Dennis wrote:
> I have a question regarding a couple of the attributes in the GENTECH
> Genealogical Data Model at...
>
> https://www.ngsgenealogy.org/ngsgentech/projects/Gdm/Gdm.cfm
>
> I am looking at the two attributes ASSERTION.Value (p. 47) and
> CHARACTERISTIC-PART.Characteristic-Part-Name (p. 50). Can someone
> explain (and possibly give examples) of the differences between the two?
>
> I guess what is confusing me is that the brief example in the
> description of ASSERTION.Value on p. 47 seems to be at odds with another
> more detailed example on p. 34.
>
> [[In an attempt to alleviate the boredom of winter, I am trying to
> design my own physical implementation of the GENTECH Genealogical Data
> Model. I need to get a life...]]
>
> Thanks,
>
As far as I can see the purpose of a CHARACTERISTIC would be to enable a
single ASSERTION to have multiple values (note the lower case "v"!).
"If the ASSERTION is about a characteristic like a person’s name, there
will be one CHARACTERISTIC for one to many CHARACTERISTIC-PARTs. If the
characteristic is occupation, for example, there will be one
CHARACTERISTIC-PART."
If there is only a single value then this could equally well be stored
in the ASSERTION.Value but if there are multiple values then they will
be held in a series of CHARACTERISTIC-PART.Characteristic-Part-Names.
(ISTM that Characteristic-Part-Value would be better terminology)
Whilst the simple example has a single CHARACTERISTIC-PART a name might
have more, for instance one with a Characteristic-Part-Type-Name of
FirstName and one with a Characteristic-Part-Type-Name of LastName with
respective Characteristic-Type-Names of John and Smith respectively.
I doubt I'd have modeled it in that way myself. For one thing, why
separate CHARACTERISTIC-PART-TYPE out into a single entity? I'd tend to
have a data type to represent a name/value pair. From a modeling
perspective I'd then make ASSERTION.Value a list of such pairs.
--
Ian
Hotmail is for spammers. Real mail address is igoddard
at nildram co uk
|
