Re: [w3c/rdf-concepts] Drop the requirement to support ill-typed literals with recognized datatype IRIs (Issue #60)

James,

See my response in the github issue: 
https://github.com/w3c/rdf-concepts/issues/60#issuecomment-2521125941

   best

On 15/11/2024 16:10, James Anderson wrote:
> yesterday, during the discussion on this topic it was argued that, because "the standard should reflect the reality", this requirement should be made optional.
>
> are there any statistics available which account for this reality?
> - how many rdf processors accept and store ill-typed literals
>    - of those which accept, how many store literal terms uniformly, that is, with no distinguished representation for known datatypes
>    - of those which accept, how many distinguish known datatypes, yet still incorporate ill-typed literals in constructed graphs
> - how many reject ill-typed literals
>    - of those which reject, how many, because it would "produce a semantic inconsistency"
>    - of those which reject, how many, because it would complicate processing an optimized term representation
> - how many rdf processors do not reject ill-typed literals, but do exclude them from constructed graphs?
>
>
> ---
> james anderson | james@dydra.com | https://dydra.com
>
>
>

Received on Friday, 6 December 2024 11:02:10 UTC