W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 1999

RE: contents of src and dst XML elements

From: Yaron Goland <yarong@microsoft.com>
Date: Wed, 10 Feb 1999 13:28:05 -0800
Message-ID: <3FF8121C9B6DD111812100805F31FC0D08792ECB@RED-MSG-59>
To: "'Jim Davis'" <jdavis@parc.xerox.com>, w3c-dist-auth@w3.org
Yup.... definitely inconsistent. I remember the argument about this, I just
don't remember which side I was on. =)

Basically the argument was that any time we have a URI we should have an
HREF element wrapping it. The counter argument is that adding more and more
layers of layering (src->href->uri) made it more difficult to parse and more
difficult to implement. At various points the spec got edited one way and
then the other. I would use the space shuttle logic here. You have three
votes (the value element, the example and the text), majority wins. The
majority choose using URI not HREF so we should just use URI.

I have added this issue to my list of issues to be considered when WebDAV
goes to draft status. I'm sure Jim Whitehead will do the same. His list,
unlike mine, is actually published.

			Yaron


> -----Original Message-----
> From: Jim Davis [mailto:jdavis@parc.xerox.com]
> Sent: Wednesday, February 10, 1999 12:55 PM
> To: w3c-dist-auth@w3.org
> Subject: RE: contents of src and dst XML elements
> 
> 
> At 11:42 AM 2/10/99 PST, Yaron Goland wrote:
> >Jim, the source of your confusion is that the spec doesn't 
> do a good job of
> >defining how we use the value argument in defining XML elements.
> 
> No, the source of my confusion is that the spec uses 
> inconsistent language.
>  The description of the link XML element clearly says the src 
> and dest XML
> elements contain href XML elements, not PCDATA.
> 
>    Description: The link XML element is used to provide the 
> sources and
>    destinations of a link.  The name of the property 
> containing the link
>    XML element provides the type of the link.  Link is a multi-valued
>    element, so multiple links may be used together to 
> indicate multiple
>    links with the same type.  The values in the href XML 
> elements inside
>    the src and dst XML elements of the link XML element MUST NOT be
>    rejected if they point to resources which do not exist.
> 
> If you agree that this is inconsistent, then why not change it?  
> 
> I suggest we just drop the last sentence altogether.
> 
> We don't need the prohibition about not rejecting the property if the
> resources don't exist because:
> 
> 1. It's just common sense.  WebDAV applications shouldn't routinely
> dereference these URIs (in order to check whether they 
> exist), and indeed
> there might be bad security implications if they did.
> 
> 2. The prohibition applies to all places src and dst might be 
> used, not
> just in link.
> 
> 3. It applies in general to other properties whose value is a 
> URI.  Lock
> tokens are URIs one can't dereference.
> 
> If you insist on keeping it, then move it to the descriptions 
> of src and
> dst, to read:
> 
>    The src/dst XML elements  MUST NOT be
>    rejected if the URI they point to  does not exist.
> 
> unless there really is some reason that this applies only to their use
> within a link element, and not elsewhere.
> 
> 
Received on Wednesday, 10 February 1999 16:56:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:49 GMT