If you're writing, say, a message editor and you don't want to deal with
UTF-8 input at all, specify this mode flag to convert any translatable UTF-8
chars (that have CP437 equivalents) to CP437 chars. Non-translatable UTF-8
chars are just ignored (treated the same as a timeout/no-input).

Does this resolve Coverity warning? I don't rember.

No functional change.

for use when all you need is the length (and validity) of the first byte in
a UTF-8 sequence.

DDMsgReader: UTF8 update for the standard Synchronet message header too.  Also, oops, dd_msg_reader.js should only be in exec/load, not xtrn/DDMsgReader.

See merge request !396

It seems each concurrent task gets a separate workspace.

DDMsgReader: UTF8 update for the standard Synchronet message header too.  Also, oops, dd_msg_reader.js should only be in exec/load, not xtrn/DDMsgReader.

For some reason, weird things happen on netbsd/openbsd

ciolib and smblib don't need hash and encode to be built.

It didn't work anyway.

For the rules still in the main file

This also splits up the dependencies better