[rereading this thread]
... I'm sure it's possible to end up with successful utf-8
decodes of latin-1 text, ...
A conversion latin-1 <-> unicode is simply impossible,
unless you (re)create a transcoding table.
(trans-coding != en-coding, trans-coding != de-coding)
I think ascii ...
- The set of the ascii code points is a subset of the unicode
code points set,
and
- The set of the ascii characters is a subset of the set
of the unicode characters,
and
- The set of the ascii encoded code points is a subset of
the set of the unicode code points encoded in utf-8.
To understand the "coding of characters", the knowledge of
mathatematical concepts, sets, injection, surjection,
bijection, is enough.
···
-------------
The set of the characters proposed by the iso-8859-1 table
is simply unusable for many users, eg you can not write texts
in French.
-------------
jmf