SG16 will hold a meeting on Wednesday, March 13th, at 19:30 UTC
(timezone conversion
<https://www.timeanddate.com/worldclock/converter.html?iso=20240313T193000&p1=1440&p2=tz_pdt&p3=tz_mdt&p4=tz_cdt&p5=tz_edt&p6=tz_cet>).

*For those in North America, please note that daylight savings time will
have begun, so this telecon will begin one hour later relative to the
last telecon.*

The agenda follows.

  * P3154R0: Deprecating signed character types in iostreams
    <https://wg21.link/p3154r0>.
  * P1729R4: Text Parsing <https://wg21.link/p1729r4>.

P3154R0 seeks to deprecate the basic_istream<char, ...> extraction and
basic_ostream<char, ...> insertion operators that treat signed char and
unsigned char as character types. This is motivated by common use of
these types as integer types (int8_t and uint8_t) and the desire to
remove surprising behavior that occurs when these types are used with
iostreams. Deprecation and eventual removal could clear the way for
eventually replacing the current operators with ones that treat these
types as integer types for all charT stream types. It is perhaps worth
noting that this is not the only place in the C++ standard that treats
these types as character types. Following adoption of P2513 (char8_t
Compatibility and Portability Fixes) <https://wg21.link/p2513> for
C++23, an array of unsigned char may be initialized with a UTF-8 string
literal. This leniency was added for compatibility with C where char8_t
is a type alias of unsigned char.

SG16 has reviewed prior versions of P1729R4 on several occasions:

  * 2023-11-11 SG16 review of P1729R3
    <https://github.com/sg16-unicode/sg16-meetings/blob/master/README-2023.md#october-11th-2023>.
  * 2023-09-27 SG16 review of P1729R2
    <https://github.com/sg16-unicode/sg16-meetings/blob/master/README-2023.md#september-27th-2023>.
  * 2019-06-22 SG16 review of P1729R0 in Cologne
    <https://wiki.edg.com/bin/view/Wg21cologne2019/SG16P1729R0>.

This latest revision incorporates changes requested by SG9 and elevates
ill-formed encoding input from garbage to errors per previous SG16
discussion (see section 4.11, Encoding
<https://wg21.link/p1729r4#encoding>). The last SG16 review included
discussion of text vs binary input. Elias can clarify, but I believe the
current proposal focuses on text support and would require custom
formatters to be written to handle binary input (which is fine). I think
other previously raised SG16 concerns have been addressed, so this
review can focus on any new concerns or we can poll to forward the paper.

Tom.