Well, there was a constraint in OM builder and OM serializing code that
every element should have a namespace and none can exist without one. I
changed that.
And there was a problem in default namespace handling as well. So fixed that
too. 

I think checking SOAPBody children will be an unnecessary perf draw back. So
shall we defer that for the time being ??

Regards,
-- Chinthaka 

-----Original Message-----
From: Sanjiva Weerawarana [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, March 08, 2005 4:49 PM
To: [email protected]; [EMAIL PROTECTED]
Subject: Re: [Axis2] OMElementImpl question

"Eran Chinthaka" <[EMAIL PROTECTED]> writes:
> Thankx Dennis. 
> 
> Will change AXIOM stuff according to that. 

Mebbe I'm missing something .. why does AXIOM need to change? AXIOM
should just do any old XML (?).

Or did you mean to say that you'll change o.a.a.om.SOAPBody to 
check that all its children are NS qualified? If so +1, as long
as it has zero perf impact .. otherwise don't bother IMO. 

IMO our decision to keep the SOAP message abstraction classes
in AXIOM will continue to cause these kinds of confusions.

Sanjiva.



Reply via email to