🚀 go-pugleaf

I think I understand what the problems are,
but I haven't seen a summary of what the solutions are.

I see two completly different problems, in points 2 and 3.

1. The low level transports.
UUCP, TCP, and ISO protocols are 8 bit. No problem here.

2. The high level transports.
SMTP is 7 bit, but there is a replacement.
But, even with a replacement, we have to deal with old software.

So the problem is encoding 8 bit data to deal with 7 bit gateways.
We do not have to deal the particular character sets
or with multi-byte encodings at this level.

How does MIME or other solutions deal with this?

3. The applications.
Applications have to be aware of the character set and encoding.
The simplest solution is to handle just the ISO 8859-1 character set.
This should be no problem with most modern software and hardware.
This can be called the "western european solution".

The second best solution is to handle all of the ISO 8859-# character sets.
The problem here is displaying the right character set,
and having all the fonts in hardware or software.
Not having all the fonts might be a problem for many people.
This can be called the "indo-european solution".

Any programmer can pick up some good books and implement most of this.
The two things that are poorly documented (or non-portable)
are locale naming conventions and selecting display fonts.

The best solution is to have full support for multi-byte character sets.
I think this is much more difficult than the previous solution,
because of the lack of library support for developing such applications.

How does MIME or other solutions deal with these problems?

		Eric Gisin, Mortice Kern Systems.