See files .../tech/blog.txt and .../tech/vim.info.

Determine encoding:
file -i x  # I find sometimes incorrectly reports ISO-8859's as  utf-8s.
"enca" or "chardet" (Python) may do a better job.

Java properties and source files use ISO-8859-1, but can use Unicode escapes
 for characters not covered by ISO-8859-1.
(Why doesn't "Java" use UTF-8, which I read is read Unicode?)

Unicode escapes \uxxxx w/hex.  Like \u2297 is an X in a circle.

UTF-8 = unicode use 2 bytes for extended characters.
vim latin1 = ISO-8859-1 uses 1 byte for each character.

At least with Konsole default encoding set to UTF_8 Shell needs LANG set like:
LANG=en_US.UTF-8
LANG=en_US.ISO-8859-1
LANG=de_DE.UTF-8

When reading text from Java, set CharSet to match encoding of the text file:
  UTF-8 or ISO-8859-1

vi uses lower-case
#:set encoding=latin1
:set encoding=utf-8     # Unless run into some problem, always use utf-8 here
:set fileencoding=latin1
:set fileencoding=utf-8

See http://www.apache.org/licenses/LICENSE-2.0.html

Available system encodings shown by "locale -a".

For foreign man pages, the instructions for Suse just don't work.
    Add the foreign man page dir to the head of the MANPATH.
    Set terminal emulator encoding to iso8859-1.
    Set LANG (or possibly LC_CTYPE, LC_MESSAGES, LC_LANG...) to desired
     encoding.

ENDCODING conversion:
    iconv -f iso-8859-1 -t utf-8 old.latin1 > new.utf8
The Suse release notes imply that ISO8859-1 is legacy.  I guess UTF8 is
what you should use if you need internationality.

locale -a   lists the locales available on UNIX