The cause of this error is that "man" converts the input to "UTF-8" although the output expects "latin1" (iso-8859-1). So the character \[la] gets the code u27E8, which does not exist in the used font.
INPUT FILE: .TH angle-test 1 2017-07-06 .pl 5 .fchar \[la] < Angle test with character \e[la] which is \[la]. COMMAND: (File "~/etc/manpath" is empty) man -D -C ~/etc/manpath -d -l <INPUT FILE> 2> <error_file> Part of the <error_file> is: final search path = /home/bg/.local/share/man:/usr/local/man:/usr/share/man ++priv_drop_count = 2 restore_cwd: 3 (null) pre-processors `t' from default page_encoding = iso88591 source_encoding = iso88591 locale_charset = ISO-8859-1 roff_device (locale) = utf8 roff_encoding = ISO-8859-1 output_encoding = UTF-8 Terminal width 80 Terminal width 80 within cat page range [80, 80] formatted_encoding = UTF-8 format: 1, save_cat: 0, found: 1 add_output_iconv: source UTF-8, target ISO-8859-1 Setting LESS to -ix8RmPm Manual page angle\.test ?ltline %lt?L/%L.:byte %bB?s/%s..?e (END):?pB %pB\%.. (press h for help or q to quit)$PM Manual page angle\.test ?ltline %lt?L/%L.:byte %bB?s/%s..?e (END):?pB %pB\%.. (press h for help or q to quit)$-I -z-2 Setting MAN_PN to angle\.test ++priv_drop_count = 3 opening - opened - trying encoding UTF-8 -> UTF-8//IGNORE --priv_drop_count = 2 --priv_drop_count = 1 The whole <error_file> is in the attachment -- Bjarni I. Gislason
geyma.gz
Description: GNU Zip compressed data