Re: Comments on NF&R

On 9 Feb 2009, at 22:07, Christine Golbreich wrote:
[snip]
> On 2009/2/9 Bijan Parsia <bparsia@cs.manchester.ac.uk> wrote:
>> Not to sound like a broken record, but if:
>>        http://www.w3.org/2007/OWL/wiki/ 
>> New_Features_and_Rationale#F9:_Key
>> is the text, it still fails to address the specifics of Jim's  
>> comment.
>>
>> It also doesn't address the rationale for non-functionality.
>>
>> I note that it's quite a mess too, with repetition and a lot of  
>> emphasis on stating *what* the features are rather than explaining  
>> *why* they are
>>
>> It needs a major rewrite and from someone who understands the  
>> rationale.
>
> ? !!

Well, I presumed, perhaps wrongly, that you didn't include various  
rationales because you were unaware or didn't quite understand them.

>> I volunteer, but not right now. It's hardly pressing as this  
>> document is not in LC.
>>
>> I still have many concerns with the document duplicating text from  
>> other documents (e.g., the grammar productions). It would be nice  
>> if we would discuss it at some point. I'm happy for there to be an  
>> issue.
>
> 1) The point raised by Jim was "why the last axiom wasn't entailed  
> by the HasKey" in the example. This  question is about the feature  
> itself not about "why" this feature.

I never said it was. I said it was about WHY *this aspect* of the  
feature. The current text does not explain why this aspect of the  
feature (unlike the text added to the syntax). I provided text that  
explains why this aspect of the feature, but it was not incorporated  
in any way.

>  He explicitely asked for more documentation to prevent other users  
> to be surprised by non intended inferences due to the sepcificities  
> of hasKey.

Yes, which the Syntax changes address but the NF&R does not.

> Therefore, I updated the NF&R section regarding key accordingly.

As I've pointed out several times, you have not address the rationale  
for this aspect of the feature. I'll happily explain again, if you'd  
like.

> We had agreed

I would appreciate that you cite precisely the agreement in the minutes.

> at the following TC that we will add more documentation on HasKey in  
> the different documents and send it in the reply.

I do not believe that was, in this form, resolved by the WG.

> 2) It was agreed by the WG

Actually, I do not believe the WG has made such a decision.

> that the New Features and Rationale will provide two aspects for  
> each feature: the "what" in subsection 'Feature' (an understandable  
> description of the feature) and the "why" in subsections  
> "theoritical perspective" and "implementation perspective".

Uhm...this was my proposal. Though, more precisely, the first part was  
to have been a "rationale from the user perspective" *NOT* another  
explication of the feature. This is a strong point of disagreement.

> I repetitively asked for contributions on the latter before the FPWD.

I'm not sure what that timing has to do with anything. I made several  
proposals that have influence the shape of the draft.

> Boris and others have volunteered (but not you).

I have provided extensive commentary about the document. I have  
proposed text. I'm not sure what relevance any of that has, though.

> Finally, after having complemented these subsections (thanks), they  
> said and it was agreed that it was fine like it.

I'm not sure what that has to do with me.

> 3) Thanks for the "mess"!  (note that you added repetitions with  
> earlier text in the Syntax as well. I did not mention it to move  
> forward.)

Repetition alone does not cause problems. If you look at the key, you  
copied a large chunk of text in the theoretical section from the  
original proposal and it doesn't fit in well.

In any case, if you would like to point out problems with the Syntax  
document, please go ahead.

> Hopefully other comments from external people, users but not only,  
> show that they do appreciate documents writen to be read by  
> 'humans', and that the overview (of "what") provided by NF&R is  
> required to help the approach of 'nicer' documents.

This doesn't follow. I agree that the NF&R is a useful document and  
have said so (e.g., at the last F2F).

> As far as I can see there has been many LC comments on the features  
> themselves but none asked  for a better explanation of the "why".

I am.

And, uhm, yes they have. There are a slew of comments claiming that  
various features are "unmotivated". These are from people who focused  
on the NF&R document. Some of this is unexpected by everyone (I  
certainly didn't think the XML Syntax needed an explicit section in  
the NF&R! or anywhere, really), some is due to the relatively immature  
state of the draft (it started much much later than the other  
documents, and has undergone great improvements; I'd like to improve  
it further...why is that a problem?)

> Looks like it's your single view.

That's irrelevant in a consensus situation. It's also irrelevant in  
determining the merits of my position.

>  Until now, you are the only one having concerns with the NF&R.

And so? I've been trying to read it carefully and thoughtfully. I'm  
sorry that's not what you want from me.

> You still have the opportunity to provide a more extended  
> explanation of the "why" in the Primer as well.

There will be, but that's a separate document. And the primer isn't  
about providing rationales for new features.

> BTW I'm still looking for  the revision of the Primer which has been  
> delayed several times.  If I remember correctely, an updated version  
> was expected early January to be reviewed.

Yes that did not happen. Fortunately the reason for that deadline (to  
engender more LC comments) has been mooted (given the flood :)). The  
primer will be updated in due course. I'm primarily waiting on the OWL  
API update so I can fruitfully make changes to the example.

Of course, that's irrelevant to my comments on NF&R. So I'm not sure  
why you bring it up.

At the moment, I've been trying to work on LC comments, which are more  
pressing, I think.

I'm sorry you feel that you can simply ignore my comments until I get  
exasperated enough to sharpen my tone. But none of my comments have  
been out of line, afaict, out of order, or, even, unhelpful.

I've note contributed more text because, thus far, you've ignored my  
contributed text.

Pulling back to the point, I do not believe the NF&R document current  
addresses any last call comments on keys.

It would help, imho, us to reach some level of consensus if you would  
examine the text I provided which I believe *does* address Jim's  
comment and concern.

It would also help if you weren't so touchy about the document. I'm  
offering my comments in good faith and intend them to be constructive.  
I apologize for my frustration at your ignoring them to bleed through  
to my tone here, but I hope you can put that behind you.

Cheers,
Bijan.

Received on Monday, 9 February 2009 22:33:53 UTC