Re: In Solid apps, let’s stop using RDF terms that 404

čt 22. 5. 2025 v 9:08 odesílatel Michiel de Jong <michiel@unhosted.org>
napsal:

> In yesterday’s CG meeting we discussed how hard and slow it is to publish
> RDF terms. While other semantic web practitioners may have a different
> consideration of the trade-off, I proposed that we as the Solid project
> should make sure that developers of Solid apps can always publish a new
> term in at most 20 minutes.


Great call yesterday, it was a breath of fresh air to have a technical
discussion with diverse viewpoints that remained constructive throughout.

We made real progress. I have a lot of confidence in the direction being
taken by the current chairs, along with Jesse and the ODI team. And I
forgot to say yesterday: many thanks to Hadrian for scribing.

I really like this idea: “developers of Solid apps can always publish a new
term in at most 20 minutes.”

(Side note: the server may not deliver a 404, or any other status code.)


>
> There are a few terms that we been using without proper dev docs for
> months or even years. Let’s make sure we fix those one by one.
>
> For newly proposed terms let’s advise app devs and spec devs to publish
> them at the same time they publish their app or spec, using something like
> webId.org instead of leaving them undocumented for months or years, stuck
> in unreviewed PRs or in unmaintained vocabs.
>

+1 you mean w3id.org + alternatives?


>
> Of course, apart from publishing their terms, app and spec devs should
> also publish their shapes and other data conventions.
>

+1 I think we agreed that logically terms come before shapes?


>
> Solid Data Modules, shape repos, inference and lenses can then help to
> achieve interop.
> But without dev docs for our RDF terms we can’t even start to properly
> work on those.
>

+1


>
> We had a good discussion about this yesterday  with Pavlik, Jesse, Melvin,
> myself, and other participants.
>
> The hope is that this proposed new recommendation will lead to a better
> developer experience and better interop between apps.
>
> Is there anyone who couldn’t attend yesterday but would like to chime in
> before we go ahead with this?
>

Perhaps it would be helpful to first outline different approaches, I've
suggested one here [1], and then list the trade-offs for each.

Thanks again for following up on this.  I think this kind of cleanup of the
devX is long overdue.

[1] https://lists.w3.org/Archives/Public/public-solid/2025May/0006.html


> Cheers,
> Michiel
>