Re: Sustainable Codes vs Volatile URIs Re: URIs / Ontology for Physical Units and Quantities

On Thu, May 7, 2015 at 9:44 AM, Thad Guidry <thadguidry@gmail.com> wrote:

> Hi Bernand and friends!
>
> This old debate of : Strings versus Things (and how do we make them useful
> and reconcilable)
>
> It seems that everyone on this thread has agreement:
>
> 1. Codes are useful
> 2. URIs can be volatile, (shit happens), but we want to be able to use
> them, irregardless, if we want.
> 3. Some folks prefer their Codes adapted in a certain way, to make it
> easier to program and reconcile against.
> 4. Reconciling processes are sometimes harder for a programmer or
> developer, if there are less reliable sources floating around that have
> correctly mapped Codes to our Codes.
>
> What is clear is that SOME of the data, identifier, reconciling sources
> that are on the web now.... probably could do a better job of alignment.
>

* RDFJS autocomplete fields (is this an ENUM?)
* RDFJS form validation
* (presentational) language mappings

NAAICT, RDF is statements about resources with URIs linked by predicates.

Is a literal or a URI better for linking between things (as "relational
keys")?

* URIs are (can be) dereferenced (optionally also with a human-readable
HTML representation (#here))
* URIs are namespace-unique (they reduce "collisions" and ambiguity)
  * When/if the underlying codeset changes, the URI (prefix) changes too
* URIs are URL-encoded
* Labels are for display
* Labels are [HTML,]-encoded
* Labels have @en (@language attributes)
  * These attributes describe a resource with a URI
* Labels and URIs are/can be similarly collated

http://www.w3.org/TR/ld-bp/#HTTP-URIS

> [paraphrasing: "Please use URIs for keys w/ RDF Linked Data."]




>
> The only issue that I see that needs to be fixed then... is # 4.  (and
> that can be fixed through better education and offering to help fix those
> sources, or don't use them, and use other sources)
>
>
> Thad
> +ThadGuidry <https://www.google.com/+ThadGuidry>
>
>