That was, if I recall correctly, about the C standard library interfaces in the Null-terminated multibyte strings section. Basically that the character at a time interfaces are not amenable to vectorization. 

On Sun, Sep 1, 2019, 11:55 Henri Sivonen <> wrote:
On Sat, Aug 31, 2019 at 10:16 PM Tom Honermann <> wrote:
> The summary for the SG16 meeting held August 21st, 2019 is now available. For those that attended, please review and suggest corrections:
> -

The minutes say:

> C interfaces can't achieve the performance that Bob Steagall demonstrated with his UTF-8 work.

What's that about? Does it mean that interfaces that assume
zero-terminated sentinel-based strings can't achieve the performance?
Surely a C interface that identifies both the source and the
destination by pointer and length can.

Henri Sivonen
SG16 Unicode mailing list