On 10/17/20 11:32 AM, Eli Zaretskii wrote:
> this is still nowhere near the compromise I suggested.
> And it loses %S, something I didn't agree to.

Attached is a revised patch that implements what you suggested, and it does not 
lose %S.

Here is what you suggested in 
<https://debbugs.gnu.org/cgi/bugreport.cgi?bug=43439#14>:

> How about a compromise: we modify doprint to exit when either it finds
> NUL or reaches the character specified by FORMAT_END?

and the attached patch implements this proposed API change.