[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: collection vs dataset EOD? Re: orthogonality (was Re: New attempt)



Umm. Wasn't Joe saying that the difference in meaning between datasets and
collections is that collections contain related objects and datasets contain
alternate views of the dataset or subsets of the dataset?

To show that differnce, it seems like there would have to be a difference
between <collection> elements and <dataset> elements. They would be identical
except that <dataset> elements can contain <access> elements and <collection>
elements cannot.

Ethan

Benno Blumenthal wrote:
> 
> Hi John and Joe,
> 
> Since I was asked, I am answering, not that I am adding anything useful.
> 
> Yes, if collections and datasets are completely interchangable in all
> machine-type ways, that works for me.   I think John gives the definitive
> summary below.  Of course, if I have a dataset that temporarily does not have
> any functioning access methods on a particular server, one may not always feel
> the need to relabel it a collection...
> 
> Benno
> 
> Quoting John Caron and Joe <address@hidden>:
> 
> > > If this is the case then I would suggest that
> > >
> > > a) this distinction be preserved by allowing both tags to be
> > > used(possibly renamed if it would clarify things); and
> > >
> > > b) data providers should be encouraged to mark up their catalogs
> > > appropriately using the two tags, so that THREDDS client UI's can take
> > > advantage of this to present catalogs in an intuitive way; but
> > >
> > > c) these tags should be completely interchangeable in all other ways
> > > (i.e. same type in the DTD/Schema, and same API calls, any tag that can
> > > go in a dataset can also go in a collection), since they are
> > > semantically equivalent at a machine level.
> > >
> > > Does that make any sense? Benno, would that satisfy you?
> > >
> > > - Joe (ready for a checkup with my ontologist)
> >
> 
> Quoting John:
> 
> > Actually Im inclined to take it a bit further.
> >
> > Currently a collection is just some collection of datasets that share some
> > common theme. If we allow it also to be a dataset (meaning it has a URL,
> > can
> > be selected, etc) then I think it should have the meaning that contained
> > datasets are subsets or specializations of it. Because if they are not it
> > seems to me that you might as well just represent the collection-as-dataset
> > as a contained dataset element. [Maybe in this whole discussion I have been
> > trying to convince myself of that :^] Does everyone agree with that meaning
> > of nested datasets inside of collection-as-dataset?
> >
> > PS: There are still semantic difference between collections and datasets: A
> > dataset has one or more access elements, a collection 0 or more.
> > Collections
> > contain datasets and nested collections.
> > OTOH, datasets and collections look so similar already in the XML, its
> > tempting to combine them (which i was playing with earlier in
> > http://www.unidata.ucar.edu/projects/THREDDS/xml/InvCatalog.0.6a.dtd)
> >
> >
> >

-- 
Ethan R. Davis                       Telephone: (303) 497-8155
Software Engineer                    Fax:       (303) 497-8690
UCAR Unidata Program Center          E-mail:    address@hidden
P.O. Box 3000
Boulder, CO  80307-3000              http://www.unidata.ucar.edu/