Re: Proposed response to Martin Merry, HP

On Tue, 2003-05-13 at 09:41, Jim Hendler wrote:
> Dan - afraid you and I see very differently on how to reply to this 
> -- I think this comment was mainly asking if we had done our design 
> right and if we are ready to move.  I think it is important to make 
> it clear to him that we have implementation experience (I can name 
> names - I am not just "making it up")
>   I am not trying to "hide behind process" but rather to make it clear 
> we meet the process needs.

Meeting process requirements is not a goal of mine.
Widely deployed technology is. The process is a means,
not an end.

>   However, the really important thing is his second comment.  We don't 
> have any new issue to raise - we already made our decisions in this 
> space.

Then let us cite them.

>   His email suggests a different way to parcel things out, and 
> I am explaining why we made the decisions we did

You can explain why *you* took the position you did, but
if you are to speak for the group, it seems to me that
you must speak with the group's words, i.e. the decision record.

Or get the group to agree to new words; i.e. put a
new issue in the issues list.


>  -- I think this is 
> the right thing to do, and our responsibility -- pointing him to 
> months of discussion and saying "sort this out" seems to me to be 
> very irresponsible on our part.

Right; we need to find the relevant decisions for him.

>   I am also not in sympathy with your suggestion that we write 
> sections and ask him to approve -- because that will require a lot of 
> our effort, and it may be we'll do it and then discover that we 
> weren't even in the space he wanted.  I wrote:
> 
> At 10:30 -0400 5/13/03, Jim Hendler wrote:
> >>  If this outline of a solution is acceptable to you, we will produce
> >>>  proposed text properly setting the expectations with respect to
> >>>  inverse and oneof, and request your approval before closing this
> >>>  comment.
> 
> in short - I tried to make it clear this was an interim answer, and 
> that we wanted his response before we went on.

OK, I read that too fast. Sorry.


> HP is an important contributor to the Semantic Web and I think they 
> deserve the kind of careful answer I provided - not just a "thanks 
> for the advice" blow off answer.

I don't think the answer I suggested is a "blow off" answer.


> I'm happy to work on a better answer, but you don't suggest any text 
> for the meat of his comments,

I haven't found the decisions yet. But I thought it best to
send my message lest the WG assume I agree.

>  just criticized me for "making things 
> up" -- btw, the three implementations I have in mind are Cerebra, 
> Euler, Fact, Racer,

Ah... good.

Evidence that these implementations handle some relevant
tests would make for a very compelling answer.

>  and I believe that cwm and Otter handle them as 
> well, although I have that only second hand.
> 
> So, I am left unsure what to do.

Please help me find the relevant decisions
and/or find evidence that those implementations
pass some relevant tests and/or add an
issue to the issues list.

>   I don't know how to fuse my 
> comments and yours as they seem to be aimed at different levels of 
> response.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Tuesday, 13 May 2003 11:06:48 UTC