W3C home > Mailing lists > Public > public-xpointer-registry@w3.org > January to March 2008

RE: New Xpointer scheme proposed: smlxpath1

From: Kumar Pandit <kumarp@windows.microsoft.com>
Date: Thu, 6 Mar 2008 18:00:52 -0800
To: Philippe Le Hegaret <plh@w3.org>
CC: "Henry S. Thompson" <ht@inf.ed.ac.uk>, "public-xpointer-registry@w3.org" <public-xpointer-registry@w3.org>, Michael Sperberg-McQueen <cmsmcq@w3.org>, "C. M. Sperberg-McQueen" <cmsmcq@acm.org>, Pratul Dublish <PRATULD@microsoft.com>, John Arwe <johnarwe@us.ibm.com>, public-sml <public-sml@w3.org>, Kumar Pandit <kumarp@windows.microsoft.com>
Message-ID: <0E4B701746DBB9418A6747852F8B9CA13837BC80C7@NA-EXMSG-W602.wingroup.windeploy.ntdev.microsoft.com>

Hi Philippe,
The SML working group had consensus on using xpath 1.0 location paths in the smlxpath1() scheme. The focus of this decision was to define the functionality that adequately met the needs of SML. The group did not exhaustively debate every case outside this set.

Regarding the 2 cases you mentioned,
#1 can be achieved as a predicate expression which is allowed in location paths. This assumes that one knows the name of the ID attribute.
#2 was disallowed right from the beginning. That is, the member submission draft specifically excluded it from the profile of xpointer() scheme it defined.


-----Original Message-----
From: Philippe Le Hegaret [mailto:plh@w3.org]
Sent: Tuesday, March 04, 2008 11:33 AM
To: Kumar Pandit
Cc: Henry S. Thompson; public-xpointer-registry@w3.org; Michael Sperberg-McQueen; C. M. Sperberg-McQueen; Pratul Dublish; John Arwe; public-sml
Subject: RE: New Xpointer scheme proposed: smlxpath1

Hello Kumar,

The xpath1 scheme indicates that it works as the S. St.Laurent proposal,
which means it takes an XPath expression, as you outlined in your third
point. The draft indicates:
Evaluating the expression contained in xpath1schemedata in that
namespace and document context returns a location-set or an error. [...]
If the result of the XPath expression is something other than a node or
node-set, the XPointer result is an error.

I was wondering if an Expr can return a node or a node-set and found out
that it's possible.

So now, I'm wondering if the SML Working Group considers the following
two cases:

1. #smlxpath1(id("foo"))
2. #smlxpath1(//a[1] | //a[3])

Since the smlxpath1() parameter is a location path, the two cases above
are disallowed, while being allowed in xpath1().

Is that the real intention of the SML WG or were those cases overlooked?



On Mon, 2008-03-03 at 20:36 -0800, Kumar Pandit wrote:
> Hi Henry,
> Thank you for your email asking for clarification on the smlxpath1() scheme..
> The smlxpath1() scheme is defined to meet the specific needs of the SML specification. It uses only a subset of the full xpath1 functionality. To be specific, it only allows xpath1 location paths to be used. The SML working group decided to define smlxpath1 scheme for the following main reasons:
> 1. The SML group did not want the SML specification to take normative dependence on xpointer schemes that are not W3C recommendations. The xpath1() scheme is not a W3C recommendation.
> 2. The xpointer schemes that is a W3C recommendations (namely, element()) does not meet the needs of SML.
> 3. The xpath1() scheme allows any xpath expression. The SML working group did not want that. The group wanted only xpath location paths.
> 4. The smlxpath1() scheme inherits namespace bindings from the containing element which greatly reduces verbosity of SML models because xmlns() portions do not need to be added to each SML reference.
> It appears that Syd's email is seeking clarification and is not a formal objection. I hope the explanation presented above provides sufficient clarification. Were any formal objections to smlxpath1() scheme received within the two week public review period?
> I am not clear on the precise nature of the concerns about the smlxpath1() scheme causing dilution of the effectiveness of the xpath1() scheme. If it is the scheme name that is at the root of the confusion, the SML working group is open to changing the name to something like sml() scheme or the smlref() scheme (or any other name that you may suggest). Alternatively, since smlxpath1() scheme is for the specific needs of SML, would you recommend some other change to its registration to make it more acceptable to you?
> Kumar
> More info:
> 1. Email messages that may provide more context on this:
> a.      http://lists.w3.org/Archives/Public/public-sml/2007Nov/0264.html
> b.      http://lists.w3.org/Archives/Member/w3c-xml-cg/2007Oct/0021.html
> 2. The full definition of smlxpath1() scheme from the SML specification (http://www.w3.org/TR/2008/WD-sml-20080114/#SMLXPath1_Scheme).
> -----Original Message-----
> From: Henry S. Thompson [mailto:ht@inf.ed.ac.uk]
> Sent: Thursday, February 28, 2008 12:19 PM
> To: public-xpointer-registry@w3.org
> Cc: Kumar Pandit
> Subject: Re: New Xpointer scheme proposed: smlxpath1
> Hash: SHA1
> Syd Bauman writes:
> > How is this [smlxpath1] different than xpath1()?
> The registrar invites discussion of this registration -- I am tending
> towards rejecting it, on the grounds that it  unnecessarily duplicates
> an existing scheme and would dilute the effectiveness of both the
> existing scheme and the new scheme if it were allowed.
> ht
> - --
>  Henry S. Thompson, HCRC Language Technology Group, University of Edinburgh
>                      Half-time member of W3C Team
>     2 Buccleuch Place, Edinburgh EH8 9LW, SCOTLAND -- (44) 131 650-4440
>             Fax: (44) 131 650-4587, e-mail: ht@inf.ed.ac.uk
>                    URL: http://www.ltg.ed.ac.uk/~ht/
> [mail really from me _always_ has this .sig -- mail without it is forged spam]
> Version: GnuPG v1.2.6 (GNU/Linux)
> iD8DBQFHxxclkjnJixAXWBoRApp0AJ4v/PsEgd3whNuASNGafzpLEkioyQCfcP0K
> PZVDqQrV/kjur7Mh1sSCzL0=
> =OddC
Received on Friday, 7 March 2008 02:06:51 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:38:56 UTC