WG21 PAPERS

Mailings Beta

Volunteer-driven · Not affiliated with WG21 or ISO

Filter and export papers from the C++ Committee’s monthly mailings by time frame and audience using the options below, or refine your export further with the Select button. Links point directly to the official WG21 papers on open-std.org.

Submit Issue C++ Standards Committee Papers

Disposition Updated

Audience
Wording
Evolution
Mailings
1 · Year 1 · Month

Select the arrow to the right of the year to filter for specific months.

2026 Jul · 1 of 5

Pick months in 2026

Filters Author: Barry Revzin EWG Clear all
3 out of 3 papers selected for export
Paper Title Authors Audience Doc Date Audit
P2287R6 Designated-initializers for base classes B. Revzin
EWG
Mailing Date 2026-07 Post-Brno
Author(s) Barry Revzin
Target Audience EWG
Tracking issue View on GitHub
Disposition Date 2026-06
Abstract [[P0017R1]](https://wg21.link/p0017r1) extended aggregates to allow an aggregate to have a base class. [[P0329R4]](https://wg21.link/p0329r4) gave us designated initializers, which allow for much more expressive and functional initialization of aggregates. However, the two do not mix: a designated initializer can currently only refer to a direct non-static data members. This means that if I have a
P2806R4 do expressions B. Revzin, B. Lopez, Z. Laine, M. Park
EWG
Mailing Date 2026-07 Post-Brno
Author(s) Barry Revzin, Bruno Cardoso Lopez, Zach Laine, Michael Park
Target Audience EWG
Tracking issue View on GitHub
Abstract C++ is a language built on statements. `if` is not an expression, loops aren’t expressions, statements aren’t expressions (except maybe in the specific case of `*expression*;`).
P4101R1 Consteval-only Values for C++26 B. Revzin, P. Dimov, D. Vandevoorde, D. Katz
EWGCWG
Mailing Date 2026-07 Post-Brno
Author(s) Barry Revzin, Peter Dimov, Daveed Vandevoorde, Dan Katz
Target Audience EWGCWG
Revisions R0
Tracking issue View on GitHub
Disposition Date 2026-06
Abstract The Reflection design from [[P2996R13]](https://wg21.link/p2996r13) was based on a model of having consteval-only types to prevent reflections from leaking to runtime. But we’ve run into issues and limitations with that approach, so we propose that, for C++26, we change instead to a consteval-only value model. This solves the same problems, but has additional benefits.
Showing 3 of 3 papers · source: open-std.org