🚀 go-pugleaf

    I discussed this solution later in my article, under the heading of
    UTF-2:

     The problems with interchanging and processing 16-bit and 32-bit encoding
     forms may be alleviated through the use of a transformation format.  Two
     formats have been defined (one in 10646), UTF-1, and another by
     JTC1/SC22/WG15, UTF-2.  The latter is currently used by the AT&T Plan 9
     operating system ...

    Perhaps you didn't recognize the name UTF-2.  It is also called FSS-UTF,
    for File System Safe UTF, although this only applies to ASCII file systems,
    and not, for example, an EBCDIC based file system.

Before everybody starts to assume that plan 9 has an ASCII file
system, let me tell you that that isn't so. The file system looks
like a UNIX file system, it is just a stream of bytes (octets in
standards parlando).  What these octests encode is arbitrary.  It
might even be an EBCDIC file. However, if the contents is text
using plan 9's native character set, the encoding is UTF-2. The
ASCII compatibility feature makes it possible that if you send me
an ASCII file, (like mail), I don't have to translate it in UTF
first.

I guess what you want to say is that this doesn't work for EBCDIC,
but then, if I hand you a file in ASCII and you talk EBCDIC, you
have to do something to turn it into EBCDIC. That is because these
two encodings are not compatible.

So, if  you want to have the same relation (UTF-2 encoded file
compatible with ASCII encoded files) of the UNICODE encoding for
EBCDIC, define an encoding format that has this property.

	jaap