W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > September 2003

Re: collecting objections

From: pat hayes <phayes@ihmc.us>
Date: Fri, 26 Sep 2003 20:27:35 -0500
Message-Id: <p06001f2dbb9a734534f2@[10.0.100.25]>
To: Dan Connolly <connolly@w3.org>
Cc: w3c-rdfcore-wg@w3.org

>On Thu, 2003-09-25 at 08:42, Brian McBride wrote:
>[...]
>>
>>  My present inclinitation is to ask I18N whether they have decided to
>>  raise a formal objection and if so, to refer us to publicly accessable
>>  text describing their objection to which we can refer in the last call
>>  announcement.  That would seem to resolve any ambiguity and gives I18N
>>  the opportunity to clearly explain their concerns.
>
>Martin gave us a pointer; I called him to confirm that this
>is a reasonable thing to use as the position of the I18N WG,
>and he confirmed.
>
>I copied it to
>http://www.w3.org/2003/09/ri434.html
>
>and updated the status from saying "draft, we will
>work on this more" to just...
>
>[[
>This document is perhaps not as polished as one would like. Some
>sections are rather well worked out, others are not. A lot of links
>could be added.
>]]
>$Date: 2003/09/26 14:32:11 $

So what do we do about this? I think some of the points made are 
reasonable, some are beside the main point for RDF, and in some cases 
seem to be based on a fundamental misapprehension about the whole 
intended use case involving the SW concept (they would be reasonable 
if RDF was intended as a language to be readable by human beings); 
and that the document fails to grasp the basic problems which arise 
from combining language tags with typed literals. In fact it seems to 
be unaware of the whole concept of a typed literal or of datatyping 
in general.

I would note that the document cites me at one point to the effect 
that I could be ready for any decision made. This may be misleading 
taken out of context. That was a remark to the WG indicating that it 
was *possible* to adapt the MT to any of the alternatives, ie a 
message to the effect that I could not see any fatal technical 
objection arising from the formal semantics for any of them, and 
could make he editorial changes corresponding to whatever decision 
was taken by the WG.  (In the past, such fatal objections have often 
risen, so the WG has come to rely on such all-clears for decisions 
with semantic consequences, before even discussing the merits of 
rival designs.) It also was symptomatic of a certain weariness with 
arguments about typed literals, and reflected the fact that I had 
already written most of them out at one time or another. 
Nevertheless, the differences between the various designs is acute 
and has far-reaching consequences for the semantics and for 
implementations of inference engines, and knock-on effects for other 
systems such as OWL.  My comments were not intended to be an 
endorsement of all the alternatives, or a vote that they were all 
equally good designs.

However, I have to say, looking at the email messages cited, 
particularly from Patrick, it seems that in May, Patrick approved 
strongly of the design where XML literals were syntactically 
distinguished and not considered typed literals; but when I 
re-suggested that we revert to this design in my 'wet fish' message 
after Martin raised this issue - using, I now see, essentially the 
same arguments that Patrick had used (typing and lang tags don't mix: 
XML needs lang tags; ergo, XML literals are not typed) it was Patrick 
who was most vehement in raising objections to it. Funny old world, 
ennit?

Pat



-- 
---------------------------------------------------------------------
IHMC	(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32501			(850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes
Received on Friday, 26 September 2003 21:29:08 EDT

This archive was generated by hypermail pre-2.1.9 : Friday, 26 September 2003 21:29:23 EDT