SG16 will hold a meeting *tomorrow*, Wednesday, April 23rd, at 19:30 UTC
(timezone conversion
<https://www.timeanddate.com/worldclock/converter.html?iso=20250409T193000&p1=1440&p2=tz_pdt&p3=tz_mdt&p4=tz_cdt&p5=tz_edt&p6=tz_cest>).

If you need a .ics file to import into your calendar, you can download
it here
<https://documents.isocpp.org/remote.php/dav/public-calendars/R7imgS2LJD9xfeWN/5FDDA8FA-1102-44AC-8E03-DC78F1B95BB6.ics?export>.

The agenda follows.

  * P3655R0: std::zstring_view <https://wg21.link/p3655r0>.
  * P2873R3: Remove Deprecated locale category facets for Unicode from
    C++26 <https://wg21.link/p2873r3>.

*P3655R0* seeks to standardize a variant of std::string_view that offers
a null termination guarantee. This is not the first paper to propose
such a type; nor the second, nor the third! Given the history of papers
(see section 3 in the paper) that stalled for one reason or another, the
motivation for SG16 review is to provide a (hopefully strong)
recommendation to LEWG (one way or the other) regarding the utility of
the proposed type from a group of text experts. This proposal was on the
agenda for the last SG16 meeting, but was not discussed. There was some
mailing list discussion <https://lists.isocpp.org/sg16/2025/04/4545.php>
that I urge participants to read before the meeting tomorrow. That
discussion concerned:

  * Support for std::char_traits.
  * Support for conversion to std::string_view and implications for
    invariants, creation of temporaries, and life-time concerns.

*P2873R3*, well, a draft of it, was already forwarded by SG16 during the
2025-03-12 SG16 meeting
<https://github.com/sg16-unicode/sg16-meetings#march-12th-2025>. The
only reason that it is back on the agenda again is to provide a
recommendation whether all of the deprecated std::codecvt facets should
be removed in C++26 or whether the ones that were deprecated during
C++26 should be retained for removal in a future standard. There was a
small amount of mailing list discussion
<https://lists.isocpp.org/sg16/2025/03/4541.php> regarding this last
month. The status quo is:

  * Introduced in C++11 (via N2238 <https://wg21.link/n2238>),
    deprecated in C++20 (via P0482R6 <https://wg21.link/p0482r6>), on
    track to be removed in C++26 (via P2873R3 <https://wg21.link/p2873r3>):
      o codecvt<char16_t, char, mbstate_t>
      o codecvt<char32_t, char, mbstate_t>
      o codecvt_byname<char16_t, char, mbstate_t>
      o codecvt_byname<char32_t, char, mbstate_t>
  * Introduced in C++20 (via P0482R6 <https://wg21.link/p0482r6>),
    deprecated in C++26 (via LWG 3767 <https://wg21.link/lwg3767>),
    should be removed in C++26?
      o codecvt<char16_t, char8_t, mbstate_t>
      o codecvt<char32_t, char8_t, mbstate_t>
      o codecvt_byname<char16_t, char8_t, mbstate_t>
      o codecvt_byname<char32_t, char8_t, mbstate_t>

Note that the latter set that convert to/from char8_t should not have
been added to the standard in the first place. UTF conversions are never
locale dependent, nor is support for file streams of char8_t (or
char16_t or char32_t) required by the standard. Nothing in the C++
standard library depends on these facets. Removal of these facets does
remove the only UTF transcoding facility currently specified by the C++
standard, but such conversions are still possible using facilities from
C (e.g., c8rtomb() followed by mbrtoc16(), at least for characters in
the BMP).

Tom.