<META> tags are HTML. They are a requirement because otherwise you
don't know what character set is being used. When you don't have a META
tags, the browser sometimes tries to guess your character set heuristically,
sometimes using character frequencies to identify the language. When it
guesses wrong, you can manually set the encoding you are browsing. But
still, a well-formed HTML page will have a META tag that says what character
set is being used. And then, no guessing, no need for manual specification
of the character set.
UTF-8 is a Unicode encoding that is backwards compatible with ASCII. That
is, if you have an english document in ASCII, it passes as a UTF-8 document
too. I have no idea what character entities are, but it appears that some
program on the way from your keyboard to the ibiblio internet line converts
your Hebrew words to HTML. But if that conversion were not to happen, we
could still send Hebrew to B-Hebrew, because we could just use a UTF-8
encoding.
Evidently, if your web client is sending: "יחד" but is
labeling it as "text/plain" something is wrong, since "&#xxxx;" codes are
HTML only. The fact that Kirk Lowery and others can send Hebrew on the
list and be read by others, and it gets saved in the archive properly, shows
that Hebrew is being handled by the list correctly. Some other programs
don't know how to interface properly with unicode. In fact, many don't!
Anyway, after doing all that, I think it's just simpler to use a
transliteration
scheme. I don't know how to read the vowels in the list transliteration
scheme
yet and many times I've used my own personal transliteration scheme. But if
I write ")otot umoftim" I usually get the message across without needing to
resort to miracles to explain to the poor guy out there why he's
seeing question
marks.