Re: Draft report and use cases

Quoting Antoine Isaac <aisaac@few.vu.nl>:


> By "use cases" do you mean here "clusters"? Of course I agree that  
> the individual use cases should not be in the report, except perhaps  
> as a list in the appendix document(s).
> As for the clusters, I am still convinced that we can still get some  
> stuff from the "Scenarios (Extracted Use Cases)" sections of the  
> cluster to populate the "use cases" section of the report. And from  
> the "Vocabularies and Technologies" and "Problems and Limitations"  
> sections, so as to populate sections 1.4 and 1.5 of the report [1].  
> Even though here they could just be used as background, as you  
> suggest.

Basically, I see the use cases as providing information for the report  
-- but I'm not sure that there is a place for the clusters or cases in  
the report itself. To me the uses cases serve to identify *issues* by  
providing some examples of projects people are attempting or  
interested in. But those projects are just examples -- and not  
necessarily representative or complete. As we develop the document it  
may be useful to use some of the use cases as illustrations, but  
unless they have a specific role in our exposition, I wouldn't try to  
force them into the report.


> OK. Correct me if I'm wrong: what you suggest amounts more-or-less  
> to getting back to  
> http://www.w3.org/2005/Incubator/lld/wiki/TopicsDiscussed ,  
> identifying priorities there, and trying to get some text for them,  
> doesn't it?
> I can't think this long list missed an important issue of our domain :-)

I haven't looked back at that list for a while, and it would be good  
to look at it again. Some of it is pretty minimal (e.g. the entirety  
of our entry on MGT.GUIDANCE is two words: "Strategic guidance." To be  
useful, we probably need to flesh that out a bit :-)) But a list is  
not a conclusion, so we need to decide what are the most important  
messages we can convey. There has to be some sense of priorities  
because all of our lists are too long to be "incubated" in their  
entirety.

One possible exercise would be to look at the topic list or the issues  
that come out of the use cases and put them into a functional order:  
for x to happen, what has to be in place? As an example, let's say  
that we feel that an issue is that we need more complete bibliographic  
vocabularies. For that to happen, certain institutions need to step up  
and create RDF for vocabularies that they own and maintain. For THAT  
to happen, those institutions need to see that as a priority, and that  
priority needs to be funded. Then, is there some requirement that must  
precede that? Maybe some standards about vocabulary creation,  
versioning, maintenance. etc.

My view is that our report needs to be grounded in reality so that  
there are actual steps that can be taken toward LLD. That means we  
can't just say WHAT we think should happen, we have to at least make  
some attempt to recommend as much of the HOW as we can. Our  
recommendations will have gaps, of course, but we need to show that  
what we are recommending is possible.

kc

>
>
>> I don't see how we can produce a report that doesn't address these  
>> core issues.
>
>
>
> Me neither!
>
> Antoine
>
> [1] http://www.w3.org/2005/Incubator/lld/wiki/DraftReport
> [2] http://www.w3.org/2005/Incubator/lld/wiki/DraftReportReviewerAssignments
>
>



-- 
Karen Coyle
kcoyle@kcoyle.net http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet

Received on Thursday, 17 February 2011 20:12:43 UTC