On Thu, Mar 26, 2009 at 4:07 PM, Scott Kurz <[email protected]> wrote:
> Simon,
>
> Still looking over your changes.... just wanted to respond that my
> impression & working understanding has been that we assume we have
> unique operation names in a bunch of places and use String.equals()
> for operation name matching... certainly within some binding
> implementations... so making this assumption in another place wouldn't
> be a big deal.   That is, unless this code needs to be used with local
> interfaces, in which case we do have the potential for operator
> overloading, however do we really need all this interface modeling for
> local interfaces?  Not sure.....
>
> ....

Ok so from your and Raymond's comments it seems like we should be OK
with operation name matching.

>
> Also I noticed you introduced
>
>  <complexType name="WireFormatJMSDefaultType"/>
>  <element name="wireFormat.jmsDefault" type="t:WireFormatJMSDefaultType"/>
>
> I did notice OASIS is talking about "wireFormat.jmsdefault" (lowercase
> 'd'), so since we're implementing OASIS here that seems to be a better
> choice even though it's still in the Tuscany NS.
>
> Scott.
>

I didn't intent to commit this change with the T2931 changes. I'd
started playing with it this as part of our discussion on creating a
separate default wire format
(http://www.mail-archive.com/dev%40tuscany.apache.org/msg06042.html)
but that's as far as I got. I hadn't noticed that OASIS had created a
wireformat.jmsdefault but I'm happy to change the Tuscany version to
lowercase "d" if we think that makes life easier for 2.x when we
convert to OASIS spec support.

Simon

Reply via email to