RE: Hierarchical fault codes and RPC from Henrik Frystyk Nielsen on 2002-01-19 (xml-dist-app@w3.org from January 2002)

From: Henrik Frystyk Nielsen <henrikn@microsoft.com>
Date: Sat, 19 Jan 2002 08:39:24 -0800
To: "Marc Hadley" <marc.hadley@sun.com>, <xml-dist-app@w3.org>
Message-ID: <79107D208BA38C45A4E45F62673A434D05A53DAD@red-msg-07.redmond.corp.microsoft.com>

I agree that RPC should use the subcode mechanism and not be part of the
"core" set.

Henrik

>While adding the resolution of issue 173[1,2] (hierarchical 
>fault codes) to the specification I noticed a problem in part II.
>
>Previously the value of a fault code element was a QName and 
>the RPC section took advantage of this to introduce two new 
>RPC specific error codes. The resolution to issue 173 makes 
>the values of the faultcode element an enumeration with a 
>closed set of values.
>
>We have two choices:
>
>(i) add the RPC specific codes to the enumeration in part 1
>
>or
>
>(ii) change the RPC section to mandate use of a specific fault 
>code value from the existing enumeration and use the new 
>subcode facility to hold the RPC specific fault code.

Received on Saturday, 19 January 2002 11:39:57 UTC