RE: RFC2518bis: xml:lang (2.6)

I don't buy that argument.

The XML spec clearly states what the scope of xml:lang is (the containing
element and all descendants). If WebDAV wants to override this rule, there
should be a better rational than "it's too much work implementing what the
XML spec says".

Regarding your questions:

> Please explain why it is important to be flexible in this case

I think it is important that XML-based specifications do not override XML's
rules without good reason. In this case, I haven't seen one yet.

>  Also, if the lang attribute must be legal in more than one location,
explain if there is any semantic difference between the multiple locations.

For each DAV:prop element, there's only one xml:lang declaration in scope.
It is completely irrelevant to which attribute it is attached.

> Also, explain what happens if the lang attribute is present in multiple
locations scoped to the same property, particularly if it has different
values in different locations!

Can't happen with the given scoping rules.

> If some servers will be slightly incompatible with RFC2518bis because of
making the location specific, I understand that's undesirable. However, I
think in the long run it will make interoperability more likely for this
feature, and the benefits will outweigh the costs.

I think this argument works both directions. The only problem with the
current spec is that it's silent on the issue, *possibly* causing interop
problems for people not reading the XML spec properly. The non-intrusive fix
to this is to clarify how the scoping rules defined in the XML spec apply to
property values, NOT to change those.

Julian

-----Original Message-----
From: w3c-dist-auth-request@w3.org [mailto:w3c-dist-auth-request@w3.org]On
Behalf Of Lisa Dusseault
Sent: Thursday, June 27, 2002 6:52 PM
To: Jason Crawford; Julian Reschke
Cc: Webdav WG (E-mail)
Subject: RE: RFC2518bis: xml:lang (2.6)


A specific location for the lang attribute is good.  It is more tightly
constrained, which makes it easier for both clients and servers to handle
XML with property values.

Please explain why it is important to be flexible in this case. Also, if the
lang attribute must be legal in more than one location, explain if there is
any semantic difference between the multiple locations.  Also, explain what
happens if the lang attribute is present in multiple locations scoped to the
same property, particularly if it has different values in different
locations!

If some servers will be slightly incompatible with RFC2518bis because of
making the location specific, I understand that's undesirable. However, I
think in the long run it will make interoperability more likely for this
feature, and the benefits will outweigh the costs.

Lisa

-----Original Message-----
From: Jason Crawford [mailto:ccjason@us.ibm.com]
Sent: Thursday, June 27, 2002 7:53 AM
To: Julian Reschke
Cc: Lisa Dusseault; Webdav WG (E-mail); w3c-dist-auth-request@w3.org
Subject: RE: RFC2518bis: xml:lang (2.6)

I agree with Geoff and Julian that the 2518bis wording is not right because
it specifies a specific location for that attribute.

http://lists.w3.org/Archives/Public/w3c-dist-auth/2002JanMar/0318.html

I also acknowledge that it's not easy to come up with a wording for this
behavior. Sorry Lisa :-)

J.

------------------------------------------
Phone: 914-784-7569, ccjason@us.ibm.com

Received on Thursday, 27 June 2002 13:54:50 UTC