W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > October to December 2005

Comments in XML-valued dead properties

From: Lisa Dusseault <lisa@osafoundation.org>
Date: Wed, 28 Dec 2005 10:13:37 -0800
Message-Id: <5cd6cb95354036ea59c277750774a89b@osafoundation.org>
To: webdav WG <w3c-dist-auth@w3.org>

We've had some discussion of the preservation of comments in XML-valued 
properties.  I believe we came to consensus on what Geoff pointed out, 
that when the XML-valued property is a live property, the server has 
very good reasons not to preserve comments -- the live property that is 
writable can be considered a configuration setting, the value of which 
the server uses to affect its own behavior.  There may be other ways 
the server wishes to normalize live XML properties (e.g. replacing 

But at any rate, I had wondered if we consider dead properties to be 
different.  A dead property is used for a client to set information 
that it can use later, or for a client to set information that other 
clients can use.  Some dead properties are even for human consumption 
(perhaps with some processing).  Thus, it's quite possible for clients 
to have a use case where the comment is important.  Following this line 
of reasoning I added to the -09 draft some "test balloon" text:

    "In dead properties (considered as content, like document bodies)
    servers are encouraged to (MAY) preserve, for any Comment Information
    Item in the value:


Julian's the only one who has commented on this and proposed removing 
that text which I admit is rather weak as it tries to land somewhere 
between MAY and SHOULD.  Saying nothing about comment preservation 
would have clients unable to rely on that:

    "XML Infoset attributes not listed above MAY be preserved by the
    server, but clients MUST NOT rely on them being preserved."

What do y'all prefer?
  - Servers SHOULD preserve comments
  - Servers are encouraged to preserve comments
  - Say nothing about comments thus clients MUST NOT rely on them
  - Other... ?

Received on Wednesday, 28 December 2005 18:14:07 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:01:34 UTC