With XmlSchema 1.4.1 we are seeing a regression in the Woden test results
for the W3C WSDL 2.0 test suite. We see 4 test cases fail which passed
successfully when Woden was last tested with XmlSchema 1.3.1.  It seems to
be related to chameleon includes (including a schema that has no TNS). In
1.3.1, no target namespace was represented in the XmlSchema object by a
null. In 1.4.1 no target namespace is represented by the empty string "".
This change in the representation of a missing target namespace is causing
the regression in Woden.  I think the 1.3.1 behaviour was correct.  I don't
think a missing TNS should be represented as an empty string. Can any
XmlSchema devs shed more light on this?

thanks,
John Kaputin.

On 4/10/08, Eran Chinthaka <[EMAIL PROTECTED]> wrote:

> +1 from me too.
>
> Thanks,
> Eran
>
> On Thu, Apr 10, 2008 at 9:42 AM, Glen Daniels <[EMAIL PROTECTED]>
> wrote:
>
> > Hi dims:
> >
> > Davanum Srinivas wrote:
> >
> > > In the interest of time, i propose we make a joint vote for all 3
> > > commons projects (XmlSchema/Axiom/Neethi). Then woden
> > > can pick up xmlschema/axiom and make their release. Then we cut Axis2
> > > 1.4 FINAL.
> > >
> > > Sounds like a plan?
> > >
> >
> > +1, sounds like a plan.  I'll grab the latest and kick some tires.
> >
> > --Glen
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>

Reply via email to