Re: urn:solid: for prototyping predicates

On 2025-05-22 07:29, Melvin Carvalho wrote:
> https://solid-lite.github.io/urn-solid/ <https://solid-lite.github.io/ 
> urn-solid/>
> 
> Would love any thoughts!


One will still have to "mint" URIs using `urn:solid`, and `urn:solid` is 
not an IANA registered namespace (not to mention the cost of pursuing that).

The primary use case is already covered by `urn:example` if one is set 
on minting a URI intended for temporary or testing/experimentation 
purposes, if "minting" HTTP URIs are for some reason not possible for 
said developers but somehow minting URN URIs are possible.

There is a difference between using a URI as an identifier and 
dereferencing it. If dereferencing is not needed, one can already use 
HTTP URIs *today*. Having an HTTP URI does not require it to be fetched. 
HTTP URIs can later be upgraded to resolve to URLs. Any placeholder URI, 
such as one based on `example.org` or even `localhost`, will suffice.

`urn:solid` has the same risks as any unscoped term. Any two authors 
could independently mint, for example, `urn:solid:foo`, without any 
guarantee the term means the same thing. That's known as a URI 
collision. So, coordination is required to ensure consistent meaning. 
But if developers are capable of that, they can already participate in 
the broader community to align on terminology. Hence, no need for 
`urn:solid` whatsoever.

That's all aside from the issue of promoting bad suggestions to the 
developer community like tightly coupling an application to a 
Solid-specific URI (even if that can be changed later). If someone 
chooses to spend time chasing `urn:solid` registration, writing 
(LLM-generated) documentation, or promoting its use, that effort could 
be better spent applying existing best practices. But I suppose each 
person can decide how to spend their time.

Convenience is being used here to justify deviation from established 
standards, without addressing long-term impact or technical debt. If 
someone wants to use `urn:solid` in their own Solid Lite experiments, 
that's their choice but it shouldn't impose a burden on the broader 
community.

There's no compelling evidence that `urn:solid` is lighter, better, or 
more aligned with Solid or Web architecture than existing alternatives.

So all of that put together: `urn:solid` is a distraction, a dead end, 
potentially damaging/fragmenting, and not particularly useful especially 
when alternatives already cover the use case and offer a path that's 
more robust and standards-aligned.

Just my two cents.

-Sarven
https://csarven.ca/#i