don at lexmark.com wrote:
> Jim:
>> I would disagree. I don't believe that all XHTML-Print enabled
> printers will necessarily bite the bullet and include a complete XML
> parser that requires support for UTF-16. I don't believe we should
> force that to occur. Perhaps you should remind the group that
> XHTML-Print is target for LOW-END printers with this embedded. No 3
> gigahertz Pentium 4's with 512 MB of memory!!!
I'm not so worried about memory usage; converting UTF-16 to UTF-8
on the input side is not expensive in terms of memory or processor.
However, reliably detecting UTF-16 and managing the endianess of
the words is a pain in the ass in the real world. Assuming that
all UTF-16 files start with FFFE or FEFF, the XML parser can handle
the UTF-16 encoding without difficulty, however certain large
convicted software monopolies regularly omit this important
information making autodetection unreliable.
Given the limited scope of XHTML-Print and the desire for maximum
interoperability, I would recommend that we stick with UTF-8 as the
only requirement so that applications that send XHTML-Print data have
to use UTF-8 and manage whatever perversion of UTF-16 they use
internally themselves...
--
______________________________________________________________________
Michael Sweet, Easy Software Products mike at easysw dot com
Printing Software for UNIX http://www.easysw.com