• From: Stephen Green <stephengreenubl@g...>
• To: Andrew Welch <andrew.j.welch@g...>
• Date: Tue, 27 Jul 2010 09:56:57 +0100

Maybe it's worth remembering too that some environments like mainframes which
can handle XML as text might have memory problems when trying to process it via
DOM, etc. It would be useful to have an identified subset of XML which
can be handled on
such platforms without the need to resort to making special
adjustments to memory
configurations. Of course in the case of the beloved mainframe it
might be a challenge
to meet a UTF-8-Only requirement, but even there the mapping table techies can
get the benefit from the subset in that they only have to map to
EBCDIC the valid UTF-8
values, even though tools have to be used to convert the EBCDIC back
and forth to
UTF-8. It helps to know that there won't be any problem characters
like '£' in the XML
to cause errors when converted to and from EBCDIC if such characters
are escaped.

Best regards

---
Stephen D Green

• References:
  • hackable xml
    • From: Andrew Welch <andrew.j.welch@g...>