Re: character maps

The technicalities have been explained and answered, but I just
wonder why an UTF-8 encoding has to be converted to to NCR. It's
quite unlikely that any XML processor isn't capable of reading UTF-8
- day 1 of Unicode is long past.

I routinely do this if generating XML or (X)HTML files for others to
use. Browsers (and other XML clients) can of course understand UTF-8 but
(as specified) they believe the encoding information in HTTP headers
which means that anyone putting the files on to their local http server
has to configure the mime types correctly. It's just easier to make the
files work with any of ascii, latin1 or utf8 than explain to people what
they need to do to, especially as they may not have control over the
server (and not every server allows local customisation with .httaccess etc)

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.