | A Constituent Study Group for Safety-Critical Applications | Ryan McDougall, Bryce Adelstein Lelbach, JF Bastien, Andreas Weis, Ruslan Arutyunyan, Ilya Burylov | 2020-01-14 | 2020-01 | WG21, Evolution, Library Evolution | ||
| P2027R0 | Moved-from objects need not be valid | Geoff Romer | 2020-01-10 | 2020-01 | Library Evolution, Library | ||
| P2028R0 | What is ABI, and What Should WG21 Do About It? | Titus Winters | 2020-01-10 | 2020-01 | Evolution, Library Evolution | ||
| P2029R0 | Proposed resolution for core issues 411, 1656, and 2333; escapes in character and string literals | Tom Honermann | 2020-01-12 | 2020-01 | Core | ||
| P2030R0 | SG19: Machine Learning 2019/10/10-2020/01/09 | Michael Wong | 2020-01-10 | 2020-01 | SG19 | ||
| P2031R0 | SG14: Meeting Minutes 2019/10/08-2020/01/07 | Michael Wong | 2020-01-10 | 2020-01 | SG14 | ||
| P2033R0 | History of Executor Properties | Jared Hoberock | 2020-01-10 | 2020-01 | SG1 | ||
| P2034R0 | Partially Mutable Lambda Captures | Ryan McDougall | 2020-01-11 | 2020-01 | EWGI | ||
| P2035R0 | Value Proposition: Allocator-Aware (AA) Software | Pablo Halpern, John Lakos | 2020-01-13 | 2020-01 | LEWGI, Library Evolution | ||
| P2036R0 | Changing scope for lambda trailing-return-type | Barry Revzin | 2020-01-12 | 2020-01 | Evolution | ||
| P2037R0 | String's gratuitous assignment | Andrzej Krzemieński | 2020-01-11 | 2020-01 | Library Evolution | ||
| P2038R0 | Proposed nomenclature for contract-related proposals | Andrzej Krzemieński, Ryan McDougall | 2020-01-11 | 2020-01 | SG21 | ||
| P2039R0 | do_until Loop | Menashe Rosemberg | 2020-01-01 | 2020-01 | LEWGI | ||
| P2040R0 | Reflection-based lazy-evaluation | Corentin Jabot | 2020-01-13 | 2020-01 | SG7 | ||
| P2041R0 | Deleting variable templates | David Stone | 2020-01-11 | 2020-01 | Evolution | ||
| P2042R0 | Alternate names for make_shared_default_init | Andrzej Krzemieński | 2020-01-11 | 2020-01 | Library Evolution | ||
| P2043R0 | Don't constexpr All The Things | David Sankel | 2020-01-13 | 2020-01 | SG7 | ||
| P2044R0 | Member Templates for Local Classes | Robert Leahy | 2020-01-12 | 2020-01 | EWGI, Evolution | ||
| P2045R0 | Missing Mandates for the standard library | Marshall Clow | 2020-01-13 | 2020-01 | Library | ||
| P2046R0 | Rangify New Algorithms | Dan Raviv | 2020-01-12 | 2020-01 | Library | ||
| P2047R0 | An allocator-aware optional type | Nina Ranns, Pablo Halpern Ville Voutilainen | 2020-01-13 | 2020-01 | LEWGI, Library Evolution | ||
| P2048R0 | Prohibit zero and NULL from being used as null pointer literals | Vittorio Romeo | 2020-01-12 | 2020-01 | EWGI | ||
| P2049R0 | Constraint refinement for special-cased functions | Andrew Sutton, Wyatt Childers | 2020-01-13 | 2020-01 | SG7 | ||
| P2050R0 | Tweaks to the design of source code fragments | Andrew Sutton, Wyatt Childers | 2020-01-13 | 2020-01 | SG7 | ||
| P2051R0 | C++ Library Issues to be moved in Prague | Marshall Clow | 2020-01-13 | 2020-01 | WG21 | ||
| P2052R0 | Making modern C++ i/o a consistent API experience from bottom to top | Niall Douglas | 2020-01-12 | 2020-01 | SG1, LEWGI, Library Evolution | ||
| P2054R0 | Audio I/O Software Use Cases | Sophia Poirier, Frank Birbacher, Timur Doumler | 2020-01-13 | 2020-01 | SG13 | ||
| P2055R0 | A Relaxed Guide to memory_order_relaxed | Paul E. McKenney, Hans Boehm | 2020-01-12 | 2020-01 | SG1 | ||
