On Jul 25, 2006, at 11:26, Andrea Preite Martinez wrote:
> Quoting Nausicaa Delmotte <ndelmott-at-eso.org>:
>
>
>> - What is the difference between 'title of an observation proposal'
>> and 'name of an observation proposal'?
>>
>> So far I would not have expected any difference but with the
>> introduction
>> of the new UCD word obs.proposal, it seems it is possible to
>> construct both
>> meta.title;obs.proposal
>> and
>> meta.id;obs.proposal
I would personally use:
meta.title;obs.proposal for the title of a proposal meta.id;obs.proposal for the proposal id
I do not understand what "name of an observational proposal" actually
means.
Where did you get this one from Nausicaa?
>> Maybe someone could provide an example, in case there is a
>> difference
>> of meaning
>> between those two UCDs? I just cannot see the difference... (given
>> that the proposal
>> code is already defined by meta.code;obs.proposal).
>>
> Don't forget catalogues! According to the Catalogue DM, a catalogue
> can have:
>
> - a name (in the description text)/a title (in the XML schema)
> - an identifier
> - a shortName
>
Yes, in the case of catalogues there is a third quantity to consider, the shortname, which does not exist for proposals.
What should we be using here?
catalog title -> meta.title;meta.table
catalog id -> meta.id;meta.table
catalog shortname -> ???
One could think of using meta.id;meta.table also for the shortname
given that shortname is an alias of the catalog id (which happens to
be just
more human-readable). But I am against that for the reasons given
below...
> My point of view:
> - The title of XYZ is a string describing the content of XYZ, as in
> a scientific paper (actually this is the origin of the ucd);
> - the name/identifier is a string uniquely identifying XYZ, as for
> humans (e.g.: Andrea Preite Martinez is my name/identifier);
To me names and identifiers are quite different. The name might not be unique, while an identifier better be.
In the human case, the name is not unique, but the name with the
birth certificate
makes that person unique. What is unique is the identifier assigned
to such
human when registering to whatever service, e.g. the social security
number.
An identifiers will differ for the same human within different services.
Alike, a catalogue which is colloquially called Veron 2006, will have different IDs in different services.
I think a meta.nickname could be useful here.
Aside note: I would drop meta.dataset in favour of meta.id;obs
> - a short name for me could be APM (and my "code" could be my
> social security number!).
> I don't see the difference between name and identifier.
meta.code is described as: "Code or flag"
Given that meta.id to me is the identifier, I am pushed to think
that Code is not an identifier. Hence I always interpret meta.code as
a flag.
As such, as a UCD user, I am not going to assign the ucd meta.code to
a social security number.
I guess that after this email,
I will be invited not to assign ucds any longer... ;-)
But the truth is that the UCD assignment is a highly subjective matter.
> Another point is that a "possible" ucd is not necessarily a "good"
> ucd!!
And how can a user tell wether her choice resulted into a "good" ucd?
Alberto Received on 2006-07-25Z13:52:01