Andrew Welch wrote:
> One theoretical question that would help:  If I choose utf-8 as my
> output encoding, there will be no BOM and characters in the original
> ascii range will be output exactly as if I had chosen ascii...? (0-127) 

Yes.
 
> So in theory, any program that takes the output, copies it byte-for-byte
> and adds some its own bytes in ascii, would maintain the original utf-8
> encoding (therefore any program told to read it in utf-8 would be
> successful).  Is this sound?

It would work, yes, as long as you can be certain that it's always
done exactly as you say.

Is it wise? I'd be imagining the day a year from now when someone else
is having to wrench something into your application, and they'll not have
done their homework and will just think a character is a character, "let's
throw some random strings in there and hey why is my XML broken!?! XML is 
dumb!"


   - Mike
____________________________________________________________________________
  mike j. brown                   |  xml/xslt: http://skew.org/xml/
  denver/boulder, colorado, usa   |  resume: http://skew.org/~mike/resume/

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list