EWGI will see this after LEWGI/LEWG has had something to say here.

Needs an sg16 tag too since the paper is talking about encodings.

Zoom call on January 21, 2025

No polls.

There is still support for a bag-of-bytes as was polled previously for P1275.

Next step: SG16 Text review and updates before additional SG18 LEWGI review.

SG16 reviewed a draft of what will be P3474R1 during the 2025-04-09 SG16 meeting. No polls were taken, but lots of good feedback was provided to the author. Concerns discussed include:

• Imposing transcoding requirements on new interfaces without first providing a transcoding
  facility that specifies how transcoding errors are to be handled. It should be possible to write
  wording for such functions that specifies transcoding using "as-if by ...". New proposal
  authors shouldn't have to determine and specify themselves how transcoding concerns are
  handled.
• The need for access to program arguments in raw form (char or wchar_t-based), in the
  execution encoding (char-based), in the wide execution encoding (wchar_t-based), in
  UTF-8 (char8_t-based), and in UTF-16 (char16_t-based as might be useful for ICU). If use
  cases for all of these exist, then UTF-32 (char32_t-based) is the only encoding left out for
  which there is a builtin character type.
• Which encodings should be supported via convenience functions.
• Distinction of the environment encoding (as identified by std::text_encoding::environment())
  and the execution encoding (locale dependent).
• Implementation concerns relating access to program arguments before main() is invoked;
  such access is not readily available for some operating systems.
• How to avoid imposing program startup costs.
• What effect mutation of program arguments provided to main() has on access to program
  arguments via std::arguments().

I'm retaining the SG16 label to review again once a new revision is published.