W3C home > Mailing lists > Public > public-rif-wg@w3.org > September 2009

Re: DTB comments from XML Query

From: Dave Reynolds <der@hplb.hpl.hp.com>
Date: Mon, 28 Sep 2009 18:44:39 +0100
Message-ID: <4AC0F607.3070405@hplb.hpl.hp.com>
To: Sandro Hawke <sandro@w3.org>
CC: public-rif-wg@w3.org
Sandro Hawke wrote:
> I put the comments from XML Query WG (via Jim Melton) on the wiki, and
> added my thoughts on each of his points:
>   http://www.w3.org/2005/rules/wiki/Response_to_JM
> Issues:
>   * Why are we using a different namespace?  (I don't remember)

I don't recall a formal decision on this but my rational reconstruction 
is that we were striving to have a simple uniform namespace for our 
users.  We use operators as well as functions and in RIF the operators 
become functions too.  Since XQuery deliberately don't offer a namespace 
for op we would have to make that up. Given that it seems easier to have 
them all in a single RIF namespace rather than always having to remember 
if this one is a rif-fn: or xq-fn:.

>   * We should explain negative guards a little; without some explanation,
>     they look buggy.


>   * Change concat to require at least 2 args, like xpath?


>   * String subscripts are numbered from 1; we made lists numbered
>     from 0.  Maybe those should match?  Ugh!  (I've hated this issue
>     for 31 years now.  Why stop now?)

Ugh indeed but since its true that fn:substring starts uses numbering 
from 1 (why?!) we ought to be consistent and change lists.

>   * Stop using the hollow square box


> There are about a dozen other points, but on the rest I think we can
> just do what the commenter asks.  (I would like a second pair of eyes on
> them, though.)

The hard one is type promotion. We are assuming type promotion for the 
numeric operators and I believe Jim that the clearest way to read the 
specs is that type promotion is done in the language and the operators & 
functions only handle the promoted types. That suggests we need to add a 
section on type promotion in RIF .

Received on Monday, 28 September 2009 17:45:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:47:56 UTC