axis-c-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From John Hawkins <>
Subject Re: HeaderBlock prefixes
Date Mon, 16 Aug 2004 09:14:24 GMT

Hi Roshan,

It appears that making an application "namespace aware" is highly
performance intensive therefore we are aware of applications that do not
bother and rely on the prefix to understand where they are. Yes, we
understand that it's wrong but we should accomodate them - at least in the
short term (next few releases?)

I like the second idea of registering the namespace/prefix - that's a much
neater interface !

John Hawkins

             <roshan@opensourc                                          To 
   >                     Apache AXIS C Developers List       
             13/08/2004 06:50                                           cc 
             Please respond to         Re: HeaderBlock prefixes            
              "Apache AXIS C                                               
             Developers List"                                              

hi all,

The earlier code base supported this "setPrefix(..)" method to be used
by the users. But then later it was removed (commented) thinking that
the prefix will be decided by the engine (the Serializer) and be added
at the time of Serializing.

But if a user really need to have his prefix (but the important one is
the 'exact namespace URI', but not the prefix), then this method could
be open for users again. What do others think?

Another approach is not to open up the setPrefix(..) method, but let the
user register the namespaceURI,prefix pair with the serializer to be
used and then the serizlier will used the registered prefix at the
appropriate places.

What are the thoughts?


On Thu, 2004-08-12 at 20:31, Fred Preston wrote:
> Hi All,
>       Has the code the allows the HeaderBlock to add prefixes to the
> localName been put back into the code base?  i.e can't create
> <prefix:localName>, <wsse:Security>, <rrbus:context>, etc...
> Regards,
> Fred Preston.

View raw message