I've experienced the problems with Latin1 characters when using
nsd8x, and read about the
current
solution.
Problem is that's temporary. We should be able to use tcl 8, and
maybe will have to in the future. Isn't there a way to
automatically translate string to/from unicode and some defined
charset? That other charset could go as global parameter, and would
be used when writing to client browsers and the database.
