Re: your request to reopen 5.26

Jos' email convinced me that I should consider reopening 5.26 -- and 
it sure would be nice to be able to have at least the bnodes for 
individuals (it's used a lot in foaf).  On the other hand, there does 
need to be some actions taken to make sure Jeremy's proof is now 
correct, that the changes won't impact other parts of S&AS, etc.  I 
am very reluctant to hold up CR, and a number of WG members have 
indicated (in the straw poll and in proxy votes I've received) that 
they want to move ahead.

  I thereby propose the following compromise in the interest of 
reaching our best solution to this issue without holding up the CR 
vote:

=====
3 part compromise:

1)First , I would propose to reopen issue 5.26 and issue 6.1 and 
agree to take an ACTION on the chairs to place this as the primary 
agenda item on the next telecon (Aug 14).

2 - then, as te process document, as quoted by Jeremy earlier, states:
>  In the Call for Implementations, the Working Group MAY identify specific
>  features of the technical report as being "features at risk."

I would suggest that we move to CR and identify the current 
restriction that bnodes cannot be used as in B1 as "at risk" -- that 
is, we state we may remove this restriction.

3 - Prior to the CR vote, I would change the CR request to read:

  The issue of mapping RDF graphs to OWL DL was discussed throughout 
the deliberations of the Working Group, but particularly at our 
January f2f in Manchester, UK and thereafter. A task force met at the 
Technical plenary in March and proposed a 20-point compromise. 18 of 
these points were closed with consensus, 2 others, concerning the 
treatment of two cases of b-nodes in RDF were closed over the dissent 
of Carroll/HP  who argued that the mapping of RDF to OWL should 
include these cases. His concern was echoed by the Jena team and by 
the RDF Core WG during the Last Call. On March 27, the WG agreed to 
close this issue over the objection. On July 12, the working group 
voted to create an issue entitled Unnamed Individual Restrictions to 
indicate our desire for such a feature, but had to close this as 
"Postponed" due to our inability to work out a solution that would 
not require a major rework of the direct semantics of OWL DL (Section 
3 of our Semantics Document).   However, there is a possibility that 
a partial solution could be found, and thus we are identifying this 
particular restriction as  a feature of the Semantics TR that may be 
removed during the CR period.    It is our hope that such a 
compromise solution would remove this objection and let us close the 
issue to the satisfaction of Carroll and the others.  However, we are 
not yet sure that such a compromise is semantically possible and do 
not wish to hold up the CR period in case we, once again, fail to 
find a solution. Thus  we ask that the Director accept our decision 
to Postpone this issue, and to confirm the WG decision despite the 
outstanding dissent.

and add a section called to the CR entitled

"Features at risk"

in which we would put a pointer to this restriction and to the issue 
on the issues list (I would need help from S&AS editors to identify 
correct pointer to the specific "at risk" restriction).

=========
I believe this would let us avoid holding up CR so implementors could 
start going, would let us resolve this issue once and for all in the 
next couple of weeks without unduly refusing to consider Jeremy's 
suggestions, and would not violate any W3C process.

  I welcome your howls
  JH


-- 
Professor James Hendler				  hendler@cs.umd.edu
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  *** 240-277-3388 (Cell)
http://www.cs.umd.edu/users/hendler      *** NOTE CHANGED CELL NUMBER ***