Hello mark,

On Tuesday June 09 2020 08:17, you wrote to me:

 MvdV>> However, as I wrote before, automatically detecting the
 MvdV>> encoding is tricky. So if more than CP437 is needed, the user
 MvdV>> needs to interfere to tell the software what the encoding is.

 ml> so umm... here's an idea... scan the message when/while it is
 ml> written... the software has to do this anyway, right? so if all the
 ml> characters are <128, then use CHRS: ASCII otherwise, convert to and
 ml> use CHRS: UTF-8... poof! no more crazy unsupported unmaintained
 ml> illogical competing controversial character set muckity muck...

Spledid idea. Why do you think I have been advocating UTF-8 in Fidonetr for
over a decade?

Just this:

1) To automatically convert from encoding A to UTF-8 one has to know what
encoding A is. This may be a problem in some setups.

2) ASCII is a subset of UTF-8. So labeling everything as UTF-8 is no technical
problem. But then if all messages are endoded in UTF-8, we can do away with the
CHRS kludge all together.

3) Fidonet is not ready yet for a complete move to UTF-8, There is still too
much software around that can not handle it. Especially the Russians will not
go along.

4) So it is a good idea, in maybe 5 or 10 years...


Cheers, Michiel

--- GoldED+/W32-MSVC 1.1.5-b20170303