MartinJ,
THanks - I have never used those DOM properties - thanks for sharing them and their results. They will prove helpful to some of projects!:-)

So it seems that, unless the root tag (xml) has a question mark at both ends, is not valid. That explains why I spent a few hours of unproductive time earlier this week trying to use Xpath to access nodes and attributes.
Thanks

>Hi Terry,
>
>

>oXML=newObject("Msxml2.DOMDocument")
>oXML.async=.f.
>cXML=[<?xml version="1.0"><row>This is my row without a question mark</row>]
>?oXML.loadXML(cXML)
>?oxml.parseError.errorCode
>?oxml.parseError.reason
>

>
>

><?xml ... ?> is control/operative tag
>
><?xml ... > is unknown tag - not valid
>

>
>MartinJ
>
>>I do not know that (?xml version="1.0") is not valid - but I do know that XMLMSDOM2.Document seems to only "see" the nodes (tags) when the form is (?xml version="1.0"?)
>>Try it:

>>oXML=newObject("Msxml2.DOMDocument")
>>oXML.async=.f.
>>cXML=[<?xml version="1.0"><row>This is my row without a question mark</row>]
>>oXML.loadXML(cXML)
>>?oXML.childNodes.length && Returns a 0
>>?oXML.childNodes(0).text && Error: No childnodes
>>* Now put a question mark in
>>cXML=[<?xml version="1.0"?><row>This is my row with a question mark</row>]
>>oXML.loadXML(cXML)
>>?oXML.childNodes.length && Returns a 2
>>?oXML.childNodes(0).text && Returns 'version="1.0"'
>>?oXML.childNodes(1).text && Returns 'This is my row with a question mark'
>>* Make sure to clear
>>oXML.abort
>>release oXML

>>Does it work? It loads - but it does not address the nodes as configured in this scenario.