W3C home > Mailing lists > Public > xml-dist-app@w3.org > November 2001

Re: updated proposal on issue #144 - array metadata in SOAP Encod ing

From: Alan Kent <ajk@mds.rmit.edu.au>
Date: Tue, 13 Nov 2001 15:55:15 +1100
To: xml-dist-app@w3.org
Message-ID: <20011113155514.C19371@io.mds.rmit.edu.au>
On Mon, Nov 12, 2001 at 06:18:25PM -0500, Rich Salz wrote:
> It is possible to communicate sparseness between serializer and
> application, but it should not be a requirement.
> 	/r$

How can sparseness be communicated? I know how to communicate whether
data is an integer, a string, or an array - but I don't know how to
communicate within SOAP that its sparse.

It still feels uncomfortable to me that many people seem to think that
its good practice to send type information over the wire (xsi:type),
that sparse arrays are different to non-sparse arrays (as illustrated
by the comment above indicating sparseness can be communicated),
but that sparseness does not need to be communicated in a SOAP packet.
Why say 'types are good' and then say 'indicating sparseness is not
necessary'?

To be *consistent*, surely there should be a *standard* way to say
if an array is or is not sparse on the wire (like other type info),
with that information being optional (like other type info).
eg: soap:sparse="yes" / soap:sparse="no" / omitted attribute.

Alan
Received on Monday, 12 November 2001 23:55:51 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:04 GMT