Re: Proposal for hierarchical fault codes

----- Original Message -----
From: "Jacek Kopecky" <>
To: "Martin Gudgin" <>
Cc: "Rich Salz" <>; "XML Protocol Discussion"
Sent: Thursday, November 08, 2001 8:11 AM
Subject: Re: Proposal for hierarchical fault codes

> Gudge, how would we report RPC faults? Would it have to be
> something like (I'm using your latest flat approach, but it can
> be rewritten into any other)
> <faultcode>
>   <value>env:Client</value>
>   <sub>rpc:MethodNotFound</sub>
> </faultcode>

Yes, it would look something like this

> Or would the enum contain all the faults defined by our spec?

I thing the top level enum should only contain the QNames defined in part 1
but I'm open to argument.

> This would be awkward since RPC is an optional part, so probably
> the former is the way. This could be a nice demonstration in-spec
> of how the faultcodes are meant to work. 8-)

I think we're broadly on the same page


Received on Thursday, 8 November 2001 11:30:18 UTC