| P2057R0 | SG14 SG19 Past, Present and Future status | Michael Wong, Ben Craig, Paul Bendixen, Matthew Bentley, Inbal Levi, Rene Riviera, Steffan Tjernstrom, Mark Hoemmen, Ronen Friedman | 2020-01-13 | 2020-01 | SG14, SG19 | ||
| P2058R0 | Make std::random_device Less Inscrutable | Martin Hořeňovský | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2059R0 | Make Pseudo-random Numbers Portable | Martin Hořeňovský | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2060R0 | Make Random Number Engines Seedable | Martin Hořeňovský | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2061R0 | Sequential consistency for atomic memcpy | David Goldblatt | 2020-01-13 | 2020-01 | SG1 | ||
| P2062R0 | The Circle Meta-model | Daveed Vandevoorde, Wyatt Childers, Andrew Sutton, Faisal Vali | 2020-01-13 | 2020-01 | SG7 | ||
| P2064R0 | Assumptions | Herb Sutter | 2020-01-13 | 2020-01 | SG21, Evolution | ||
| P2065R0 | naming and aliases | Kirk Shoop | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2066R0 | Suggested draft TS for C++ Extensions for Transaction Memory Light | Jens Maurer, Michael L. Scott | 2020-01-13 | 2020-01 | SG1, Evolution | ||
| P2067R0 | Allowing trailing commas in ctor-initializer | Marc Mutz | 2020-01-13 | 2020-01 | EWGI | ||
| P2068R0 | Using ?: to reduce the scope of constexpr-if | Marc Mutz | 2020-01-13 | 2020-01 | EWGI | ||
| P2069R0 | Stackable, thread local, signal guards | Niall Douglas | 2020-01-13 | 2020-01 | SG1, SG12, EWGI, LEWGI | ||
| P2070R0 | A case for optional and object_ptr | Peter Sommerlad, Anthony Williams, Michael Wong, Jan Babst | 2020-01-13 | 2020-01 | SG12, LEWGI, Library Evolution, Library | ||
| P2071R0 | Named universal character escapes | Tom Honermann, Peter Bindels | 2020-01-13 | 2020-01 | SG16, Evolution | ||
| P2072R0 | Differentiable programming for C++ | Marco Foco, Vassil Vassilev, Max Rietmann, Michael Wong | 2020-01-13 | 2020-01 | SG19 | ||
| P2073R0 | Debugging C++ coroutines | Dmitry Duka, Ivan Shutov, Konstantin Sadov | 2020-01-13 | 2020-01 | SG15 | ||
| P2074R0 | Asynchronous callstacks & coroutines | Dmitry Duka | 2020-01-13 | 2020-01 | SG15 | ||
| P2075R0 | Philox as an extension of the C++ RNG engines | Pavel Dyakov, Ilya Burylov; Ruslan Arutyunyan; Andrey Nikolaev | 2020-01-13 | 2020-01 | SG6 | ||
| P2076R0 | Previous disagreements on Contracts | Ville Voutilainen | 2020-01-13 | 2020-01 | SG21 | ||
| P2077R0 | Heterogeneous erasure overloads for associative containers | Konstantin Boyarinov, Sergey Vinogradov; Ruslan Arutyunyan | 2020-01-13 | 2020-01 | LEWGI | ||
| P2078R0 | Add new traits type std::is_complex | Bob Steagall | 2020-01-13 | 2020-01 | SG6, Library Evolution | ||
| P2079R0 | Shared execution engine for executors | Ruslan Arutyunyan, Michael Voss | 2020-01-13 | 2020-01 | SG1 | ||
| P2080R0 | Polymorphic allocators: There is no such thing as One True Vocabulary Type | Ville Voutilainen | 2020-01-13 | 2020-01 | Library Evolution | ||
| P2081R0 | Rebase the Library Fundamentals v3 TS on C++20 | Thomas Köppe | 2020-01-13 | 2020-01 | Library Evolution, Library | ||
| P2082R0 | Fixing CTAD for aggregates | Timur Doumler | 2020-01-13 | 2020-01 | Core | ||
| P2087R0 | Reflection Naming: fix reflexpr | Mihail Naydenov | 2020-01-12 | 2020-01 | SG7 | ||
| P2088R0 | Reflection Naming: Reification | Mihail Naydenov | 2020-01-12 | 2020-01 | SG7 |