On Mon, Feb 16, 2026 at 08:07:38PM +0100, Rene Kita wrote:
I think somewhere in the discussion leading to this patch was a rational why wcwidth is special here, I already can't remember. Maybe it would be worth it to add that rational to the commit message.
That was from Steffen Nurpmeso in <20260212194458.BQUwHGoc@steffen%sdaoden.eu>, and in the sub-thread under <[email protected]>, Daniel Tameling and Steffen noted the backup implementation we currently have might continue to be a suitable replacement for systems missing it.
I'll update the commit to note the discussion on tentative 2.5.0 plans and give pointers to that discussion.
AFAIU all the wc-funcs are now part of ISO C or POSIX and systems not supporting it are legacy. Deprecating is fine, maybe don't put 2.5.0 as a hard deadline (just in case). If someone somewhere is really building a modern mutt on such a legacy system they can come to the ML and complain.
Agreed. That's why we're doing the two-step dance of printing a warning, and giving a whole release cycle to give people to complain.
I'm hoping we may be past that point. There are some people who still compile recent mutt on older systems, but maybe "older" is capable of the wc-funcs by now...
-- Kevin J. McCarthy GPG Fingerprint: 8975 A9B3 3AA3 7910 385C 5308 ADEF 7684 8031 6BDA
signature.asc
Description: PGP signature
