• From: "Michael Kay" <mike@s...>
• To: "'Richard Salz'" <rsalz@u...>
• Date: Mon, 11 Dec 2006 19:27:31 -0000

> Well that's not quite right; since he is defining a subset, 
> any XML-reading tools should work fine; there might be 
> problems with XML-writing tools, but a minor post-processing 
> edit seems likely to clean that up.

We're seeing a lot of traffic at the moment from people asking for tweaks to
serializers to make them avoid using constructs that recipients can't
handle. For example, there's a thread on the JDOM list from someone who
can't cope with CDATA sections (banned apparently because they are deemed a
security risk!), and I had one recently on the Saxon list from someone who
wanted particular line endings, and another not so long ago from someone who
didn't want empty element minimization. Minor post-processing edits seem to
be beyond the skill level of the people raising these posts; and they
shouldn't be necessary. People should accept XML. I intend to continue
responding to such requests with "tell your customer to install a real XML
parser".

Michael Kay
http://www.saxonica.com/

• Follow-Ups:
  • RE: How to handle "newline" characters in an XML parser.
    • From: Richard Salz <rsalz@u...>

• References:
  • RE: How to handle "newline" characters in an XML parser.
    • From: Richard Salz <rsalz@u...>