I had noticed that some Unicode characters were not displayed (showing the "rhombus with question mark" instead), precissely those that have a 0x8D byte when encoded in UTF-8. Examples: the cyrillic e (¨), 0xD1 0x8D; the "thumbs up" emoji (¨), 0xF0 0x9F 0x91 0x8D.

I've found that this issue can be fixed by adding "-keepsoftcr" to the corresponding line in jamnntpd.xlat, just like with CP866:

================================
read UTF-8   utf-8   -keepsoftcr
================================

Carlos

--- Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101 Thunderb