Re: What to do with WoT Profiles 1.0

On Mon, 29 Apr 2024 at 08:36, Luca Barbato <luca.barbato@luminem.it> wrote:

> If you have knowledge of the profile you can process the implicit TD and
> produce an explicit TD that refers to vocabulary terms from bindings
> (that aren't existing now, but can and probably should exist), but if
> you do not have knowledge of profiles you might be confused with no way
> to exclude implicit elements, in general.
>

I hope this is not a problem, because non-conformant Consumers should
either ignore or be oblivious features they don't support.


> Ben stated that the profiles he crafted were carefully made so no
> confusion may happen, so we might try to see if nothing got overlooked
> and the situation isn't as problematic.
>

Exactly, let's double check this is the case.


>
> for node-wot since it has already a mean to address default overrides
> through subprotocols the transformation would require to add
> profile-specific subprotocols (and vocabulary terms to bind together
> coupled ops).
>

I would really like to avoid this if we can. I think it will create a mess
and result in less interoperability by default.

In the long term if we merge profile protocol bindings and binding template
defaults it won't be necessary for profiles to override the defaults
because they will be the same thing.


>
> Or we all agree that the suggested degraded consumption of async actions
> is "fire and forget" and profile Things authors have to accommodate this
> constrain.
>

I would prefer this approach, and have filed
https://github.com/w3c/wot-profile/issues/403 and
https://github.com/w3c/wot-profile/issues/405 related to this topic.

I prefer "progressive enhancement" to "degraded consumption" (which is also
the current best practice for wider web standards).

Kind regards

Ben