Somehow I always forget that the more global LC_ALL takes precedence over
authormartijn <martijn@openbsd.org>
Fri, 7 Jan 2022 10:20:11 +0000 (10:20 +0000)
committermartijn <martijn@openbsd.org>
Fri, 7 Jan 2022 10:20:11 +0000 (10:20 +0000)
commit46548d5bc3a72f77ece16854cb2754d8984814e7
treeeff18e9ba26e7f470000d398035238e99150ae85
parent05cb8c9e4f7a9b14b9aab2075268222030b870f8
Somehow I always forget that the more global LC_ALL takes precedence over
the more specific LC_CTYPE. Things are weird that way.

The problem here was that "eval" and "LC_ALL=" were swapped, not the
priority of variables.

pointed out by naddy@
OK tb@
regress/usr.sbin/snmpd/snmpd.sh